risk management: a fiduciary's guidebook - national association of

Russell Research

Risk management:A fiduciary’s guidebookIntroduction by Bob Collie Managing Director, Investment Strategy and Consulting Russell Investments

/ p 2

Nothing contained in this material is intended to constitute legal, tax, securities, or investment advice, nor an opinion regarding the appropriateness of any investment, nor a solicitation of any type. The general information contained in this publication should not be acted upon without obtaining specific legal, tax, and investment advice from a licensed professional.

The opinions expressed in this material are not necessarily those held by Russell Investment Group, its affiliates or subsidiaries. While all material is deemed to be reliable, accuracy and completeness cannot be guaranteed. The information, analysis, and opinions expressed herein are for general information only and are not intended to provide specific advice or recommendations for any individual or entity.

Table of contentsTITLE AUTHOR(S) PUBLISHED

1 Introduction Bob Collie, FIA

3 There’s nothing normal about risk Bruce Curwood, CFA, CIMA September 2009

19 Risk management is the cornerstone of investing Bruce Curwood, CFA, CIMA February 2010

43 Prudence is process, not performance John H. Ilkiw, CFA November 1996

49 Thoughts on ERISA’s requirements regarding risk management

Bob Collie, FIA March 2010

53 Resist the Amygdala! Bob Collie, FIA and D. Don Ezra, FIA

March 2007

69 Can a DB plan provide a competitive edge? D. Don Ezra, FIA and George Oberhofer

February 2008

79 Risk management Nathan Dudley, CFA, Chris Hensel and Jennie Tyndall

February 1999

87 Narrowing the knowing-doing gap in investments through effective fund governance

Bruce Curwood, CFA, CIMA October 2006

101 Investment governance: A pragmatic update D. Don Ezra, FIA January 2010

113 Roadmap for fiduciary risk management: the importance of an effective investment management program

Janine Baldridge, CFA, CAIA October 1999

127 Mind the ‘governance gap’ Sorca Kelly-Scholte, FIA and Shashank Kothare, FIA

November 2009

153 A model agenda for an investment committee Sorca Kelly-Scholte, FIA April 2004

161 Mostly brick Bruce Curwood, CFA, CIMA and D. Don Ezra, FIA

July 2010

170 Not yet a bonfire Bruce Curwood, CFA, CIMA and D. Don Ezra, FIA

July 2010

177 About the authors

178 About Russell

Russell Investments // Risk management: A fiduciary’s guidebook

/ p 1

1 Knight, Frank H. (1921). Risk, uncertainty and profit. Hart, Schaffer & Marx, Boston. Houghton Mifflin Co., The Riverside Press, Cambridge.

INTRODUCTION

Risk management: A fiduciary’s guidebookInside this guidebook you will find 14 articles by Russell consultants and strategists that have been published over the past fourteen years on the subjects of risk management and fund governance. Some of the articles are short, others longer. Some concentrate on very specific questions, others cover broader ground. Together, they provide a wonderful education for fiduciaries charged with the oversight of a public or corporate pension plan, an endowment, a foundation, or any other type of institutional investment program.

As I write this introduction in mid-2010, few investors need to be reminded of the importance of risk management. Risk was not invented in 2008, of course, but it did begin one of its periodic spells on center stage. Risk is an elusive concept. It means different things at different times to different people. Stock market investment felt like a bleak prospect in the depths of March 2009 when valuations were little more than half their peak value and the possibility of further declines with no floor in sight felt all too real. Risk in that situation is both quantitatively and qualitatively different than in the midst of a bull market.

But we learn our lessons slowly, if at all. When you take a risk you shouldn’t (or you see someone else do it) and success follows, the behavior is reinforced. Or in reverse: after a bad experience, we can become unduly risk-averse. Risk tolerance is often a matter of emotion, not reasoning. Someone who placed their faith and their wealth in the housing and stock markets and then saw both savings and home values devastated in the financial crisis likely perceives risk quite differently from the winner of one of the state mega-lotteries, who took a less rational risk decision but saw it pay off. Their experiences will affect their future choices.

And risk is not just about the probability and impact of a particular outcome. What were the odds of a Ponzi scheme lying behind Bernie Madoff’s extraordinary track record? The question makes no sense; the scheme was there all along, but undetected. Thus a victim of that episode will have a yet different perspective on risk than our unfortunate stock market investor or our lucky lottery winner. It is clear that only some risks are neatly defined by a mathematical distribution. Others, like Madoff’s Ponzi scheme, defy measurement. This distinction was captured way back in 1921 by Frank Knight when he famously differentiated between risk (by which he meant risk that is measurable) and uncertainty (which is not), arguing that “There is a fundamental distinction between the reward for taking a known risk and that for assuming a risk whose value itself is not known”.1

Knight’s argument, incidentally, was that it is only uncertainty (not risk of the measurable kind) which can lead to a profit. Discussion of that argument would represent a side road from which we may be a long time returning and I will avoid the temptation to take us down it here. What we should note, though, is that while it is common to refer to both

By: Bob Collie

Managing Director, Investment Strategy and Consulting

Russell Investments // Risk management: A fiduciary’s guidebook // Introduction

/ p 2

2 Adams, Douglas (1982). Life, The Universe and Everything. Pan Books. Adams goes on to explain: “The technology required to actually make something invisible is so complex and unreliable that it isn’t worth the bother. The ‘Somebody Else’s Problem field’ is much simpler and more effective, and ‘can be run for over a hundred years on a single torch battery.’ This is because it relies on people’s natural predisposition not to see anything they don’t want to, weren’t expecting, or can’t explain.”

(quantifiable) risk and (unquantifiable) uncertainty with the single term “risk,” a clearer distinction between the two would serve us well.

So one theme that runs through this guidebook is the variety and breadth of the risks (and uncertainties) that an investor faces. More perspective is better. It is dangerous to attempt to simplify risk, to assume that it can necessarily be measured, or to reduce it to a few dimensions.

Another theme is that good risk management begins with good decision making and sound governance. That’s because the most fundamental element in an institution’s risk profile is not the degree of exposure to the stock market or to interest rates, to particular counterparties, to style factors or to anything else you might care to mention. Rather it is how those exposures were chosen, how they are monitored and when and how they are changed. Without sound decision making, there can be no effective risk management.

Unfortunately, it is difficult to get people excited about good governance. It seems to be shielded by a force that pushes attention away from it, a real-life example of Douglas Adams’ SEP field, described in the Hitchhiker’s Guide to the Galaxy series.2 Not quite the invisibility cloak of later Harry Potter fame, the SEP field is almost as effective: “A SEP is something we can’t see, or don’t see, or our brain doesn’t let us see, because we think that it’s somebody else’s problem.” That is why it can seem easier to buy a new system than to address questions such as the suitability of the trustee board make-up or the effectiveness of delegation.

These and other thoughts on risk management are expanded in the articles that follow. Each is prefaced with a short introduction explaining the historical context in which it was written and the angle it takes. Each was written as a standalone article, but they add up to a composite picture that tells a richer story.

And as we look to the future, whatever it holds, we can be sure that fashions will come and go but the underlying importance of doing things right will not change. You can get away with second-best practices sometimes—and sometimes best practices can lead to bad outcomes—but in the end it’s the programs that are best run that will win.

Russell Investments // Risk management: A fiduciary’s guidebook // Introduction

/ p 3Russell Investments // Risk management: A fiduciary’s guidebook // There’s nothing normal about risk

There’s nothing normal about risk

By the middle of 2009, the global financial crisis was in full swing and had left investors—both individuals and institutions—reeling. Talk of black swans and fat tails was widespread and fast becoming the received wisdom. In this paper, which we have chosen to lead off our guidebook and set the stage for what will follow, Bruce Curwood captures the main arguments for why traditional approaches to risk management had consistently failed, not only in 2008 but on many occasions before then too. There are lessons to be drawn from that crisis and its aftermath, for those who choose to learn them.


There’s nothing normal about riskThis paper is the first of a three-part series on risk and its management. It will discuss important myths and facts about risk and articulate why a renewed approach to risk management may be necessary for most long-term institutional investment funds. We will look at risk largely from a defined benefit (DB) pension plan perspective, but the approach can also be adapted to other long-term funds, such as endowments. The second paper will follow up with an outline for this new direction in risk management for long-term institutional investors by addressing the issues identified here. The final paper, a case study, will demonstrate the application of risk management in a client-based setting, using many of the ideas presented in the first two papers. In short, these three papers will highlight the broad aspects of risk, suggest a viable approach to managing risk and give a practical demonstration of how to incorporate risk management practices effectively in the governance process.

BackgroundThe past 10 years have been a roller-coaster ride that most institutional investors would choose to remember only so as not to repeat. Not only were pension funds hit with the perfect storm of 2000–2002, when the Tech Wreck hit and interest rates fell, but they have been ravaged by the effects of the recent economic crisis as well. Both of these crises would be considered by most to be once-in-many-lifetimes occurrences, exceptionally rare (that is, in the extreme left tail of the normal distribution). Institutional funds have seen the mirage of their well-funded status (fully funded DB plans/endowments easily exceeding their policies for preservation and spending) evaporate twice in recent years, and it has been painful for trustees. As such, trustees are undoubtedly searching for a better understanding of the market events that have occurred and a new direction for managing risk within their funds, for a much gentler ride amidst any future market turmoil.

Why didn’t the conventional approaches of assessing and managing risk work this time? What did we miss in the process? How can we learn from these events? What should be the focus in the future? Our paper will address those questions and others of interest. It will also discuss some recent economic and investment research surrounding risk management. To start, we highlight some of the issues the current credit crisis has illuminated such as the importance of risk management to good governance.

By: Bruce Curwood

September 2009


What can we learn from the current crisis?As British Prime Minister Benjamin Disraeli once observed, “There is no education like adversity.”1 We believe the current crisis has taught us a great deal—about ourselves as long-term investors and about the market in general. First and foremost, we quickly learned that as investors we were too often unaware of some of the risks to which our funds were exposed. Yes, excessive leverage, poorly designed incentives and loose regulations got the economic crisis rolling, but who knew that counterparty risk might invade a money market fund through derivatives, or undermine the collateral in a securities lending program so extensively, or result in significant losses in a passive equity index fund? In stretching for yield during a period of low volatility and by levering our portfolios through aggressive strategies, had we merely been collecting pennies in advance of the steamroller? We did not expect the deterioration in the market to transpire so rapidly, and we did not understand that liquidity would be so important. We ignored the traditional adjuration to hold a few safe government bonds in our portfolios until the correlation of equity subcategories and alternatives converged during the recent crisis and made its wisdom evident. Even the ability to conduct basic rebalancing became an issue for many funds, particularly those that had followed the endowment model with a heavy allocation to alternatives. A lack of transparency had no doubt clouded our judgment, but it has become quite clear that our traditional risk models, which failed to focus on “tail events,” left much to be desired. Not only did many plan sponsors determine this was uncomfortable and that excessive short-term volatility was overly painful, but many vowed never to let it happen again! Sponsors didn’t enjoy that Investment Committees were questioning their investment beliefs and that some, although acknowledged long-term investors, converted risky assets to safe havens in the midst of the crisis. Discipline and stronger risk management practices would, no doubt, be a necessity in the future.

TO DO SO HOWEVER, WE MUST FIRST ACKNOWLEDGE SOME FUNDAMENTAL FACTS:

1. Risk is uncertain and multidimensional

2. Irrational human behavior may override what rational economic models suggest

3. There is therefore nothing normal about risk (tail events are just as important as the normative market conditions)

4. Thus, complex mathematical tools that assume normality should be enhanced with due consideration for fat-tail events

5. We cannot control return, but we can manage risk (or at least determine an acceptable level) and, perhaps, behaviors that might be deemed irrational

6. Investment committees should therefore focus greater effort on risk management and have a better understanding of possible adverse outcomes

7. Current governance spends too much time on administration, return generation and attribution, and needs to spend more time on risk management

Yes, there was a great deal to learn from this crisis, including the importance of focus and discipline in long-term investing. Unfortunately, many of us learned these lessons the hard way.

We expand on these themes in the rest of this paper. First, we look at the conundrum that long-term investors face and discuss some myths and facts about investment risk. Second, we explore how investor sentiment could affect investment decisions in light of recent

1 Turner, Bob (February 9, 2009), Letter to Clients, Turner Investment Partners.


findings in the field of behavioral finance. Third, we highlight shortcomings of many risk management tools currently in use. Finally, we offer some suggestions for what long-term investors should consider focusing on.

The long-term investor’s conundrum (can investors bear the short-term pain, to reach the long-term gain?)A defined benefit pension plan’s ultimate objective is to fund its liabilities over the long term, with assets exceeding the pension liabilities. (Similarly, the objective of an endowment or foundation is to have sufficient assets to consistently meet or exceed its spending and capital preservation policies). There are two main sources of assets: the contributions to the fund and the earnings (returns on the contributions). Due to the power of compound interest and the long-term nature of the liabilities (a perpetuity for endowments), the earnings generally exceed the contributions by a wide margin.

Although less-risky assets, like treasury bills and domestic real-return bonds, tend to match the inflation-sensitive liabilities better (with greater certainty),2 their total returns over time have typically been lower than those for riskier assets, like equities and alternatives. Long-term institutional investors therefore usually invest in riskier assets in an attempt to reduce contributions by increasing return, but in so doing, they also make a tradeoff to reduce long-term costs at the expense of short-term variability. As soon as they move away from a matched position (assets responding closely in value to the liabilities), they incur investment risk. To reduce overall fund volatility, most institutional investors therefore use a combination of risky and less-risky assets to diversify their asset mix, or they invest in assets that are less correlated with each other. The degree of risk assumed will depend on the objective, time frame and risk tolerance of the fund. The trouble here is that many funds in stretching for return to reduce cost, underestimate the pain of short-term variability in determining their long-term asset mix. That, in short, is the long-term investor’s conundrum!

Myths and factsThere are many half-truths and misconceptions to confront when it comes to the complex process of assessing and managing risk. Let’s start with the asset/liability study. We immediately recognize that correctly identifying the fund’s collective risk tolerance is an extremely difficult task. It’s akin to understanding the terror of the rollercoaster from the safety of the ground. Sure, you think you can tolerate the sensation and contain your fear during a dramatic vertical drop, but experience is the only true test of resolve! Hopefully, the short-term funding pain doesn’t force a bailout of the long-term investment strategy at exactly the wrong time.

Moreover, comprehension is obfuscated by the distilled quantitative models the actuary or investment consultant uses to help fiduciaries with the process. While these models are widely used as foundations for making asset allocation, portfolio optimization, diversification and risk management decisions, they are often treated as ”black box” models by fiduciaries who take their outputs at face value without properly understanding the models’ shortcomings and limitations. Perhaps the most significant shortcoming is the assumption that asset returns are “normal.” The number of market crashes and crises in the past hundred years or so has far exceeded the number that would have been expected if returns were to come from a normal distribution. Moreover, there appears to be some evidence that asset returns are correlated from one period to the next. Historical returns exhibit short-term cycles and long-term trends. These patterns of time-dependence are ruled out in conventional quantitative models.

2 There is rarely a perfect asset match for the liabilities, since the liabilities are not static and most assets exhibit some basis risk, especially during extreme events.


Even though asset consultants often highlight the probability of negative returns or funding shortfall in an asset/liability study, clients’ eyes are naturally drawn to the average and possibly to the upside of the probability distributions. But good risk management involves focusing equally, or even more, on the downside, as illustrated by Sam Savage in his book, “The Flaw of Averages: Why We Underestimate Risk in the Face of Uncertainty”. In fact, when evaluating the long-term benefits, we often don’t even consider probing the experience of nearer-term volatility—what it will be like to incur the poorer returns first. In addition, because the asset/liability study is generally done once every three years or so, and reviewed only annually, fiduciaries seldom devote enough time to a thorough comprehension of the assumptions, drivers and potentially extreme variability of outcomes. Even when they do, valuable lessons are too often forgotten, perhaps several years later in the heat of market movement. With busy schedules and investment committee meetings jam-packed with manager and other issues, fiduciaries often want to get right to the recommendations (“tell me the answer now”) without understanding that the journey is as important as the destination. The fact that there are multiple objectives (solvency or going-concern liabilities) and outcomes (contributions, assets, liabilities, fund surplus, etc.) also interferes with fiduciaries’ comprehension. A further confusion is investors’ penchant for looking at secondary goals, such as adding value to the fund policy benchmark through active management or above-average fund performance relative to a peer group.

Moreover, it is commonly assumed that risk tolerances are constant over time, as we tend to set the asset mix for the long term. But is that really true? Most investors, encouraged after several years of favorable market performance, are more prepared to take on additional risk (increase risk tolerance at the wrong time) by increasing their exposure to risky assets, while just the opposite (decreasing risk tolerance) generally occurs after a downward market shock, or several years of negative returns. We witnessed this low-return/ low-volatility environment in 2007, when many funds reached for additional return, unaware of the additional risk incurred. Many even swapped known bond returns for the allure of alternatives, whose risks were less well known. In addition, because of constant committee turnover, many fiduciaries at the helm during the current economic debacle, may not be the ones who actually set the risk policy, years earlier. Without proper documentation and good supporting research, the new committee may abandon the policy at an inopportune time or decide they have a lesser risk tolerance.

What can explain such investor behavior? Perhaps this is a good juncture at which to discuss some findings in behavioral finance which show that investors’ decisions are often based on more than just economics.

Fear, greed and the market cycleBeing sensible investors, we all like to think of the world as being a rational place: economic fundamentals drive the business cycle, and as professional investors generally study the leading economic indicators, the stock market tends to be anticipatory and will often lead the business cycle by six to 12 months. In practice, however, reading the economic entrails is more difficult than it sounds, and few professional investors can successfully “time” the stock market on anything like a consistent basis. Even when investors get the direction right, they often read the timing and magnitude wrong.

Why are accurate numerical predictions so difficult? First, economic predictions are holistic; they incorporate multiple and diverse variables. The economy is as dependent on trade and productivity as it is on weather patterns, politics and the occasional pestilence. For every variable that can reasonably be forecast with some degree of accuracy, there is a purely random variable that no computer can “predict” with any degree of confidence. At various times these factors reinforce or offset each other. Economic data is often inaccurate, and it is revised frequently. As different variables are considered and combined in intricate models, the uncertainties multiply. Thus, differences in predictions that are based on small changes


to existing models can be widespread. Nor do forecasts accurately anticipate turning points in the economy. There are long-term secular trends as well as the more closely followed shorter-term business cycles, whose influence periodically prevails in the stock market for varying lengths of time.

In his recent text “The Future of Everything”, David Orrell3 noted that Organisation for Economic Co-operation and Development (OECD) forecasts of G7 gross domestic product (GDP) growth (1986—1998) were notoriously incorrect, with a standard deviation of .95, or analogous to naïve predictions. How many economists got the call right last year (2008), or in 1998 or 1991, for that matter? The answer is, unfortunately, very few.

To complicate matters further, at different times, fear and greed can be the main drivers of the market. In Exhibit 1, ‘The Stock Market Pendulum,’ you can see that while markets are unpredictable, it’s investor behavior that may really hurt! At these market extremes, the correlation of risky assets tends to go to 1. Modeled predictions based on normal events can therefore be dramatically deficient under such circumstances. Are there findings in behavioral finance that institutional investors can draw upon to improve the risk management process?

Exhibit 1: The Stock Market Pendulum

“Markets are unpredictable but it’s investor behavior that may really hurt!”

Behavioral financeA Russell Research paper, “Resist the Amygdala!,”4 provides an introduction to the subject of behavioral finance and a brief description of 16 findings from that field which are noteworthy for institutional investors. The paper ends with some thoughts on the application of these findings to institutional investment programs.5 Many findings from this field are, in fact, directly applicable to the subject of risk management, and have been useful in establishing the major themes of this paper. We highlight some of the findings relevant to risk management below.

Stock Market Valuation Based onAnticipating Economic Fundamentals

“Markets are unpredictable but it’s investor behavior that may really hurt!”

Stock Market Valuation Based onAnticipating Economic Fundamentals

Investor Behavior Based on Emotion

Bear

OvervaluedUndervalued

Fear Greed

Bull

(negative) (positive)

3 Orell, David (2007). The Future of Everything. Thunder’s Mouth Press.4 Collie, Bob and D. Don Ezra (2007). “Resist the Amygdala!” Russell Research. 5 These thoughts are grouped into three areas: the operation of an investment committee, the asset allocation

decision and the selection and monitoring of money managers.


THE NATURE OF THE RISK MANAGEMENT TASKNot all tasks are the same; the best approach to tackling the challenges in one decision-making scenario may not be workable for the next. D. Don Ezra’s and Bob Collie’s “Amygdala” paper framed a distinction between decisions that are better made by experienced individuals using judgment and those that are better formulated through a rigorous analytical process. This distinction is captured in the following table.

Task conditions Recognition-primed decisions Rational choice strategy

Greater time pressure More likely

Higher experience level More likely

Dynamic conditions More likely

Ill-defined goals More likely

Need for justification More likely

Conflict resolution More likely

Optimization More likely

Greater computational complexity More likely

Source: Klein, Gary. Sources of Power. Massachusetts Institute of Technology Press, 1998.

A general observation that can be drawn from this table is that the majority of the very human decisions investors make fall into what Klein calls the “recognition-primed decision model.” In other words, most of the decisions institutional investors make happen quickly, and they are often based on ill-defined goals and in the absence of full knowledge of the parameters potentially affecting the outcome (these parameters are likely to be constantly changing, in any case.) As a result, the rational choice strategy of classical economic theory can be of little value in this context.

Which type of task, then, is risk management, and which approach to this task is most likely to be successful? Of the task conditions listed above, the two that most obviously define risk management are the dynamic conditions and the computational complexity: one from each column. This highlights the difficulty of this challenge. On the one hand, risk management does require a solid rational basis and rigorous processes. On the other, it is a skill that ultimately relies on far more than simply understanding and applying those processes. To be done well, risk management requires both types of decision processes to be in play—and that in itself brings the challenge of determining which process should dominate at which stage of the task. A hefty governance assignment indeed.

OVERCONFIDENCE AND OTHER BEHAVIORAL TRAITS THAT MAKE US UNDERESTIMATE RISKOf the many human traits studied in the behavioral finance field, one that is clearly applicable to risk management is our tendency to be overconfident. We tend to think we are more skilled at a task than we really are, and that we are more likely to succeed than we really are. Allied with this trait is the illusion of control—the feeling that we have more influence than we really do and that the world somehow revolves around us. It is a trait compounded by hindsight bias: we are more prone to recall to memory our successes than our failures, and to paint an inaccurately rosy picture of our past performance of any given task. Add to this a leaning toward confirmation bias, our tendency to seek evidence that confirms our views and to overlook or discount evidence that contradicts them. The presence of these traits means that as humans,


we are inclined to underestimate the extent of the risks we face, and to overestimate the likelihood that an outcome will be in our favor.6

The best prescription in the face of all this appears to be two-fold. First, replace subjective assessments with objective measurement wherever possible. And, second, continually ask this question: “Why might we be wrong?” This simple principle does indeed underlie many of the arguments we advance in this paper for improving management of risk by institutional investors. For example, when we explore the role of risk models in the next section, we will see how these models may become a hindrance to effective risk management if they serve to reduce the habit of asking the wider questions.

The problem with existing models: There is nothing normal about risk!We’ve all learned directly via university studies or indirectly through CFA and finance writings about pioneering applications of mathematics to financial markets. In fact, many well-known scholars have won Nobel Prizes for their theories. Markowitz, Sharpe, Scholes and Merton, through detailed, rigorous mathematics, helped build the foundations of modern finance that underpin current risk management practices. With cogent dissertations on various financial models, they enlightened and converted many investors (who previously relied on rules of thumb or hunches) to using mean-variance analysis and the tools of mathematical optimization. In many of their mathematical models, a simplifying assumption is made that investment returns are normally distributed and that volatility is a good measure of investment risk (value at risk, tracking error, standard deviation, etc). According to these theories, a fund manager can build an efficient portfolio by targeting a specific return with a desired level of risk. These inputs form the basis for asset/liability modeling (optimization).

A few brave scientists, like Mandelbrot, noted however that although the mathematics were elegant and aided in understanding, the concepts were built on shaky assumptions. Investors are not always rational (clear-thinking) and self-interested (trying to get rich); they are not alike (homogeneous); prices jump regularly and significantly (rather than moving continuously); price changes are not independent “random walks,” nor do they typically look at all like a normal bell curve (normal distribution). Eugene Fama, widely regarded as a founder of the academic discipline of modern finance, noted that big price changes in the stock market were far more common than what most financial models would suggest.

“Large changes, of more than five standard deviations from the average, happened two thousand times more often than expected [the concept of large fat tails]. Under Gaussian rules, you should have encountered such drama only once every seven thousand years; in fact, the data showed, it happened once every three or four years.”7 Financial securities prices (stocks, commodities, currencies, etc.) and therefore stock indexes, only rarely followed the predicted normal pattern. To paraphrase Mandelbrot, the traditional, quantitative market models were not just wrong, but dangerously wrong.

Exhibit 2 documents more than 25 risky events that have occurred in the last 20 years. These market disruptions have been far greater in number and far more significant than what the models would have predicted. Furthermore, they seem to have increased in magnitude and frequency in recent years. This may be attributable to developments in technology, increased globalization and trade, derivatives and the overall interconnectedness of global financial markets.

Exhibit 2: How risk has been manifested over the last 20 years

6 Overconfidence, the illusion of control, hindsight bias and confirmation bias are all described, with references to key academic studies, in “Resist the Amygdala!” (op. cit.).

7 Mandelbrot, Benoit, and Richard L. Hudson (2004). “The (Mis)Behavior of Markets: A Fractal View of Risk, Ruin and Reward.” Basic Books, NY; p. 96.


1987 October Black Monday Market crash (portfolio insurance?) DJIA –29.2% in one day (probability 1/[10]^50) or highly unlikely)*

1988 Hammersmith and Fulham Interest rate swaps/void contracts (–$0.6B)

1994 Orange County Derivatives reverse floaters (–$2.0B)

1994 Proctor and Gamble Exotic interest rate derivatives (–$0.1B)

1994 Kidder Peabody Rogue bond trader losses (–$0.1B)

1995 Barings Rogue futures trader Nick Leeson (–$1.0B)

1995 Daiwa Rogue bond trading losses Japan (–$1.0B)

1997 NatWest Bank Bad swap options model (–$0.1B)

1997 Asian Financial Crisis DJIA falls 7.7% in one day (probability one in 50 billion)*

1997 BreX Gold stock fraud ($–4.4B)

1998 Russian Debt Crisis

1998 LT Capital Leverage/wide credit spreads (–$4.0B) August market collapse three days <–3.5% (probability one in 20 million)*

2000 Technology Bubble

2001 Terrorism 9/11

2002 Corporate (Enron, WorldCom, Tyco, etc.) DJIA three steep falls in Accounting Fraud seven days (probability one in four trillion)*

2002 Allied Irish Bank Speculative FX trades (–$0.7B)

2003 Lancer Hedge fund fraud (–$0.4B)

2005 Bayou Hedge fund fraud (–$0.9B)

2006 Amaranth Trader gas futures losses (–$6.0B)

2007 Subprime Crisis ABS and quant meltdown—a 6 sigma event

2008 Soc Gen Rogue trader losses (–$7.0B)

2008 Madoff Ponzi scheme / fraud (–$50.0B?)

2009 Stanford Group Fraud (–$8.0B?)

Source: Russell unless otherwise noted.

*Source: Mandelbrot, Benoi, and Richard L. Hudson (2004). “The (Mis)Behavior of Markets”.


Exhibit 3 shows Mandelbrot’s demonstration of how poorly modern finance and the normal distribution assumption have represented actual market movements.8

Exhibit 3: Actual market movements compared to a standard model

Source: Mandelbrot, Benoit, and Richard L. Hudson (2004). “The (Mis)Behavior of Markets: A Fractal View of Risk, Ruin and Reward.” Basic Books, NY; p. 93.

An entire new wave of literature has delineated a host of problems concerning current risk management practices (Bookstaber, Damodaran, Rebonato, Taleb, etc). Objections focus largely on the non-normality of markets and the ways by which modern finance underestimates downside risk. In essence, a traditional mean-variance framework fails to incorporate:

1. Fat tails, leading to underestimation of downside risk

2. Serial correlation, leading to errors in estimating the true volatility of asset returns;

3. The convergence of correlations during periods of high market volatility, leading to overstatements of the benefits of diversification

These findings put in doubt the whole Gaussian/normal approach to managing risk, given that the normal distribution assumption is inconsistent with financial price variation. In Fooled by Randomness, Nassim Taleb demonstrated that many supposedly successful investors who traded on economic variables may have just been fortunate, or that what was deemed successful was merely coincidence. One can make money in the financial markets for limited periods, totally from randomness. Performance records are often exaggerated by the observer. Also, due to problems such as survivorship bias (only those who performed well

8 Ibid, p. 93.

20

15

10

5

0

10

5

0

19151920

19251930

19351940

19451950

19551960

19651970

19751980

19851990

19952000

Changes in the Dow, in standard deviations

Changes in the Browian Motion, in standard deviations


survive, and therefore bad performers are excluded from performance calculations), data mining or regression to the mean, there is often a misconception about the importance of randomness. Traders are trading on noise for most days in the market, yet it is a few surprise events that have usually changed the shape of financial history.

HERE IS TALEB ON NORMAL DISTRIBUTION:“The problem is that we read too much into shallow recent history, with statements like ‘this has never happened before,’ but not from history in general (things that never happened before in one area tend eventually to happen). In other words, history teaches us that things that never happened before [the black swan or large negative event] do happen… The practice of ‘financial engineering’ came along with massive doses of pseudoscience. Practitioners of these methods measure risks, using the tool of past history as an indication of the future… which “makes the entire concept seem like a costly (perhaps very costly) mistake.”9

Genius also seems to meet with its fair share of failure when modern finance theories are applied to the capital markets. Long-term capital is a prime example; two Nobel laureates destroyed their firm (and very nearly a significant part of the financial system) in 1998 by focusing on normative quantitative techniques. John Meriwether was quoted in the Wall Street Journal as saying that their “whole approach was fundamentally flawed” and their “whole focus is on the extremes now.”10 This is one of numerous examples, as with portfolio insurance in 1987, where modern finance has sometimes failed miserably, due to overconfidence in the accuracy of financial models.

So why do we continue to use such misconceived statistical models? Various explanations have been put forth, ranging from habit, convenience, lack of understanding and failure to challenge the conventional wisdom. The latter problem is particularly intractable, given that these tools were created by legends in the industry and are still taught in schools of finance around the world. As human beings, we want to believe in the physics of economics. Mathematics provides us a perceived structure and thus a measure of comfort in an uncertain financial world. Despite this, perhaps it is time to confront traditional models and question the integrity of their foundations, given that they are built upon the normal distribution assumption that has been proven wrong time and time again. We need to determine the level of acceptable risk through a broadened set of risk metrics and to ensure that fiduciaries take a similarly broadened but focused approach to risk management.

A better focusWe have looked at the vagaries of the market and outlined the importance of discipline to overcoming behavioral shortcomings in investing. We have also drawn attention to deficiencies in conventional quantitative models and underscored the importance of a broadened approach to assessing and managing risk. Clearly, investor discipline requires focus, and there are three key areas which are unique to every investor: time horizon, goal setting and risk tolerance. These three aspects must be clarified and resolved at the outset of any risk management process.

Breakdowns in decision making can frequently be traced to inadequate consideration of the impact of short-term, negative events on funding and liquidity; unclear or multiple goals (or a failure to tie decisions closely enough to those goals); or a failure to properly address risk tolerance. This can apply, for example, in the context of strategic asset allocation, which is one part of the broader risk management process for an institutional investor. The challenge

9 Taleb, Nassim N. (2004) “Fooled by Randomness.” Random House Inc., NY; p. 115. 10 Mandelbrot and Hudson, ibid.; p. 107.


is in how investors can better focus on these three key areas. We briefly summarize some suggestions, taken from an earlier-referenced Russell paper, “Resist the Amygdala!”, below:

1. TIME HORIZON Since the readers of this paper are long-term institutional investors, we needn’t discuss time horizon in great depth. We might simply reiterate that given the long-term investor’s conundrum, it is inherent that any sound risk management practice should probe an investor’s short-term pain threshold extensively. The essential question is which is the greater risk, short-term volatility or failing to meet a long-term objective, and how do we reconcile the two factors (less risk and more cost)?

2. GOAL SETTING The decision process begins with setting goals. We have shown earlier in this paper that goals are often inadequately considered. This observation seems well founded in the case of many asset allocation decisions. Indeed, it seems to us justifiable to spend the majority of the time devoted to an asset allocation study on this first stage.

At a superficial level, asset allocation goals are easy to set: maximize expected return, subject to acceptable risk. In reality, the quality of the final decision will rest largely on how well “risk” is defined, and how confidently the investor can decide what is “acceptable.” But clearly however, in most cases some degree of risk must be accepted to reach many long-term investors’ goals. So investors must guard against being overly conservative or frozen by unrealistic aspects of the downside.

The goals of an institutional investor are complex, and they can conflict with each other. There are many stakeholders to be considered. For instance, in the case of a corporate defined benefit pension plan, the plan sponsor has certain objectives which may differ from the objectives of the plan’s fiduciaries. Also, different cohorts of beneficiaries will have different goals: for older members of the plan, security of principal is the dominant goal (to ensure that benefits can continue to be paid), while for younger members, capital appreciation is necessary to ensure that the plan remains viable and that future benefits can continue to accrue. To confuse matters, there are several accounting and reporting conventions and requirements that may be another source of conflict.

In addition, investors of all types are influenced by their peers. Even though it has been many years since it was generally considered acceptable to set a goal as merely to

“outperform the peer group,” most committees find it hard to reach decisions that deviate too far from the practices of others. In part, this instinct is reasonable: those who pay no regard to peers may be guilty of overconfidence. One of the tests of prudence that courts apply in practice is a comparison of how others in similar circumstances behaved. The onus is on those acting differently to prove that their actions are justified. Those who follow the herd will rarely if ever be blamed for doing so, but will their primary objective (full funding) be attained?

Peer groups are rarely part of the formal goal-setting process, and we are not arguing that they should once again become a major factor in asset allocation decisions. However, where material divergences from peers’ opinions do arise, it is worth the extra effort of identifying why we have reached different conclusions, remembering that our aim is always to ask, “Why might we be wrong?”

These issues should not be regarded as complications or obstacles getting in the way of the “pure” asset allocation decision you’d like to base on a single measure of success. More time should be spent on understanding the multiple goals of asset allocation policy, the nature of risk and what constitutes success and failure (the primary objective or mission). That is at the heart of the asset allocation decision.


3. RISK TOLERANCEHaving considered what risk is, an investor must then decide how much of it to take. Defining risk tolerance has always been one of the least rigorous aspects of the strategic asset allocation process. Decades ago, in the early days of mean-variance analysis, risk tolerance was a simple number that was plugged into Markowitz’s formula. So if your risk tolerance was 100 (which was quite high), you were considered prepared to accept up to 100 units of additional variance11 for an additional unit of expected return. Depending on the exact asset class assumptions used, this might lead to an equity allocation of 90% or more. If your risk tolerance was 10 (quite low), you were deemed less willing to take on additional variance in return for a higher expected return: you would accept only 10 units of additional variance for the same additional unit of expected return. Your resulting equity allocation might be 30% or less. Most investors, strangely enough, turned out to have risk tolerances of something like 55 or 60. The majority of investors therefore ended up with equity allocations around 60%.

Of course, what was really happening a great deal of the time was that investors were choosing the risk tolerance that produced the asset allocations they wanted to adopt, based on their preferred return estimate. The question “How much additional variance are you prepared to take on for an additional unit of expected return?” is not particularly intuitive. Very few people had ever thought in those terms until the question was asked.

The best asset allocation processes today avoid making risk tolerance an abstract input. It is a concept underlying much of the modeling, below the surface. But an investor’s risk tolerance can be explored in more meaningful terms by looking at the impact on funding status, return volatility and so on, and of the various choices open to them. Indeed, it is clear today that risk has many dimensions, that these dimensions cannot simply be aggregated into one number and that the investor almost certainly tolerates some forms of risk more easily than others.

So whereas risk tolerance has long been expressed in terms that seem very palatable to the prefrontal cortex (“55”), today it is not. There are tools available today that allow investors to be more rigorous in this critical area. But we must remember they are still just models and not crystal balls. One needs to be highly skeptical of the variables, the assumptions, the outputs and the possible outcomes.

“Never again!”We know that the most frequent failings in decision making in general are inadequate definition of the right goals at the outset, and failure to tie the final decision back to those goals. In the case of asset allocation, these tendencies manifest themselves through inadequate definition and oversimplification of what risk really is. In an institutional investing world that is becoming ever more complex the right asset allocation policy begins with a clear understanding of what it is we are trying to achieve.

As we indicated at the outset, most long-term investment funds have fared poorly over the last decade. Some harsher critics have even proclaimed that past risk management policies and practices have failed entirely. Others have noted the extremely unpredictable circumstances of the market that we have recently experienced. In any case, fiduciaries have been unable to protect their funds from the effects of the recent stock market carnage, and a great number of fiduciaries are vowing that “this will never happen again.” They no doubt recognize the limitations of traditional models and realize that there are more dimensions to risk than meet the eye. But aren’t these the same hollow words that echoed

11 Even though the natural trade-off to use would be standard deviation of returns versus expected return (because each is measured in the same unit: percent per annum), the mathematics are more tractable if variance (the square of standard deviation) rather than standard deviation is used, and the end result of the optimization process is the same.


worldwide after the Tech Bubble burst, and weren’t these same words quickly forgotten as the global stock markets recovered? “Well, not this time!” How do we ensure that the debacle is not repeated?

Quite simply, investors must plan for long-term success but still be ever cognizant of the downside. They must realize that risk is by definition uncertain and unpredictable, and therefore that risk management is a dynamic and rigorous process, which must be considered more seriously than emotion or hunches would suggest. The current governance process, which often involves spending the majority of time on return analysis and administration, must become more focused on risk management. No single statistical asset/liability model can possibly capture the multiplicity or multidimensionality of investment risk. Investors must spend more time, energy and resources on its management, rather than just relying on a triannual asset/liability study. Also, since risk is not necessarily based on normal distributions, investors must use a variety of tools such as stress testing, scenario analysis and downside analysis to assist in the process. We must be vigilant against letting our human frailties dominate and resist becoming over-aggressive in our investment strategies (markedly more risk-tolerant) as the market strengthens. In fact, fiduciaries may want to preserve the wealth gained after several years of good performance, when their funds are closer to meeting their primary objective. But this type of singular focus takes discipline and a well-coordinated game plan.

Above all, investment committees must know their collective risk tolerance and solve the long-term investors’ conundrum. They must balance contribution cost against the risk of not meeting their overriding objective of the assets exceeding the liabilities. Although investing for the long term, they must also take into account the effects of short-term market volatility on their stakeholders. In short, they must probe to find their short-term pain threshold, while keeping their eyes on the potential long-term prize.

In our next paper, we will look at how to build governance structures that are better equipped to manage risk—and at the importance of knowing that the time to be most vigilant is often when there seems to be little to worry about. Discipline and an effective governance approach are necessities, lest emotion overrun regard for the investment objectives. Unfortunately, this is not as simple as it sounds and often requires a cultural change throughout the organization!

References Bernstein, Peter L. “A Fresh Look at the Super Part of the Debt Supercycle.” Economics and Portfolio Strategy, March 15, 2009.

Bookstaber, Richard. “A Demon of Our Own Design: Markets, Hedge Funds, and the Perils of Financial Innovation.” John Wiley & Sons, Inc., 2007.

Collie, Bob and Ezra, Don. “Resist the Amygdala!” Russell Research, March 2007.

Damodaran, Aswath. Strategic Risk Taking: A Framework for Risk Management. Wharton School Publishing, 2007.

Kazemi, Kambiz. “2008: Volatility in Action.” Toronto CFA Society Analyst, March 2009).

Mandelbrot, Benoit, and Richard L. Hudson. The (Mis)Behavior of Markets: A Fractal View of Risk, Ruin & Reward. Basic Books, 2004.

Mauboussin, Michael J. “Long-Term Investing: How I Learned to Stop Worrying and Ignore Volatility (excerpts from comments from the Greenwich Roundtable, December 11, 2008).” Legg Mason, December 15, 2008.

Orrell, David. The Future of Everything: The Science of Prediction. Thunder’s Mouth Press, 2007.


Otar, Jim. “Monte Carlo Simulators: Are They Worth the Gamble?” July/August 2007. Available at http://www.retirementoptimizer.com/articles/mcimca0001.PDF (accessed 8/19/2009).

Pilkey Orrin H. & Pilkey-Jarvis, Linda. Useless Arithmetic: Why Environmental Scientists Can’t Predict the Future. Columbia University Press, 2007.

Rebonato, Riccardo. Plight of the Fortune Tellers: Why We Need to Manage Financial Risk Differently. Princeton University Press, 2007.

Savage, Sam L. The Flaw of Averages: Why We Underestimate Risk in the Face of Uncertainty. John Wiley & Sons, 2009.

Taleb, Nassim N. Fooled by Randomness: The Hidden Role of Chance in Life and in the Markets. Random House, 2004.

/ p 19

Risk management is the cornerstone of investing

The title of this recent piece captures its message perfectly. The argument we are putting forward in 2010 is that risk management is not just a component in an investment strategy, or a box to be checked. It is the very foundation of what a well-run investment program should be doing. Investment is risk management.

This view allows an investment program to be governed differently. The decision-making framework can be constructed from the perspective of risk management. By doing so, better structures can be built and better decisions made.

Russell Investments // Risk management: A fiduciary’s guidebook // Risk management is the cornerstone of investing

/ p 20

Risk management is the cornerstone of investingThis paper is the second of a three-part series on risk and its management. The first paper, “There’s nothing normal about risk”, discussed several myths and facts about risk, as well as problems with existing risk models. One of its main purposes was to communicate the confusion and anxiety investors felt when the 2007–2009 financial crisis revealed unintended exposures in their funds. This second paper explores the major reasons many institutional investors have trouble managing risk and proposes a new way to think about risk management and governance. (Please note that while we look at risk primarily from the perspective of defined benefit (DB) pension plans, much of what we discuss applies to other long-term funds, such as endowments.) Our final paper will be a case study demonstrating practical implementation of the concepts outlined in the first two papers. In total, this body of research will highlight the broad elements of risk, propose a viable approach to managing risk and provide an example of how to effectively incorporate risk management practices into the governance process.

Overview The root cause of failures in pension risk management is poor governance. Quite simply, most fiduciaries place too much emphasis on return and don’t spend the time it takes to understand and manage the risks in their funds. Everyone seems intent on rushing through the process of determining a plan’s asset allocation, so they can focus on manager selection and monitoring performance. Risk management is put on the back burner, and behavioural issues that affect decision making are overlooked. When returns fall short of expectations, the “blame game” begins and changes are made. The more volatility there is in the market, the more turnover in the fund, primarily because the investment committee doesn’t have the conviction of logic to stay the course.

Now is the time for fiduciaries to adopt better risk management processes and to consider the impact of human behaviour on investment decision making. However, this requires an entirely different approach to risk management and governance. We’re not talking about structural tweaks, but rather the hard work of building a better risk management framework and culture. Sometimes it’s only when you stare across the abyss that you truly recognise the enormity of the chasm between knowledge and action, and are then prepared to make a radical transformation in approach. The transformation begins with a greater understanding of risk and the willingness to change for the better.

By: Bruce Curwood

February 2010

Russell Investments // Risk management: A fiduciary’s guidebook // Risk management is the cornerstone of investing

Now is the time for fiduciaries to adopt better risk management processes and to consider the impact of human behaviour on investment decision making. However, this requires an entirely different approach to risk management and governance (not just structural tweaks).

/ p 21Russell Investments // Risk management: A fiduciary’s guidebook // Risk management is the cornerstone of investing

A pervasive problem Recently published research conveys a sense of the enormity and urgency of the problem we all face when it comes to managing risk. The following are excerpts taken from the main conclusions of three lengthy research reports written in 2008 and 2009. Interestingly, they represent concurring opinions from two different perspectives—from the macro view, on the financial economy as a whole; and from the micro view, on just the pension industry. Both vantage points make it clear that risk has been broadly mismanaged across corporations and pension funds alike.

THE MACRO VIEW First, let’s look at the macro view as perceived by the Organisation for Economic Cooperation and Development (OECD). In a 2009 report on lessons from the financial crisis, the OECD said:

“This article concludes that the financial crisis can be to an important extent attributed to failures and weaknesses in corporate governance arrangements. When they were put to a test, corporate governance routines did not serve their purpose to safeguard against excessive risk taking in a number of financial services companies. A number of weaknesses have been apparent. The risk management systems have failed in many cases due to corporate governance procedures rather than the inadequacy of computer models alone: information about exposures in a number of cases did not reach the board and even senior levels of management, while risk management was often activity rather than enterprise-based. These are board responsibilities. In other cases, boards had approved strategy but then did not establish suitable metrics to monitor its implementation. Company disclosures about foreseeable risk have also left a lot to be desired, even though this is a key element”.1

(See Appendix A for additional findings from this study.)

THE MICRO VIEW From a more micro view, MetLife published a research report on the pension industry at about the same time that the OECD released its study. The MetLife report concluded:

“The climate for change tends to be greatest when decision-makers acknowledge the need to improve current practices…. This survey suggests a wide gap between the importance plan sponsors ascribe to each risk area and the sponsors’ own reported success at managing those risks. Overall, more than two-thirds of all plans studied indicate some degree of inconsistency in how they view and manage pension plan risk, and about 15% show significant problems in this area.

For example, many respondents affirm their close attention to ‘Plan Governance,’ defined as the ‘exercise of effective, independent oversight, supported by internal controls within all areas and at all levels of plan management.’ However, the balance of the study’s findings suggest otherwise.

This research indicates the possibility that DB plan decision-makers are failing to effectively align importance and success”.2

(See Appendix B for a synopsis of the study’s major points.)

Recently published research conveys a sense of the enormity and urgency of the problem we all face when it comes to managing risk.

1 “The Corporate Governance Lessons from the Financial Crisis”. OECD Financial Market Trends, 2009; p. 2.Available at http://www.oecd.org (accessed Dec. 15, 2009).

2 “MetLife U.S. Pension Risk Behavior Index: A Study of Risk Management Attitudes and Aptitude Among Defined Benefit Pension Plan Sponsors”. MetLife, 2009. Available at http://www.metlife.com (accessed Dec. 15, 2009).


Finally, Morgan Stanley conducted a survey of public fund chief investment officers (CIOs) in July and August 2008, just prior to the market debacle. It summarised what practitioners had to say as follows:

“The results of the Morgan Stanley Public Fund CIO Survey show that the demands associated with this increased [market] volatility and complexity have put a considerable strain on U.S. public funds’ resources by introducing new issues (shortages of specialised staff, increasingly sophisticated risk-management needs) and exacerbating old ones (limitations imposed by politics, investment policies and the time needed to educate boards)”.3

What went wrong? How could all of these bright senior executives have gotten it so wrong and failed to protect their organisations? What behavioural impediments clouded their judgment? What masked their detection of the true nature of the problem?

From the macro perspective, a March 2009 article in Harvard Business Review outlined six ways that companies mismanage risk:

• Relying on historical data—history is an imperfect guide.

• Focusing on narrow measures—commonly used risk metrics don’t capture catastrophes and/or ignore risks that count.

• Overlooking knowable risks—these risks include market concentration and illiquidity.

• Overlooking concealed risks—an organisation’s risks are not transparent, or those responsible for assuming risk don’t report it.

• Failing to communicate—there are delays or distortions in the reporting of risk exposures, or risk reports are overly complex.

• Not managing in real time—the risks associated with embedded derivatives can be dizzying in times of market turmoil.4

These shortcomings are apparent in the pension industry. They are also strikingly similar to what Russell has identified as the main reasons many plan sponsors fail in their fiduciary duties:

• Defining the problem incorrectly by focusing too much on expected return while largely ignoring risk.

• Narrowly interpreting the solution set, thereby seeking a risk management tool instead of a risk management process.

• Underestimating the dynamic complexity of risk.

• Being impatient and looking for shortcuts, two related human behaviours.

• Allocating time inadequately by overemphasising what’s easy, such as administrative tasks and dealing with the historical.

The issues facing plan sponsors are further complicated by frequent trustee turnover and poor documentation, as lessons go unlearned and institutional memories become short. In addition, a lack of investment expertise increases the odds of group behavioural biases.

3 Morgan Stanley Public Fund CIO Survey. Morgan Stanley Investment Management Publications, 2008. 4 Stulz, Rene M. “Six ways to mismanage risk”. Harvard Business Review, March 2009.

These and other deficiencies in pension risk management have been well documented in research conducted over the past 20 years. The element common to all of them is poor governance…


These and other deficiencies in pension risk management have been well documented in research conducted over the past 20 years (see Exhibit 1). The element common to all of them is poor governance—poor governance resulting from insufficient attention to behavioural issues and inadequate risk management processes. The remainder of this paper will delve into these concerns in more detail, to better diagnose the problem and formulate a more prescriptive response.

Exhibit 1: The pension governance deficit is still with us

O’Barr and Conley (in their text Fortune & Folly), 1992“the aim of pension fund governance appeared to be focused more on responsibility deflection and blame management, rather than good governance and creating value for fund stakeholders”

Ambachtsheer, Capelle & Scheibelhut, 1997“They found a positive correlation between pension fund governance quality and organizational performance”

Ambachtsheer, Boice, Ezra & McLaughlin, 2004“excellence shortfall in their organizations… resulted in a median response of 66 basis points by the fifty senior pension executives who participated in the survey”

Clark, Caerlewy-Smith & Marshall, 2005“pension fund trustee decision making was less than ideal”

Clapman et al, 2007“cited a litany of governance related problems in U.S. pension funds”

Source: Ambachtsheer, Capelle & Lum “The Pension Governance Deficit: Still With Us.” Rotman International Journal of Pension Management, Fall 2008.

Insufficient attention to behavioural issues Richard Thaler, professor of behavioural science and economics at the University of Chicago, astutely remarked that “One of the most important insights from behavioural research is that we need to distinguish between ‘normative’ theories that tell us how rational agents ‘should’ behave and ‘descriptive’ theories that tell us what real people do”.5 In short, theory and practice often differ. All too frequently, however, we forget that investors are not always rational, and that human behaviour can impede investment decision making.

INDIVIDUAL BEHAVIOUR One of the most fundamental characteristics of human nature is to think we are better than we really are. We make level-headed estimates of other people’s odds of success, but we typically overestimate our own chances of success, in a tendency called “optimistic bias”. For example, our positive view of investment vehicles like hedge funds in 2002 may have been coloured by their encouraging performance during the Tech Wreck relative to the rest of the stock market. Optimistic bias, combined with the so-called “recency effect”, caused us to invest more heavily in hedge funds than we should have, given their lack of transparency and our limited understanding of the risks. We also unconsciously mimic others or put too much trust in whatever is familiar to us, all of which leads to “herding” and

“home bias”. We overestimate the power we wield over our own circumstances, an “illusion of control” that makes us complacent and results in too little planning. With “hindsight bias”, we convince ourselves that we foresaw what was going to happen, when actually we had no idea what the future would hold. As a result, we tend to fool ourselves into believing that

5 Mitchell, Roger. “Risk: The Final Frontier”. CFA Magazine, March–April 2006; p. 25. Available at http://www.cfapubs.org.

All too frequently, however, we forget that investors are not always rational, and that human behaviour can impede investment decision making.


we or others can make accurate predictions in a chaotic market. Above all else, we have a terrible time admitting that we don’t know something, as it lowers our self-esteem. (I can almost sense readers thinking, “Now that may be true elsewhere, but it certainly doesn’t occur in our investment committee meetings”. I believe that’s called “denial”!)

GROUP DYNAMICS In discharging their duties, investment committees are prone to the same host of emotions and decision-making biases as individuals. Investment firm Vanguard documented this extremely well in a paper entitled “Group Decision-Making: Implications for Investment Committees”.6 While acknowledging that groups have larger collective memories and greater information available, Vanguard also noted that group biases and behavioural hurdles can quickly derail these advantages. The four most notable hurdles put forth were groupthink, group overconfidence, committee composition and group polarisation. Groupthink occurs when members of a cohesive group are more interested in avoiding conflict than in realistically appraising various courses of action. Confirmation bias and shared-information bias, byproducts of overconfidence, ensure the group does not leverage its full investigative resources, which can lead to ineffective decision making. Groups that are too large degrade performance through poor coordination or motivational issues. Group polarisation demonstrates how a committee may often make riskier decisions than individuals within the group would have made on their own.

THE REALITIES OF OUR INVESTING BRAINS “Your Money and Your Brain”, by Jason Zweig, sheds additional light on why smart people make foolish and imprudent financial decisions.7 I am going to borrow heavily from this book in the remainder of this section, although I highly recommend that readers study the full text themselves.

Apart from discussing the normal gamut of emotions, including fear, greed, surprise and regret, Zweig clearly points out that the investing brain is far from the consistent, efficient and logical device we like to pretend it is. Even though the brain functions superbly for most purposes in daily life, it can lead us astray when confronted with the challenge of choices within financial markets. Humans have a phenomenal ability to detect and interpret simple patterns, a skill that helped our ancestors survive the hazardous primeval world and still aids us today in meeting the stresses of daily life. But when it comes to investing, our incorrigible search for patterns leads us to assume that order exists when often it doesn’t. We humans believe we’re smart enough to forecast the future, even when we have been explicitly told that it is unpredictable. Our investing brains:

• Search for patterns when confronted with random data.

• Leap to conclusions, with two in a row of almost anything making us expect a third.

• Seek shortcuts through instinctive “solutions”, even when we think we are engaged in sophisticated analysis.

This type of processing is unconscious, automatic and largely uncontrollable; you can’t turn it off or make it go away. For example, many investment funds determined their asset mix on the basis of desired return and a belief in low correlation to diversify, without properly testing their assumptions—an unfortunate shortcut. As Charles Jacobs has noted, “At best, logic is just a way to justify conclusions we have already reached unconsciously”.8

6 Mottola, Gary R. and Stephen P. Utkus. “Group Decision-Making: Implications for Investment Committees”. Vanguard Investment Counseling & Research, 2009. Available at https://institutional.vanguard.com (accessed Dec. 15, 2009).

7 Zweig, Jason. “Your Money and Your Brain: How the New Science of Neuroeconomics Can Help Make You Rich”. Simon & Schuster, 2007.

8 Jacobs, Charles. “Management Rewired: Why Feedback Doesn’t Work and Other Surprising Lessons from the Latest Brain Science”. Portfolio (The Penguin Group), 2009; p. 2.

Broadly speaking, institutional investors have underestimated the significance of behavioural issues to their detriment.


Broadly speaking, institutional investors have underestimated the significance of behavioural issues to their detriment. Riccardo Rebonato captured this point beautifully in “Plight of the Fortune Tellers” when he wrote:

“There is a well-established science that teaches us how to extract the best information from the data. It is called statistics, and its use in risk management is well-known. There is, however, another science that deals with how actual human beings reach financial decisions once the data have been gathered. It is a branch of experimental psychology, and it is also well-established. Unfortunately, its use in risk management is nowhere near as widespread as that of statistics”.9

Inadequate risk management processes The occurrence of multiple risky events of significant magnitude and impact in the last decade—with models predicting that each would happen only once in a lifetime—has exposed weaknesses in pension risk management. To be able to address these deficiencies, plan sponsors must first comprehend the nature of the problem.

MISDIAGNOSIS OF THE PROBLEM Many funds have simply failed to realise that risk is the cornerstone of investing, and instead have focused too heavily on expected return. Group overconfidence and the illusion of control have blinded fiduciaries to the fact that returns are largely unpredictable and outside their control. It is far more difficult to estimate trends than to determine an appropriate level of risk. Fiduciaries have forgotten that returns trend over time, vary with time and are subject to statistical noise. As such, they fell prey to the recency effect and were lulled into a false sense of security, thinking the exceptional market returns of the 1980s and 1990s would continue indefinitely. Good absolute returns masked the problem. But those days of irrational exuberance and heady market performance were driven by an unprecedented period of falling interest rates, which crested in 1981 and continuously abated thereafter.

Today we must realise that there is indeed a “new normal” which is truly uncertain. Part of the uncertainty is because we’re in a low-return environment and interest rates, which are close to zero, are likely to rise. Furthermore, the potential outcomes for the financial markets are quite dichotomous and may well be politically motivated, making them even harder to fathom. Take the U.S. debt load, for example. Policymakers can either tax the populace to pay down the debt, or inflate their way out of it (retiring the debt with dollars that are worth less). Each option has political implications, as well as vastly different financial consequences for investors. The markets came close to falling off a cliff in 2008 but did not. Now we are in a situation where the whole cliff could still break off—or everything could return to a more pension-friendly environment.

NO SILVER BULLET When fiduciaries do look at risk, they tend to define risk management narrowly, as a tool or compliance vehicle rather than a process. Contributing to this problem is the fact that most of the people talking about risk management are trying to sell investors black-box models. Their wares are appealing to funds looking for a quick-and-dirty solution that lets them feel like they have effectively addressed the issue.

But buying the black box is not the answer. Perhaps no one makes this point better than Benoit Mandelbrot, a Yale mathematician who uses statistics to poke legitimate holes in modern financial theory. For example, in his “Ten Heresies of Finance”10 (see Appendix C),

9 Rebonato, Riccardo. “Plight of the Fortune Tellers”. Princeton University Press, 2007; p. 235.10 Mandelbrot, Benoit, and Richard L Hudson. “The (Mis)Behavior of Markets: A Fractal View of Risk, Ruin and Reward”.

Basic Books, 2004; pp. 226–52.

When fiduciaries do look at risk, they tend to define risk management narrowly, as a tool or compliance vehicle rather than a process.


Mandelbrot challenges the notion that market prices are normally distributed and move continuously, assumptions that form the foundation of many conventional risk models. He asserts that markets are much riskier than most people realise, and that current models are used not because they work well, but because the simplicity of the mathematical assumptions behind them is not disputed.

THE COMPLEXITY OF RISK There are numerous facets to risk, and fiduciaries have generally failed to appreciate its intricacies. Russell believes that one of the best ways to visualise risk is to think of it as a three-dimensional matrix, something like a Rubik’s Cube (see Exhibit 2). Envisioning risk in this way demonstrates some very important concepts, such as its complexity, multidimensionality and interdependency. Then consider that market risk is not stable over time and that investors’ risk tolerances are also subject to dramatic change, and you start to understand the true extent of the predicament. In other words, the dimensions of each of the 125 equity risk boxes in Exhibit 2 are not constant and each or several can explode in size at any time. So trying to solve the risk riddle with a simple, two-dimensional (mean/variance) optimisation tool designed for normal markets is suboptimal and prone to failure. Making incremental changes to outmoded models will not suffice.

Exhibit 2: Managing asset class risk is a multi-dimensional process

Therefore, using triangulation that incorporates various diagnostics and making trade-offs between multiple estimates of risk may prove more fruitful than attempting to achieve a precise estimate of risk-adjusted return. As we noted earlier, determining the level of risk is somewhat manageable, but accurately forecasting return is not. Given the complexity of risk, perhaps it is better to be almost right about achieving a somewhat manageable objective, such as a given level of risk, than to be precisely wrong about a totally uncontrollable variable, such as a forecast of risk-adjusted return. This, however, takes a complete reorientation of the fiduciary thought process.

Source: Russell Investments

Bonds 64-factor matrix

Equities 125-factor matrix


Rethinking risk management Despite the abundance of research on risk management, findings have largely been ignored by pension plans. As plan sponsors face the uncertainties related to a new normal for the financial markets, it appears obvious that now is the time to focus on a better, more robust approach to managing risk.

PUTTING RISK IN THE PROPER CONTEXT In our first paper in this series, we stressed the importance of clearly defining time horizon, goal setting and risk tolerance. These areas, unique to each investor, must be fully explored and clarified. Don’t underestimate the time it takes to do this right, as the evaluation of each of these variables may affect your future response to risk.

Time horizon is easiest to define, given the long-term focus of institutional investing. Nonetheless, it may not be apparent that any risk management solution must incorporate a long-term strategic asset allocation time frame of 10 to 20 years, an intermediate time frame of five to 10 years and a short time frame of less than five years, including crisis periods. The latter is particularly important to ensure that a fund can weather a potential market calamity to attain the expected long-run benefits. This requires having alternate ways to evaluate risk, which generally encompass risk management techniques such as stress testing and scenario analysis. Rebonato suggests always framing the issue by asking three simple but pertinent questions:

• “Can I survive the worst plausible outcome?” (Are the downside results considered unacceptable by the system of assessing performance?)

• What is my breakeven?” (How badly do things have to go to yield a return no better than the riskless one?)

• How much do I gain in the best plausible scenario?” (After understanding the risks, what could be gained?)”11

Goal setting is more of a challenge. Since most investors have competing investment objectives, goals must be prioritised, with the primary objective kept solidly in mind at all stages of decision making. That said, we might have to back up even further—to plan design—and ask ourselves, “What is the mission of the fund in regard to the overall organisation and our human resources policy?” And “Can we afford it?” Take a manufacturing company with easily replaceable, unskilled labor requirements and high employee turnover. The company may decide that a pension plan is not that important in hiring and retaining workers and therefore might favor a DC plan. At the other extreme is a management consulting firm whose services are defined by the sophistication and professionalism of its employees. To attract and retain the best and brightest associates and to differentiate itself within the industry, the consulting firm may decide that a rich DB plan is a competitive advantage. In other words, the better an organisation can envision the big picture, the likelier it is to arrive at the appropriate design.

Risk tolerance is even more difficult to discern. Fiduciaries need to address its three components separately, but must also recognise the linkages between them. The three interwoven components of risk tolerance are:

• Risk psychology—a psychological trait, like intelligence, personality or aptitude;

• Risk capacity—a financial consideration establishing how much risk the organisation can afford to take in its pension fund; and

11 Rebonato, ibid, pp. 245–49.

Since most investors have competing investment objectives, goals must be prioritised, with the primary objective kept solidly in mind at all stages of decision making.


• Perceived risk—a subjective judgment about risk, such as the feeling that investing in stocks in a bull market is not as risky as investing in stocks in a bear market.

During the process of selecting or designing a portfolio, the client and consultant may discover gaps between required risk (the risk inherent in the return required to achieve the goals); risk capacity (the maximum affordable risk); and risk psychology (the preferred risk/return tradeoff). Exploring these relationships is usually a powerful educational experience about risk and return for the client. This diagnostic process should be led by the consultant, whose role is to suggest alternatives, illustrate outcomes and recommend—but not to decide. This process is commonly referred to as “gap analysis” and is usually handled with clients through a combination of:

• Increasing resources (returns) by earning more and/or spending less;

• Contributing more via employer and/or employee contributions;

• Delaying, reducing, discarding and prioritising goals; and/or

• Assisting in the determination of the appropriate level and preference for risk, so that it is enough to achieve the long-term goals, but not so much that the client might panic and sell in a downturn.12

Risk capacity should also lead to a discussion about the financial health of the plan sponsor, the nature of the organisation (corporation or government entity) and the potential impact of the pension fund liabilities/contributions on the organisation’s financials. Over the last decade, poor investment returns (decreasing assets) and declining interest rates (increasing liabilities) have devastated the funded status of many sponsors and forced them to try to find ways to make contributions at a time when they can ill afford to do so. Again, discussions between the consultant and client on this issue are critical.

Let’s review two polar extremes. On one hand, a countercyclical consumer staples firm, such as a brewery, may actually find that its corporate profits increase during economic downturns. At times like these, the consumer feels less wealthy and opts to drink beer instead of more expensive wine and spirits. Consequently, the firm can well afford increased pension contributions when the economy is weak. Compare that to a discretionary consumer products company, such as a restaurant chain, whose profits are highly sensitive to the economy. All else being equal, the former should be able to afford a more risky asset allocation in the long term, as it will probably endure far less stress in a short-term crisis.

But the assumption of all else being equal rarely holds in real life, so the magnitude of the pension liability on the sponsoring organisation must be considered in great depth, too. Are pension liabilities and potential contributions minimal in relation to the corporate financials, as might be the case at a growing tech company with a small, young workforce? Or are they extremely large, such as we would expect to see at a mature car manufacturer? Each corporation’s capacity to absorb the pension downside may differ considerably. Given the well-publicised challenges of managing a pension fund, corporations may also wish to understand their exposures compared to competitors and broader peer groups. After all, in any competitive environment, it behooves us to understand how we might gain an edge on our rivals, and the pension area is no exception. As the old joke goes, the hunter may not have to outrun the bear if he can run faster than his hunting buddies.

WEIGHING THE ALTERNATIVES Once investment risk is put into context relative to the overall firm, and the nuances of risk have been thoroughly considered, the merits of various alternatives can then be

12 Black, Pamela, J. “The Three Faces of Risk; What You Need to Know about Risk Tolerance”. The Global Association of Risk Professionals (www.GARP.com) newsfeed, 2009.

Risk capacity should also lead to a discussion about the financial health of the plan sponsor, the nature of the organisation and the potential impact of the pension fund liabilities/ contributions on the organisation’s financials.


appraised. Two broad options immediately spring to mind. The sponsor can either “de-risk”13 the pension fund through pension buyouts, plan redesign or risk mitigation, including immunisation (although most find it too costly to de-risk). Or the sponsor can seek to better comprehend and manage the risks.

In our examination of the recent research, we have seen two principal schools of thought on how risk should be managed for those funds that decide not to de-risk. They fundamentally differ based on their market interpretation and beliefs.

1. Writers like Nassim Taleb, author of “The Black Swan”, believe that the market is unpredictable and that any attempts to outsmart it are largely futile. In short, they argue that you shouldn’t bet what you can’t afford to lose, because the market may indeed be irrational, but it is always “right”. Allow me to coin the argument, “be somewhat conservative”, since risky investors are often just picking up pennies in front of the proverbial steamroller. The idea is to take no more risk than you are comfortable with and is deemed necessary to meet your fund’s objective. Subscribers to this philosophy should explore more deterministic investment strategies, including liability-driven investing (LDI),14 liability-responsive asset allocation (LRAA)15 and buying cost-effective insurance, such as options, to hedge some of the risk.

2. Others, many of them investment managers, also believe the market is uncertain. That said, they think that through better research and understanding, informed professionals can determine when the market is cheap and rich. By weighing the probabilities and placing small, calculated bets (buying some risky assets) to increase market exposure when it is cheap, and reducing bets when the market is rich, value can be added over a deterministic approach. In summary, informed judgment, being more nimble with better risk intelligence and avoiding critical mistakes should lead to superior outcomes. More dynamic investment strategies, such as enhanced asset allocation (EAA)16 and strategic tilting,17 might be appropriate here.

Both of these schools of thought—conservative vs. informed judgment—and their corresponding investment strategies—deterministic vs. dynamic—agree on one overriding fact, which is that markets are uncertain. However, none of the strategies profess to eliminate risk; rather, they hope to assess and possibly contain it. Of course, effective risk management requires much more than simply figuring out which camp you are in and then picking an investment strategy. Fiduciaries need a framework for proactively managing the risks in their funds on an ongoing basis.

13 Please remember that all investments carry some level of risk. Although steps can be taken to help reduce risk it cannot be completely removed.

14 LDI, or liability-driven investment, is a set of strategies designed to reduce risk by managing the volatility in the gap between liabilities and assets by better matching expected cashflows. See Scott M. Donald, “An Introduction to LDI”, Russell Research, April 2007.

15 LRAA, or liability-responsive asset allocation, adjusts asset allocation based on changes in a defined benefit plan’s funded status, to reduce risk as the goal of full funding narrows or is achieved. See Sorca Kelly-Scolte’s, “Liability-responsive asset allocation”, Russell Research November 2009.

16 EAA, or enhanced asset allocation, provides an additional source of potential active returns by aiming to identify which markets are poised to out- or under-perform one another, and take advantage of these views. This information can be used to temporarily adjust or “tilt” a portfolio’s exposure from its long-term default strategic asset allocation in response to changes in expected returns. See “Enhanced asset allocation: a disciplined and strategic response to market dislocations”, November 2009.

17 Strategic tilting aims to deviate from the long-term strategic asset allocation to improve portfolio performance, only when markets are at some unsustainable extreme. See Andrew Pease and Geoff Warren, “Strategic Tilting”, Russell Research, October 2008.


ESTABLISHING THE RIGHT FRAMEWORK Common sense dictates that fiduciaries pursue a fund-wide approach to risk management, an all-encompassing process that incorporates both quantitative and qualitative risks. This was the topic of my paper entitled “A Comprehensive Risk Management Framework for Investment Funds”, which appeared in the summer 2007 issue of The Journal of Investment Consulting. In it, I concluded that a fund-wide investment risk framework, which addresses the following nine desirable attributes and six risk management elements, is a necessity for most plan sponsors.

Nine desirable attributes of an ideal risk management framework:

Risk considerations

1. Has a fund-wide scope, encompassing all investment (quantifiable) and process (qualitative) risks.

2. Demonstrates that return and risk are interwoven.

3. Identifies and describes a general hierarchy of risks.

4. Allows for variations across funds in the importance of specific risks and/or risk tolerance.

5. Enables an assessment of risk using both quantitative tools and common sense.

Organisational considerations

6. Is “owned” by the governing fiduciaries.

7. Is clearly articulated and understandable to all fiduciaries.

8. Is practical and proactive in dealing with risk.

9. Can be implemented with results that can be reviewed and evaluated.

Six elements of a standard risk management process:

1. Identify and define all the sources of risk within a fund. (Note that we have focused on DB pension plans in our risk matrix in Exhibit 3, which lays out 57 risks. However, through minor modification, the matrix can easily be applied to an endowment fund by incorporating preservation of capital and spending/payout policy, or to a defined contribution plan by incorporating communication, advice and record-keeping risk factors.)

2. Understand the nature of the risks within your total investment context. Possible risk causes and scenarios should be developed.

3. Measure the potential magnitude and impact of those risks on your fund. Separate minor, more acceptable risks from major risks, and provide information to assist in the evaluation and treatment of risks.

4. Assess how and whether those risks have been addressed or mitigated by past actions. Develop a list of risks, in order of priority, that require further action.

5. Manage the largest unaddressed risks and utilise a hierarchical approach to sequentially deal with the remainder. Risk treatment involves identifying the potential range of options and preparing and implementing risk treatment plans. Document the process and your rationale.

6. Review the risks, continually utilising the hierarchy, while adding new risks as they are identified. Monitor the delegated risks to the desired outcomes. Set up a

Combining these six elements with the nine desirable attributes yields a fund-wide investment risk management framework that is practical and understandable for fiduciaries.


procedure for monitoring risk and evaluating the effectiveness of your risk treatment plan. Such an ongoing review is essential to ensure the relevancy of your approach.18

Combining these six elements with the nine desirable attributes yields a fund-wide investment risk management framework that is practical and understandable for fiduciaries. But this alone won’t get plan sponsors where they need to be. An entirely different approach to governance is also needed.

Exhibit 3: Russell’s general hierarchy of governing fiduciary concerns

18 Curwood, Bruce B. “A Comprehensive Risk Management Framework for Investment Funds”. Journal of Investment Consulting, Summer 2007; pp. 51–52.

Decision/risk Fiduciary Asset/liability Structural Implementation Operational

Governance 1. Legislative2. Political3. Decision-making4. Imprudent

delegation5. Publicity6. Documentation

10. Benefits 18. Maverick19. Research20. Initial due diligence

19. Procedural control 51. Resources52. Systems/

technology

Objective setting 7. Policy

Asset allocation

8. Socially responsible investing

9. Actuarial/ funding

11. Mismatch12. Assumptions13. Asset

mix/classes14. Model15. Downside16. Diversification

21. Rebalancing 30. Cash flow31. Leverage

Asset class strategy 17. Regret 22. Active/passive23. Currency24. Benchmark25. Derivatives

32. Hedging33. Tactical asset

allocation34. Timing35. Credit

Manager/ portfolio structure

26. Style27. Sector28. Country

36. Manager37. Holdings

concentration38. Tracking error39. Counterparty

Manager research and selection

40. Residual41. Call

53. Liquidity

Custody/execution 42. Trading43. Transition

54. Custodial55. Securities

lending56. Valuation

Performance/process evaluation

44. Tolerance45. Monitoring46. Control47. Contract48. Audit/accounting49. Oversight

50. Ongoing due diligence

57. Guideline breach

n High impact n High risk n Moderate risk n Low risk


Rethinking governance Good risk management policies and effective governance are intricately intertwined. The OECD report mentioned earlier communicated this when saying, “Deficiencies in risk management and distorted incentive systems point to deficient board oversight”. And, “Risk policy is a clear duty of the board in any organisation”.19 The Group of Twenty (G-20) finance ministers and Central Bank governors recently noted, “Failures in corporate governance must be addressed because they allowed the financial crisis to develop… good governance, overseen by responsible shareholders, addresses management of risks in a way that underpins prudential supervision and regulation”.20

The belief that managing risk is an essential part of good governance is reinforced by pension regulators, such as the Canadian Authority of Pension Supervisory Authorities (CAPSA). One of CAPSA’s guiding principles states, “The plan administrator should provide for the establishment of an internal control framework, commensurate with the plan’s circumstance, which addresses the pension plan’s risks”. In addition, CAPSA’s self-assessment questionnaire asks:

• Have you identified the pension plan’s risks?

• Do you have a process to manage these risks?21

What is important to understand is that CAPSA and other regulators are not offering up a precise prescription for how fiduciaries should manage risk. Instead, they are asking big-picture questions in an effort to motivate fiduciaries to think about risk from a broader perspective.

ADOPTING AN ENTERPRISE RISK MANAGEMENT CULTURE Exhibits 4a to 4c outline three modern depictions of a robust risk management process, all of which see risk from a more holistic, enterprise-wide view. In the modern context,

“Enterprise risk management is a structured and disciplined approach that supports the alignment of strategy, processes, people, technology and knowledge with the purpose of evaluating and managing the uncertainties an organisation faces as it creates value.

• Aligns with strategic intent and related objectives.

• Includes all risks, not just financial.

• Integrates into the management process, becoming every manager’s responsibility.

• Addresses both the hard and soft sides of risk management”.22

Fiduciaries must understand the complexities of risk and, to be effective, dramatically change the way they think and act. This means adopting an enterprise risk management culture, embracing appropriate behaviours and attitudes and developing and rewarding competencies. It also involves having all of the fiduciaries agreeing on their risk management beliefs. There are no shortcuts to changing a culture. Effective risk management requires collaborative and cohesive governance, which fosters top-down oversight, bottom-up involvement for risk-taking and coordinated monitoring of the process. In short, we must have effective communication around risk issues.

19 “The Corporate Governance Lessons from the Financial Crisis”. OECD Financial Market Trends, 2009; p. 17. 20 “Calls for governance to take center stage at G20”. Corporate Secretary, May 2009; p. 13. 21 “Regulatory Principles for a Model Pension Law”. Canadian Association of Pension Supervisory Authorities

(CAPSA), 2004.22 Makomaski, Joanna. “Harnessing the Power of ERM”. The Conference Board of Canada, 2009.


Exhibit 4: Depictions of robust risk management processes Exhibit 4a: Driving value through enterprise risk management

Source: Foley & Lardner, National Directors Institute — 2009 Web Conference Series, “Risk Management in the Boardroom”, March 18, 2009.

Exhibit 4b: The key to successful enterprise risk management practices depends on the behavioral attributes of the organisation at all levels

Source: Risk and Insurance Management Society (RIMS). Image copyright 2007 by the Risk and Insurance Management Society, Inc. All rights reserved.

The bottom line is that good fund governance is critical to acquiring and retaining the necessary skill sets to decide, implement and oversee the appropriate investment policies.

Core ERM Model

Ri

sk g

over

nanc

e

Risk awareness

Risk improvement

Risk strategyand objectives

Risk identification and prioritization

Riskassessment and

quantification

Risk responsestrategy

Risk responseimplementation

ERMsustainability Growth Profitability

Continuity

Conceptual skillsPlanningOrganizingDecision-makingManagement processEthical judgementOrganizational architectureStrategic thinking

Technical skillsRisk management processRisk analysisRisk controlRisk financingEnterprise risk managementProject managementInsurance knowledgeVendor relationsErmis and claims management

Interpersonal skills:LeadershipMotivatorNegotiationsConsensus builderTeam builder

Personal skills:MotivatedInnovativeExperiencedCommunicationConsultative

AccountingEconomicsFinanceLegalComplianceHuman resourcesAudit

ManagementInformationTechnologyMarketingOperationsStatisticsSecuritySafety

Business skills:

Corecompetency skills

ERM frameworks help organizations:Demonstrate proactive understanding and management of risk

• Advance Management and Board-level understanding of existing business risks and results and emerging risks and future prospects

Improve organizational health• Enhance corporate governance

• Streamline existing risk processes

• Improve business decision-making

• Support financial reporting

Gain competitive advantage• Anticipate and effectively respond to risk

• Improve the risk/reward ratio

• Better allocate resources


Exhibit 4c: Asset managers need to modernise risk management processes across four fronts

Source: U.S. Institute Risk Management Survey, “Recovering From the Storm”. McKinsey & Company, Inc. 2009.

In a risk-aware culture, communication is frequent and inclusive, resulting in multiple views of risk for increased transparency. So we must look beyond the normative, use multiple scenarios and stress test our assumptions. We must consider capital adequacy and liquidity in the event of systemic shock or unexpected volatility. We must strike the right balance between the quantitative and the qualitative, to establish risk-adjusted performance measures to influence behaviour and strategy. Risk management should be a dynamic, forward-looking and comprehensive process; however, it is just evolving. Nonetheless, the Risk and Insurance Management Society recently endorsed this approach in its Enterprise Risk Management Report, which stated that “organisations seeking better performance need to broaden and deepen their programmes to mature in the competency drivers that support front-line risk ownership, linkage and governance oversight…to build a culture of risk-adjusted decision making throughout an organisation”.23

Bill Martin, Chief Risk Officer of the Commonfund, summed it up nicely when he noted that “the key elements that drive the effective and proactive management of risk for any organisation fall into two dimensions:

1. Action—all staff members are able to act with confidence if they are supported by:

• Accurate, timely data.

• Information flow that gets the right analysis to the right person at the right time.

• Decision rights that clarify decision-making authority while eliminating unnecessary steps or inefficient processes.

• An ability to execute that focuses on minimising errors and providing feedback on actions taken.

23 Coffin, Bill. “The 2008 Financial Crisis: A Wake-up Call for Enterprise Risk Management”. Risk and Insurance Management Society, 2009; p. 7.

2. Governance, organization and processesAre the structures, systems, controls and infrastructure in place to manage risk and comply with regulatory requirements?

How robust is governance model?

3. Measurements and reportingAre risk factors effectively identified, measured and monitored as well as the correlations between them?Do reports highlight key risks?Are risk systems, reports, models and stress testing adequate for the sophistication of assets being managed?

4. CultureDoes the firm’s culture reinforce

risk management principles?What formal and informal mechanisms

support the right mindset and behaviors?

Integratedrisk-return

management

1. Insight and strategyWhat are the major risks to future performance, including contingent exposures?What is the firm’s desired risk profile? Are business decisions made with a clear view of their impact on the risk profile?Which risks should be owned? Which should be transferred or mitigated? Is risk capacity aligned with strategy?


2. Discipline—a quality control mechanism that seeks a higher level of consistency and predictability… during periods of acute uncertainty”.24

The bottom line is that good fund governance is critical to acquiring and retaining the necessary skill sets to decide, implement and oversee the appropriate investment policies. It is grounded in a deep understanding of markets and requires both expertise and education in managing risk. The human element therefore permeates all aspects of effective governance and risk management.

ADDRESSING THE HUMAN ELEMENT It’s critical to acknowledge that human beings are emotional and inconsistent, and that investment decision-makers, alone or in groups, may therefore fall victim to human frailty. Even our moods can momentarily affect our behaviour, so it must be remembered that our investment decisions generally extend far beyond the moment. Behavioural impediments are now, more than ever, being regularly discussed by sophisticated investment managers, academics and leading investment strategists, including Bill Gross, Warren Buffett, Keith Ambachtsheer and D. Don Ezra. When we fail to listen to both investment theory and practice, we do so at our peril. Therefore, wherever possible, we must rethink the consequences of our actions and not let our feelings rule our investment decisions. It’s vital to put sound practices in place to protect us from our emotions.

Jason Zweig provides a number of folksy but proven ways to make our brains work for us instead of against us, as shown in Appendix D. At the top of the list is “Control the controllable”, which speaks to our central premise—that risk management, and not expected return, is the cornerstone of investing. Similarly, the Vanguard article previously mentioned (“Group Decision-Making: Implications for Investment Committees”) outlines several recommendations for realising a group’s full potential. The firm’s research hints at several basic and easy-to-implement tactics that committees should consider, including inviting a devil’s advocate into the discussion of key issues. (See Appendix E.) These suggestions are just the tip of the iceberg. The main point is that discipline, prudence and adherence to basic guidelines, rooted in the fundamental research of decision making, can reduce behavioural problems and leverage the full power of individuals and groups.

Conclusion Poor governance in the pension industry has led to the mismanagement of risk, resulting in painful consequences for plan sponsors. Yet the need for better governance is not a new or revolutionary concept. Numerous strategists have conducted ample research on this topic over the last 20 years, documenting the need for change in fund and fiduciary behaviour. This is encapsulated in Exhibit 5 and well chronicled in Exhibit 1 and in a paper entitled “Narrowing the Knowing-Doing Gap in Investments through Effective Fund Governance”.25 Nevertheless, evolution and progress have been slow—some might say glacier-like.

Better governance starts with the understanding that risk management is indeed the cornerstone of investing—that investment management must begin with risk considerations and not with simply pursuing returns. Failing to grasp this, most plans have focused on optimising (uncertain) returns rather than managing the level of acceptable risk. Beyond misdiagnosing the fundamental problem, investors have narrowly interpreted the solution set, underestimated the complexity of risk and fallen prey to behavioural biases. As a result, they have spent too little time and effort on managing risk. The vast majority of their time

24 Martin, Bill. “Connecting the dots: Managing risk in an environment of unprecedented uncertainty”. Commonfund Mission Matters, Winter 2009; p. 14. Available at http://www.commonfund.com (accessed Dec. 15, 2009).

25 Curwood, Bruce B. “Narrowing the Knowing–Doing Gap in Investments through Effective Fund Governance”. Russell Monograph, October 2006.

Better governance starts with the understanding that risk management is indeed the cornerstone of investing—that investment management must begin with risk considerations and not with simply pursuing returns.


has been spent in routine administration and control, instead of in education, training and strategy assessment around risk management. This has led to a dearth of resources for addressing risk management and a widespread inability to focus and delegate.

Risk management cannot be summed up in a number, nor is it an exercise in compliance and box-ticking. It involves a well-defined process and intangibles like human behaviour and organisational culture. Douglas Hubbard, author of “The Failure of Risk Management”, made this point quite nicely when he said:

“…if you were to implement better methods for measuring risks, then you would have much better guidance for managing risks. To achieve an improvement, however, your organisation has to have a way to deal with barriers that are not part of the quantitative methods themselves. You need to break organisational silos, have good quality procedures, and incentivise good analysis and good decisions”.26

The behavioural aspects can be especially challenging. This is because good governance is required to overcome bad behaviour, but unconscious bad behaviour causes fiduciaries to take the easy way out and not focus on governance. This devious loop is analogous to what many of us have encountered when constructing a complex Excel spreadsheet and the infamous error message “Circuitous Reference” appears. You must either go back through countless data entries, many interconnected by complex formulae, to solve the loop, or scrap the spreadsheet entirely and start over again. Neither alternative is easy, but you can’t proceed until you correct the error. And no matter which route you choose, you must always keep the error itself in your mind, to solve it or avoid it. That is the key to fixing the Excel spreadsheet issue—and it may also be the key to fixing the governance-behaviour loop. If plan sponsors are to progress in solving this Catch 22 and, by doing so, enable a better risk management process, they must be consciously aware of their emotions and harness them through a better governance process. In other words, they must be consciously aware of the unconscious. Now we know the true challenge of risk management.

Is it any wonder that mega funds such as the Caisse in Canada and CalPERS in the United States have recently instituted massive governance/risk management initiatives, established the position of Chief Risk Officer and beefed up their risk management capabilities by adding over 20 new hires in the area?27 The question that remains is how mid-sized funds lacking economies of scale will compete, and whether they can create a competitive advantage.

26 Hubbard, Douglas W. “The Failure of Risk Management”. John Wiley & Sons Inc. 2009; p. 241. 27 “Caisse shakes up structure after big 2008 losses”. Reuters. 30 April 2009. <http://www.reuters.com/article/

idUSN3053734720090430>. “Risk Management Committee Chair and Vice Chair Elected”. CalPERS press release. 29 April 2009. <http://www.calpers.ca.gov/index.jsp?bc=/about/press/pr-2009/april/risk-management-chair-elected.xml>.


Exhibit 5: The chair and CIO play a vital role in fund governance

Source: Based on the “Proposed Board Effectiveness Model” of Richard Leblanc, PhD.

RESOURCES

TIMELY

DECIS

IONSRESEARCH / BELIEFS

A COMMONVISION

BOARD COMMITTEEMEMBERSHIP

Tenure / Recruitment / OrientationBehaviorExpertise

ExperienceTraining / Development / Appraisal

GOVERNANCEPROCESS

Proper business planAgendas and documentation

Investment educationand innovation oversite

Prioritize by importance

GOVERNANCESTRUCTURE

Unambiguous accountabilityFocust on what matters• Primary objective• Risk management• Strategy

Delegation to Professionals

BOARD EFFECTIVENESS


Appendices

Appendix A: Corporate governance lessons from the financial crisis—OECD report 2009

• Risk policy is a clear duty of the board.

• Deficiencies in risk management and distorted incentive systems point to deficient board oversight.

• A solid risk culture throughout the firm is essential.

• Boards need to be educated on risk issues and be given the means to understand risk appetite (tolerance) and the firm’s performance against it.

• Above all, boards need to understand the firm’s business strategy and have a forward-looking perspective.

• The composition of any risk committee is an important issue.

• The issue is not just independence and objectivity, but also capabilities (and may require formal education and training).

• A number of the members of the risk committee should be individuals with technical financial sophistication in risk disciplines.

Source: The Corporate Governance Lessons from the Financial Crisis”, OECD Financial Market Trends, 2009.

Appendix B: MetLife U.S. pension risk behaviour survey executive summary January 2009 This report represents 168 plan sponsor views on current risk management practices and identifies inconsistencies:

• “The study found most plan sponsors focus on a few risk factors rather than addressing the full range of relevant risks.

• On many occasions, plans were not addressing the risks they viewed as most important.

• The survey suggests a wide gap between the importance plan sponsors ascribe to each risk area and the sponsors own reported success at managing those risks.

• Many affirm their close attention to Plan Governance …however, the balance of the study’s findings suggest otherwise.

• This research indicates the possibility decision-makers are failing to effectively align importance and success across risk factors”.

“The results of this study should be a call to action for pension plans not yet comprehensively and systematically managing risk”.

Source: “MetLife U.S. Pension Risk Behaviour Study: A Study of Risk Management Attitudes and Aptitude Among Defined Benefit Pension Plan Sponsors”, MetLife, 2009.

Appendix C: Mandelbrot’s ten heresies of finance 1. Markets are turbulent.

2. Markets are very risky —more risky than the standard theories imagine.

3. Market timing matters greatly—big gains and losses concentrate into small packages of time.


4. Prices often leap, not glide—that adds to the risk.

5. In markets, time is flexible.

6. Markets in all places and ages work alike.

7. Markets are inherently uncertain, and bubbles are inevitable.

8. Markets are deceptive.

9. Forecasting prices may be perilous, but you can estimate the odds of future volatility.

10. In financial markets, the idea of “value” has limited value.

Source: Mandelbrot, Benoit, and Richard L. Hudson. The (Mis)Behaviour of Markets: A Fractal View of Risk, Ruin & Reward. Basic Books, 2004.

Appendix D: Zweig’s suggestions for making our investing brains work for us • Control the controllable—we cannot control the return on our investment, but we can

control our expectations, the amount of risk we take and our readiness or patience.

• Stop predicting and start restricting—don’t make too many large bets and consider smaller allocations over time via dollar-cost averaging.

• Ask for evidence—ensure you have good investment research to support your decision.

• Try to prove yourself wrong—the refusal to second-guess yourself can lead to huge losses or crippling regret.

• Know yourself—your attitudes toward financial risk can differ drastically depending on how they are framed.

• Get reframed—seeing risk estimates from alternative perspectives in both percentage and personal terms may limit overconfidence (e.g., a 78% chance of success versus 22 out of 100 investors get it wrong).

• Track your feelings—writing down your past emotions may help you realise that you are often overenthusiastic when markets are rising and vice versa.

• When the market blinks, blink back—doing advance research may enable you to take advantage of investment strategies that are undervalued or currently out of favor (i.e., take advantage of mean reversion).

• Only fools invest without rules—don’t be governed by guesswork but by clear policies and procedures.

• Write it down, right away—document your reasons for investing before you buy.

Source: Zweig, Jason. Your Money & Your Brain: How the Science of Neuroeconomics can Help Make You Rich. Simon & Schuster, 2007.

Appendix E: Vanguard’s recommendations for realising a group’s full potential • Engender healthy debate with heterogeneous groups—member diversity in knowledge,

skills, abilities, personality, attitudes and demographics can help.

• Invite a devil’s advocate or outside members, such as expert professionals, to discuss key issues—differing opinions can remedy the negative effects of groupthink, reduce confirmation bias and limit group polarising effects.

• Minimise dysfunctional behaviour like social loafing—mutual trust and high involvement in group activities make members feel personally responsible.


• Make committee members aware of decision-making biases—knowledge, which comes from reporting, self appraisals and external evaluations, can begin to remedy these biases.

Source: Mottola, Gary, R. & Utkus, Stephen, P., “Group Decision-Making: Implications for Investment Committees”, Vanguard Investment Counseling & Research, 2009.

Related readingAmbachtsheer, Keith, Ronald Capelle and Hubert Lum. “The Pension Governance Deficit: Still With Us”. Rotman International Journal of Pension Management, Fall 2008.

Armin, Janine. “A Little More Conversation”. Corporate Secretary, May 2009.

Bernstein, Peter L. “A Fresh Look at the Super Part of the Debt Supercycle”. Economics and Portfolio Strategy, March 15, 2009.

Bernstein, Peter L. “The Moral Hazard Economy”. Harvard Business Review, July / August 2009.

Bhansali, Vineer. “The Equity Risk in a Bond Manager’s World”. PIMCO Canadian Quarterly Review, fourth quarter 2008.

Black, Pamela J. “The Three Faces of Risk; What You Need to Know about Risk Tolerance”. The Global Association of Risk Professionals (GARP.com) newsfeed, 2009.

Bookstaber, Richard. “A Demon of Our Own Design: Markets, Hedge Funds, and the Perils of Financial Innovation”. John Wiley & Sons Inc., 2007.

Campbell, Andrew, Jo Whitehead and Sydney Finkelstein. “Why Good Leaders Make Bad Decisions”. Harvard Business Review, February 2009.

Coffin, Bill. “The 2008 Financial Crisis: A Wake-up Call for Enterprise Risk Management”. Risk and Insurance Management Society Inc., 2009.

Collie, Bob, and D. Don Ezra. “Resist the Amygdala!” Russell Research, March 2007.

Crowley, Scott. “Build the Plan. Success Will Follow”. Meyers Norris Penny (Canada), 2009.

Curwood, Bruce B. “A Comprehensive Risk Management Framework for Investment Funds”. Journal of Investment Consulting, Summer 2007.

Curwood, Bruce B. “Narrowing the Knowing–Doing Gap in Investments through Effective Fund Governance”. Russell Monograph, October 2006.

Damadoran, Aswath. “Strategic Risk Taking: A Framework for Risk Management”. Wharton School Publishing, 2007.

Eddy, Kevin. “A Little Less Complication”. Corporate Secretary, May 2009.

Finkelstein, Seth, and Pierre Couture. “Get Funded, Stay Funded: A More Active Approach to Pension Risk Management”. ING PlainSpeak Series, January 2009.

Hubbard, Douglas W. “The Failure of Risk Management: Why It’s Broken and How to Fix It”. John Wiley & Sons, 2009.

Ilkiw, John H. “Pension Fund Governance: Prudence is Process, Not Performance”. Russell Canada Monograph, February 1996.

“Calls for governance to take center stage at G20”. Corporate Secretary, May 2009.

Johnson, Keith L., and Frank Jan de Graaf. “Modernising Pension Fund Legal Standards for the Twenty-First Century”. Rotman International Journal of Pension Management, Spring 2009.

Jorion, Philippe. “Risk Management Lessons from the Credit Crisis”. European Financial Management, April 2009.

Kazemi, Kambiz. “2008: Volatility in Action”. Toronto CFA Society Analyst, March 2009.

Li, Geng. “The Missing Link”. CSJ, August 2008.


Lyons, Sean. “The Changing Face of Corporate Defence in the 21st Century”. Corporate Governance Quarterly, Spring 2009.

Makomaski, Joanna. “Harnessing the Power of ERM”. The Conference Board of Canada, 2009.

Mandelbrot, Benoit, and Richard L. Hudson. The (Mis)Behaviour of Markets: A Fractal View of Risk, Ruin and Reward. Basic Books, 2004.

Martin, Bill. “Connecting the dots: Managing risk in an environment of unprecedented uncertainty”. Mission Matters, Winter 2009.

Mauboussin, Michael J. “Long-Term Investing: How I Learned to Stop Worrying and Ignore Volatility” (excerpts from comments at the Greenwich Roundtable, December 11, 2008). Legg Mason Capital Management, December 15, 2008.

“What Do We Need to Fix?” Legg Mason Capital Management, July 16, 2009.

“MetLife U.S. Pension Risk Behaviour Study: A Study of Risk Management Attitudes and Aptitude Among Defined Benefit Pension Plan Sponsors”. MetLife, 2009.

Mitchell, Roger. “Risk: The Final Frontier”. CFA Magazine, March–April 2006.

Mottola, Gary R., and Stephen P. Utkus. “Group Decision-Making: Implications for Investment Committees”. Vanguard Investment Counseling & Research, 2009.

Orrell, David. “The Future of Everything: The Science of Prediction”. Thunder’s Mouth Press, 2007.

Otar, Jim. “Monte Carlo Simulations: Are They Worth the Gamble?” The Monitor, July/August 2007.

Persaud, Avinash D. “Risk models just run with the herd”. RBC Dexia Investor Services Perspectives, September 2009.

Pilkey, Orrin H. and Linda Pilkey-Jarvis. Useless Arithmetic: Why Environmental Scientists Can’t Predict the Future. Columbia University Press, 2007.

Prybylski, Hank. “The Hard Work of Culture-Building”. Risk Professional, October 2009.

Morgan Stanley Public Fund CIO Survey. Morgan Stanley Investment Management Publications, 2008.

“The Corporate Governance Lessons from the Financial Crisis”. OECD Financial Market Trends, 2009.

Rebonato, Riccardo. “Plight of the Fortune Teller”. Princeton University Press, 2007.

“Recovering from the Storm: The New Economic Reality for U.S. Asset Managers”. McKinsey & Company, Inc., and Institutional Investor U.S. Institute survey, 2009.

“Regulatory Principles for a Model Pension Law”. Canadian Association of Pension Supervisory Authorities (CAPSA), 2004.

Savage, Sam L. “The Flaw of Averages: Why We Underestimate Risk in the Face of Uncertainty”. John Wiley & Sons, 2009.

Schacter, Harvey. “Recognising the Emotional Side of Business”. The Globe and Mail Report on Business, Oct. 21, 2009.

Stulz, Rene M. “Six ways to mismanage risk”. Harvard Business Review, March 2009.

Taleb, Nassim N. “Fooled by Randomness”. Random House, 2004.

Taleb, Nassim N., Daniel G. Goldstein and Mark W. Spitznagel. “The Six Mistakes Executives Make in Risk Management”. Harvard Business Review, October 2009.

Zweig, Jason. “Your Money and Your Brain: How the Science of Neuroeconomics Can Help Make You Rich”. Simon & Schuster, 2007.

/ p 43

Prudence is process, not performance

This article was first published as a Russell Canada Monograph in 1996. Its mantra is one that underpins best practices in risk management, but that is too easily forgotten and too frequently overlooked in the face of constant pressure to achieve short-term performance. The foundation for long-term success must lie in an effective structure. Decisions should be based on sound information, and made by informed experts. Implementation should be effective and, again, carried out by those properly equipped to do so. Monitoring and review provide more information and allow the cycle to continue and improve. In the absence of such a process, strong short-term performance is still possible but sustainable success is not.

Russell Investments // Risk management: A fiduciary’s guidebook // Prudence is process, not performance

/ p 44Russell Investments // Risk management: A fiduciary’s guidebook // Prudence is process, not performance

Prudence is process, not performanceIn a pension fund framework, governance is the management of plan assets by responsible fiduciaries. Pension fiduciaries have unilateral discretion over other people’s assets; one of their legal requirements is to act prudently. Prudence can be described as the exercise of good judgment to avoid decisions that result in undesired consequences. This Conversation Piece explains governance practices that will help fiduciaries exercise good judgment on behalf of plan members who rely on fund assets for their retirement income, but have little or no control on how those plan assets are managed.

Good judgement hopefully results in good performance—but it is not assured because investments are always subject to disappointing performance. This principle was established in the 1830 Harvard College decision that introduced the prudent man concept. The Massachusetts Supreme Court concluded that trustees had not breached their fiduciary duty by investing in stocks even though the value of the stocks had declined. “Prudent men” will consider “probable income” and “probable security of capital” when making investments, but neither income nor capital can be assured even with

“safe” investments. In the words of the court: “Do what you may, the capital is at hazard.”1 The prudent person concept has evolved into a common pension governance theme in the United States, Canada, the United Kingdom, Australia and New Zealand and can be summarized in the more contemporary aphorism: prudence is process, not performance.2

Experience indicates that three governance practices foster good judgment:

• Using three types of fiduciaries

• Following a four-step Understand-Plan-Do-Review fiduciary cycle

• Implementing an accountable reporting structure

By: John H. Ilkiw

November 1996

1 Longstreth, B., Modern Investment Management and the Prudent Man Rule, Oxford University Press, 1987, pp. 27–28. 2 Section 404(a)(1)(B) of the Employment Retirement Income Security Act (ERISA) of 1974 as amended. The Pension

Benefits Act, 1987 (Ontario) codifies the prudent person concept in two sections: Section 22(1): The administrator of a pension plan shall exercise the care, diligence and skill in the administration and investment of the pension fund that a person of ordinary prudence would exercise in dealing with the property of another person. Section 22(2): The administrator of a pension plan shall use the administration of the pension plans and, the administration of the investment of the pension fund, all the relevant knowledge and skill that the administrator possess or, by reason of his or her profession, business or calling, ought to possess.


Three fiduciary typesWhile all fiduciaries must act prudently, not all fiduciaries have the same expertise or level of responsibility. Governing fiduciaries—members of corporate boards and boards of trustees—are ultimately responsible for ensuring pension assets are prudently invested. Governing fiduciaries usually employ managing fiduciaries—often in the form of an Investment Committee— to provide advice and oversee policy implementation. Managing fiduciaries in turn employ a number of operating fiduciaries with specialized knowledge and skills to implement and manage investment policies on a daily basis. For larger funds, operating fiduciaries could include the Director of Investments who is usually the working arm of the Investment Committee, the custodian/master trustee who holds the assets, settles transactions and reports on fund activities, and the investment manager who invests fund assets. Non-fiduciary agents such as actuaries, lawyers and investment consultants are used to advise fiduciaries on investment related issues.

In the mistaken belief that they are lessening their exposure to liability for breach of fiduciary duty, many governing fiduciaries retain decision powers that are best delegated to managing and operating fiduciaries. In such cases governing fiduciaries assume additional responsibilities because they do not have the time, knowledge or skills required to make timely and sufficiently informed decisions. They are more likely to satisfy their fiduciary duty by appropriately delegating decisions to managing and operating fiduciaries.

Four-step fiduciary cycleSystematic and continuous application of the four-step Understand-Plan-Do-Review fiduciary cycle illustrated in Exhibit 1 can assist fiduciaries in governing and managing their fund.3 It is particularly important that governing and managing fiduciaries follow a cycle like this because they share the final responsibility for investment policies and, if ever the occasion arises, having followed this type of process provides evidence of the fulfillment of fiduciary duty.

Step one requires a clear understanding of the duties and standards of care required by common law and legislation and of key actuarial, investment, organizational and regulatory principles that apply to pension plans as well as the plan document itself. Step two is planning how assets are to be managed, including establishing governance, funding, investment, accounting policies and benchmarks. Step three is implementation which involves documenting decisions and employing and supervising operating fiduciaries to execute tasks. Step four is the review of fund activities and performance to determine how well the pension system is functioning relative to benchmarks established in step two and documented in step three. A good review process provides governing and managing fiduciaries with an ever-improving understanding of the system for which they have responsibility. It is a mechanism for developing and exercising good judgment.

We have observed that some governing and managing fiduciaries make decisions before having an adequate understanding of the issues being decided. They believe their occupational and personal experiences provide them with sufficient knowledge to make prudent pension investment decisions. Those who do recognize the need for a better understanding of pension-specific investment issues are often unable to do so because of other pressing responsibilities.

Most professional operating fiduciaries invest a significant portion of their resources to improve and expand their understanding of their chosen specialty areas because their expertise provides them with their living. Most governing and managing fiduciaries may not

3 I have added “Understand” to the Plan-Do-Review fiduciary cycle introduced by Janine Baldridge and D. Don Ezra in “The Fiduciary’s Guide to Investment Management,” Russell Research Commentary, December 1993. While it is obvious that governing fiduciaries should understand the principles that drive pension plans for which they are responsible, experience indicates that this is not always the situation.


have the same pecuniary incentive, but they arguably have a more pressing incentive: an inadequate understanding of issues for which they have accepted responsibility might result in an unintended imprudent decision and exposure to a potential liability.

Exhibit 1: Four-step fiduciary cycle

Accountable reporting structureAn accountable reporting structure is the third feature of successful and prudent fund governance. This can be achieved by meshing the three fiduciary levels with the four-step fiduciary process as illustrated in Exhibit 2. The structure is characterized by the downward delegation of decision making, and informative and timely upward reporting of compliance and performance.

Exhibit 2: Accountable reporting structure

In this framework, governing fiduciaries retain responsibility for those issues they are best positioned to address, and delegate to managing and operating fiduciaries responsibilities for those issues they are better equipped to decide. For example, governing fiduciaries invariably retain the responsibility for approving the long-term fund objectives but almost always delegate the strategic asset management to managing fiduciaries and day-to-day asset management to operating fiduciaries. In general, the smaller the impact a given decision has on fund objectives, the further down it can be delegated. That is why most governing fiduciaries decide a fund’s long-term asset mix

1

2 PlanReview

Do

Understand

3

4

Fiduciary level Understand Plan Do Review

Governing

Managing

Operating

Informative and timely reports Prudent and

profitable delegation


policy—which is the principal determinant of long-term performance—but delegate to operating fiduciaries security selection and marketing timing decisions—which have a meaningful but nonetheless secondary impact on long-term performance.

Informative and timely reporting of compliance and performance is the feedback mechanism for verifying and evaluating investment activities and their contribution to achieving established fund objectives. “Informative” means a package of relevant information designed to allow a supervising fiduciary to verify compliance with established investment policies and understand the impact of key investment decisions.

“Timely” means a reporting frequency that corresponds to the period of time over which decisions can be reasonably monitored and evaluated. In general, reporting detail and frequency increases as one moves down the chain of delegation. Governing fiduciaries may meet as infrequently as once a year to verify compliance and evaluate performance using high-level reporting formats. Managing fiduciaries usually meet quarterly and, in many cases monthly, and utilize more comprehensive reports. Operating fiduciaries meet most frequently and their reports are more detailed because they often make decisions within very short time frames, often daily. For example, many investment managers routinely verify compliance of a portfolio with legislated and client-imposed restrictions at least weekly and, often daily, during periods of increased transactions or market activity.

ConclusionThe Understand-Plan-Do-Review cycle is an important step toward prudent fund governance. Most people will recognize that this four-step process is not unique to pension fund management, but is integral to the successful management of any organization. This process is of paramount importance to fiduciaries because it maximizes the likelihood that they will exercise good judgement on behalf of plan beneficiaries who rely on the fund’s assets for their retirement income but have little or no influence on how those assets are managed.

/ p 49

Thoughts on ERISA’s requirements regarding risk management

The legislation that governs fiduciary responsibilities varies by investor. Each country has its own laws, and these in turn can differ according to whether an investor is a corporate pension plan, an endowment or foundation, an insurance company, a public entity of some sort, or whatever. In each case, fiduciaries should take care to understand their own legal obligations.

In this short paper, we explore these obligations from the perspective of U.S. corporate retirement plans, as defined by the Employee Retirement Income Security Act of 1974 (ERISA).

Russell Investments // Risk management: A fiduciary’s guidebook // Thoughts on ERISA’s requirements regarding risk management

/ p 50

Thoughts on ERISA’s requirements regarding risk management Issue: As risk management becomes a greater focus for pension plans, fiduciaries should be aware of their legal duties in this regard. What does ERISA have to say about risk management?

Response: The letter of the law is only a starting point for best practices in risk management; ERISA does not address all of the considerations that should be taken into account in this area. Nonetheless, fiduciaries should be sure they know what the law says.

The fiduciary responsibilities defined by ERISA are based on prudence, a concept best looked at through a risk lens. It has been well said that “prudence is process, not performance.” Process can be demonstrated through documentation.

ERISA also sets diversification requirements for fiduciaries.

Background

FIDUCIARY RESPONSIBILITIES SET OUT IN ERISA According to the Employee Benefits Security Administration:

“Fiduciaries have important responsibilities and are subject to standards of conduct because they act on behalf of participants in a retirement plan and their beneficiaries. These responsibilities include:

• Acting solely in the interest of plan participants and their beneficiaries and with the exclusive purpose of providing benefits to them;

• Carrying out their duties prudently;

• Following the plan documents (unless inconsistent with ERISA);

• Diversifying plan investments; and

• Paying only reasonable plan expenses”.1

These responsibilities are clearly focused on the risk perspective. The maximization of return (an objective often treated as the de facto primary goal of the investment program) is not listed here, and in fact ERISA does not set out obligations for fiduciaries in that regard. Nor should it. In a recent Russell research piece, Bruce Curwood made the case that risk management—not return maximization—is the cornerstone of investing.2 That research looks at the operation of an institutional investment program from the perspective of best practice: how should it be run? Best practice goes well beyond the

Russell Investments // Risk management: A fiduciary’s guidebook // Thoughts on ERISA’s requirements regarding risk management

By: Bob Collie

March 2010

1 U.S. Department of Labor Employee Benefits Security Administration (2008).2 Curwood (2010).

/ p 51Russell Investments // Risk management: A fiduciary’s guidebook // Thoughts on ERISA’s requirements regarding risk management

bare requirements of the law, but fiduciaries should be aware of those requirements as a starting point. ERISA’s emphasis on prudence points to the risk-based approach to running an investment program as being consistent with the law’s requirements.

PRUDENCE: IT’S ABOUT PROCESS, NOT PERFORMANCE Prudence concerns process.3 The list above explicitly states, for example, that plan documents must be followed. That’s one part of a good process. There’s obviously much more to prudence than that, though.

The origin of the “prudent man” concept for investment fiduciaries is commonly traced back to an 1830 decision of the Massachusetts Supreme Court concerning an investment loss that had been incurred in the endowment of Harvard College. Justice Samuel Putnam ruled: “All that can be required of a trustee to invest is that he shall conduct himself faithfully and exercise sound discretion. He is to observe how men of prudence, discretion and intelligence manage their own affairs…”.4

ERISA added a twist to the prudent man concept. It requires that a fiduciary should act “with the care, skill, prudence, and diligence, under the circumstances then prevailing, that a prudent man acting in like capacity and familiar with such matters would use in the conduct of an enterprise of a like character and with like aims” (italics added).5 The italicized phrase is the key here, and has led to the concept of the “prudent expert”—prudence must be supplemented by a degree of familiarity.

The law, then, places emphasis not so much on outcomes, but rather on process and on documentation. Good documentation serves not only to support a sound process but also to demonstrate it, in the event of a legal challenge.

APPLICATION TO DEFINED BENEFIT (DB) AND DEFINED CONTRIBUTION (DC) PLANS The same law covers both DB and DC arrangements. However, the process by which investment expertise is applied in DC decision making is typically different than in DB, and different agency considerations also apply (such as, for example, who it is that bears the direct impact if expenses are unnecessarily high). As a result, we can characterize legal activity around DC as having more of a consumer-protection flavor than in DB. Nonetheless, the underlying legal requirements are the same in both cases.

DIVERSIFICATION: SIMPLE, CHEAP AND (USUALLY) EFFECTIVE Diversification is almost as old as the hills,6 and lawmakers around the world have long looked to diversification as a source of protection against risk. And, indeed, for most investment programs, diversification really is a simple and generally effective tool for reducing overall portfolio volatility.

We would caution, however, that diversification should not be seen as an end in itself. Rather it is a means of managing risk in most but not necessarily all circumstances. The precise requirement of ERISA is that fiduciaries invest “by diversifying the investments of the plan so as to minimize the risk of large losses, unless under the circumstances it is clearly prudent not to do so” (italics added).7 One example of where it can be prudent

3 See for example Ilkiw (1996).4 Justice Samuel Putnam, Harvard College v. Amory, Massachusetts 1830.5 ERISA section 404(a)(1)(B).6 See, for example, Murphy (2009). 7 ERISA section 404(a)(1)(C).

The law, then, places emphasis not so much on outcomes, but rather on process and on documentation.

/ p 52Russell Investments // Risk management: A fiduciary’s guidebook // Thoughts on ERISA’s requirements regarding risk management

not to diversify would be when a defined benefit plan is matching its investment strategy as closely as possible to the liabilities in order to minimize uncertainty around funded status and contribution requirements, hence protecting the security of benefits. A fully hedged strategy in this example may consist of large amounts of fixed income or annuity assets. While ERISA predates the advent of liability-driven investing, the Department of Labor has issued an advisory opinion stating that it is reasonable to take liability characteristics into account in setting an investment strategy.8 Fiduciaries should note, however, that the onus is on them in this case to be sure that they can demonstrate prudence.

In short, ERISA requirements as regards risk management are built on the concept of prudence. Fiduciaries who apply themselves to building soundly governed programs that follow clear, logical processes should find themselves in compliance with the law’s demands.

Related readingCurwood, B. (2010). “Risk management is the cornerstone of investing.” Russell Viewpoint, February.

Ilkiw, J. (1996). “Prudence is process, not performance.” Russell Conversation Piece, November.

Murphy, A. (2009). “The long march of diversification —1500 BCE to today.” Great Moments in Financial History: Russell Communiqué, fourth quarter.

U.S. Department of Labor Employee Benefits Security Administration (2008). “Meeting Your Fiduciary Responsibilities,” October. Available at http://www.dol.gov/ebsa/publications.

Walkey, D. (2006). “Is LDI consistent with ERISA diversification requirements?” Russell Practice Note, October.

We would caution, however, that diversification should not be seen as an end in itself. Rather it is a means of managing risk in most but not necessarily all circumstances.

8 This opinion is appended to Walkey (2006).

/ p 53

Resist the Amygdala!

In keeping with our argument that a great deal of the risk faced by investors is unquantifiable and behavioral, we believe that one component of a solid grounding in risk management is an understanding of the way in which the human animal makes decisions. The field of behavioral finance received a great deal of attention in the early 2000s as the investment world started to pay attention to decision-making research and significant developments in the field of neuroscience. The following pages are excerpts from an overview of behavioral finance that was published in 2007.

Russell Investments // Risk management: A fiduciary’s guidebook // Resist the Amygdala!

/ p 54

EXCERPTS FROM:

Resist the Amygdala! Improving institutional investment decision makingContext: Our stone age brainsWe know very little about our brains. We know what comes out of them, but we don’t know why. We haven’t mapped the inside, the neural network, the way we’ve mapped our cardiovascular system. But it appears that we’ve made a start.

Researchers1 are using the latest breakthroughs in technology to trace the exact circuitry our brains use to make decisions. It turns out that our modern skulls house a Stone Age mind. For example, there’s an area called the amygdala that is responsible for generating emotional responses with lightning speed. This was invaluable for survival when we needed to react instinctively to a vivid or sudden sight or sound, because our survival depended on it when we lived by hunting and gathering, not by monthly pay-checks. If it turned out that we were mistaken about the need to panic, it was no big loss; at least we survived.

Today, though, if the amygdala kicks in as a response to a sudden stock market decline we might dump all our stocks, an action with a genuine long-term negative consequence. What’s more, activity in the amygdala triggers the release of adrenaline, which has been found to “fuse” memories, thus preserving them. And a financial losing streak heats up action in the hippocampus, a part of the brain next to the amygdala that helps program our memories of fear and anxiety.

That’s the physical explanation of why market crashes, which actually make stocks cheaper to buy, also make us less willing to buy them. And why the actual experience of risk and loss is so much more powerful than our rational anticipation of it. That’s the physical explanation for why regret is so powerful.

THE MORE RECENT PAST: HOW BEHAVIORAL FINANCE NEARLY HIT THE HEADLINES IN 1948, AND THE CONCEPT OF ACTING “AS IF”As long ago as 1948, it was observed by Friedman and Savage that some of the same individuals who buy insurance policies also buy lottery tickets2. On the surface this appears inconsistent, they argued, because in the one case a premium is being paid to avoid risk while in the other a premium is paid to take on risk. They went on to produce an implied utility function which explained this action within the traditional rational framework. They acknowledged that the utility function they produced is “rather peculiar,” in that it implies that the marginal utility of $1 increases at some level of wealth above the purchaser’s current level. (It is much more common to assume that utility falls as wealth increases: that an extra $1 is worth more to a pauper than to a millionaire.)

By: Bob Collie and D. Don Ezra

February 2007

1 Zweig, J. (2002). Are you wired for wealth? MONEY Magazine, September 27.2 Friedmann, M and Savage, L.J. (1948). The utility analysis of choices involving risk. The Journal of Political

Economy, 56(4), 279-304. http://home.uchicago.edu/~vlima/courses/econ200/spring01/friedman.pdf.

Russell Investments // Risk management: A fiduciary’s guidebook // Resist the Amygdala!

/ p 55Russell Investments // Risk management: A fiduciary’s guidebook // Resist the Amygdala!

In and of itself, the apparent contradiction between purchasing insurance and purchasing lottery tickets is not of particular interest here. For one thing, the definition of a lottery ticket as a high risk investment seems questionable. The potential loss on a $1 ticket is $1. That is not a lot of risk. For another, institutional investors (our subject in this paper) have shown little interest thus far in lottery tickets as a potential new asset class.

What we do find interesting is a digression the authors make towards the end of the article. They ask: “is it not patently unrealistic to suppose that individuals consult a wiggly utility curve before gambling or buying insurance, that they know the odds involved…, that they can compute the expected utility… and that they base their decision on the size of that utility?”

To which the obvious answer seems to be “yes”. But Friedman and Savage appear not to have wanted to pursue the implications of that answer. Instead, they argued that: “this objection is not strictly relevant. The hypothesis does not assert that individuals explicitly or consciously calculate and compare expected utilities… the hypothesis asserts rather that, in making a particular class of decisions, individuals behave as if they calculated and compared expected utility and as if they knew the odds.”

They go on to assert, despite the contortions required to justify an in-out-shaking-all-about utility function, that: “Whatever the psychological mechanism whereby individuals make choices, those choices appear to display some consistency”. This despite the fact that their paper was built around an inconsistency.

Fifty years later, there is much greater willingness to accept behavioral explanations. This makes us more willing to acknowledge (indeed to seek out) inconsistencies. If we can identify inconsistencies in our behavior, the cause of these inconsistencies can be explored and this in turn opens the door to improving our decisions.

In other words, whatever the process that individuals actually follow, it may in most cases produce a result “as if” they were correctly applying the rational utility maximization theory of traditional economics, but in some instances that process fails. The purchase of a lottery ticket is such an instance of failure.

After all, prior to the invention of games of chance, human beings never dealt with situations in which odds, or payoffs, were known, or even constant. There is no such thing as a fair coin in nature, still less a situation in which a one in 20 million event can be, or needs to be, distinguished from a one in five million. The ability to accurately ascribe a utility to a one in 20 million chance of winning $10 million is not an ability that the stone-age brain would possess, and neither should it be expected to act “as if” it could.

MORE CONTEXT: HEURISTICS AND WHY THEY ARE GENERALLY GOOD While the traditional economists’ concept of the rational person depends on the idea of “as if” in order to be applied to actual decisions, the behavioralists’ concept of decision making is built around the idea of heuristics. Heuristics are simply rules of thumb. While the term is not often used outside the behavioral field, it is so widely adopted within it that we have caved in and used it throughout this paper.

Behavioral finance attempts to answer the question: “So if we are not actually consulting a utility curve when making a decision, what are we doing?” Most of the time, we are following very basic rules. These rules speed up and simplify the decision by, for example, eliminating almost all of the available information. As a result, we are well suited to making adequate decisions very quickly. Malcolm Gladwell has popularized this idea with blink.3 Gladwell does not merely observe that most of our decisions are made by the subconscious mind; he explores ways in which snap judgments and unconscious reactions can be educated and controlled.

3 Gladwell, M. (2005). blink. Little, Brown.


Just about every behavioral text will acknowledge that the use of heuristics produces many good decisions. They are a very efficient way to make decisions, because they require less information and are faster than more complex approaches. And a well-designed heuristic, even a very simple one, can produce decisions almost as good as far more complex processes, occasionally even better. At other times, however, these simple instincts lead to biases and sub-optimal decisions.

Gary Klein has analyzed how decisions are made in the field by experienced firefighters, doctors and tank platoon leaders, among many others. The following table shows how he categorizes the uses of what he terms the Recognition-Primed Decision Model (i.e. identifying a single reasonable course of action based on experience and pursuing it) and the Rational Choice Model.4

Rational choice model

Task conditions Recognition-primed decisions Rational choice strategy

Greater time pressure More likely

Higher experience level More likely

Dynamic conditions More likely

Ill-defined goals More likely

Need for justification More likely

Conflict resolution More likely

Optimization More likely

Greater computational complexity More likely

This table allows us to see different roles for different types of decision making. If time is limited, the situation changing, objectives ill-defined or important data missing, then some alternative to the rational choice strategy is needed, i.e. a heuristic-based approach. Because most decisions we face fall into those categories, that is in fact our normal approach.

As Klein’s table shows, this should not be regarded as a battle between “rational” decision making and “natural” decision making. There is a role for both. Investment is not a mathematical puzzle to be solved; it is a skill. We will expand upon this point further in the final section of this paper. But investment is a field (professional card playing and computer code-writing would be other examples) in which the role of rational analysis is greater than in most fields and in which our natural instincts can easily do harm. Indeed, Klein specifically mentions investment portfolio analysis as an example of a computationally complex decision. Intuition alone, even that of an experienced expert, is not necessarily going to lead to an appropriate decision.

It is our goal in this paper to highlight the instances where heuristics fail in the investment field. That is why we have started by acknowledging that the way we work is not always “as if” we were economically rational persons but rather less mathematical and altogether more human beings.

4 Source: Klein, G. (1998). Sources of Power. Massachusetts Institute of Technology Press.


MORE NEUROSCIENCE, SOME GOOD NEWS AND A RALLYING CRYWe have discussed the amygdala, which offers a physical explanation for the important human instincts of fear and panic. There is also a physical explanation for how the human brain enabled us to move beyond hunting and gathering, to reach the moon, to build civilizations and to design savings and investment structures that provide for our needs in old age. That explanation seems to lie in the thinking, or planning, part of the brain called the prefrontal cortex. This allows us to store sets of events in the form of memories, to draw general conclusions from particular data, to help forecast the consequences of our actions and to compare past and current experiences. In this way the prefrontal cortex helps us to reach more balanced judgments. We have to give it more power. We need to realize how our minds work. And since we can’t change our brains, we have to cope with their characteristics.

Over-reaction is damaging in an investment context and we have to discipline ourselves to avoid panicky action. Even though investment is a skill, not a mathematical puzzle, real investment decisions are too often clouded by the emotions of panic and over-reaction. So the amygdala is a part of the brain which seems to make little if any useful contribution to effective investment. As such, it provides a convenient target and a valuable rallying cry for the main theme of this paper. To all those involved in the investment of institutional capital, we therefore say:

Resist the amygdala! Assist the pre-frontal cortex!


The application of behavioral findings to institutional investorsIn this section, we will begin to explore how the findings described so far can be applied by institutional investors. If behavioral finance is to become more than an interesting diversion, we must look to how we might apply its lessons to improve investor behavior.

In looking for ways to improve, we inevitably focus on weaknesses. This should not be taken as meaning that we regard current standards of decision making as low; we do not. It is simply a recognition that the decision-makers are human beings. Human beings are emotional and they are inconsistent. They are prone to distorted decision making, overconfidence and all of the other characteristics described in this paper [these descriptions are not included in this excerpt]. They get on better with some other human beings than they do with others. We’re all like that.

And simply by recognizing that, we have made a valuable start. By reading the summary of findings in this paper, you will become aware of some natural tendencies we all have. You are more likely to regard your initial responses to situations with some skepticism. You are less likely to deny the role of chance in the outcomes you have experienced. You are more likely to ask “why might I be wrong” before jumping to a half-thought-out conclusion. You are more willing to express doubt in a group setting, when everybody else appears to be in agreement.

That is valuable. You are a better investor.

We will close our paper with three short essays, each setting out our thoughts on the application of these findings to a specific area of institutional investment decision making.

The first essay looks directly at the way in which investors make decisions and, in particular, at the committee structure typically employed. Committees represent a fruitful field for students of human nature and of sub-optimal decision making.

We then consider the way in which asset allocation policy is set. The use of sophisticated asset and liability models, optimizers and advanced scenario analysis means that asset allocation policy is already largely an analytic output, not a behavioral one. There are many lessons yet to be applied, however.

The hiring, monitoring and firing of investment managers is our third area of discussion. This is frequently a frustrating field—hard to know if it’s working, hard to be confident that decisions are as good as they can be. It’s also an area of huge importance and one which takes up a significant proportion of investment committees’ time.

We have not looked at the application of behavioral findings at the investment manager level. That application has been explored by many others already and continues to be so. After all, it opens up many ways in which markets might be found to be inefficient and the discovery of inefficiencies in markets is a very highly rewarded pastime. So, while that is a very important area of application for these findings, it is extensively investigated elsewhere and we have regarded it as beyond our scope. We have concentrated instead on the relatively unexplored perspective of the institutional investor.

Human behavior and the operation of an investment committeeCommittees are a necessary part of institutional investing, but not always an effective one. However, groups do enjoy some advantages over individuals, and the right environment can maximize the benefit of those strengths. Similarly, steps can be taken to address the challenges of group decision making. The role of the group leader is critical, but every committee member can help to create the right environment for an effective committee.


NABC: NOT ANOTHER BLOODY COMMITTEE“A committee”, opined Sir Barnett Cocks, the British political writer, “is a cul-de-sac down which ideas are lured and then quietly strangled.”

Here on the other side of the Atlantic, George Will is equally unkind. “Football”, he said, “combines the two worst aspects of American life: violence and committee meetings.”

Committees are easy to criticize, frequently with good reason. Committees often try to do too much and often do it ineffectively. They can take on a life of their own and become an end, not a means. The whole can easily become less than the sum of the parts, defeating the very point of creating the committee in the first place.

In part, the problem stems from the same source as so many of the other issues highlighted in this paper: our Stone Age brain. We are not designed to make decisions by committee.

A committee provides an inefficient replication of the individual’s decision-making process, exaggerating the biases built in to the heuristics we use. Our attention is diverted to personal agendas, building informal alliances, gaining status by making widely-accepted points and so on. The final decision then becomes a side-effect of these competing concerns, bearing little resemblance to the optimization that, in economic theory at least, this process is supposed to be acting “as if” it is replicating.

Institutional investment, however, demands that committees not individuals are used for many roles. There are a broad range of interests that must be represented and the responsibility involved is so large that it ought to be shared. Committees provide greater security and stability than individuals can.

The approach that we must take is therefore one of asking: how can we take maximum advantage of the strengths of a committee? What can we do about the challenges presented by the (unnatural) group setting?

THE VALUE OF POOLING INSIGHTSStarting with the strengths, a committee can bring a much wider range of perspectives than any individual can. To maximize the value of this, we should seek diversity in the make-up of the committee.

Diversity in this context is easily misunderstood. Diversity is not primarily about the racial or gender mix of a committee. Desirable as this type of broad mix is, it does not automatically lead to a diverse group in the sense we are talking about here. What we are seeking is as wide a set of perspectives as possible.

For example, trivia nights (or “pub quizzes”) are popular around the world. These involve teams (typically of four or five) being presented with a list of questions under various headings. A good team consists of individuals with complementary strengths. It does not really matter how many questions a potential team member can answer; what matters is how many questions they can answer that the other team members cannot.

Similarly, a good committee is made up of individuals each of whom can add something that the other members cannot.

TAKING ADVANTAGE OF VARIED PERSPECTIVESEven when different perspectives are present, they may well not be expressed or exploited. Indeed, the committee setting can subtly discourage that sort of thing. Taking advantage of differing points of view demands skill and sensitivity from the group leader as well as a willingness to co-operate from the group members.


First of all, group members must feel free to express their own perspective. An individual’s most valuable contribution can be the preliminary view formed before hearing others’ opinions. This is the view that is the most likely to be complementary to the views of others. Typically, we would hesitate to share these preliminary views—we like to be able to reassure ourselves that they are reasonable before putting them forward.

For important decisions, it can therefore make sense to have group members give independent consideration to the question before coming together. They should write down their initial reactions. The purpose here is not for them to reach firm conclusions about the final decision, but rather to identify what they believe to be the important arguments and pieces of information to be considered.

Secondly, potential disagreements should be brought out and explored, not suppressed. Group members should feel able to express doubts. A clear conclusion of the behavioral findings is that better results are achieved when decision-makers stop to consider why a proposed course of action might be wrong. Fully testing important decisions is a good thing to do. Often, that will mean that doubts are raised, answered and dismissed. The one raising the doubt in that situation should not feel discouraged.

This type of open debate, however, can turn into division, entrenched positions and factions. Part of the problem is that, once we have made a point, it becomes our point; we hear others refer to it as “Don’s point” or “Bob’s point” and it becomes personal. Our emotions become involved. It’s not an idea that is being discussed (or ignored) any more, it’s us. This tendency needs to be overcome. Bob’s point may turn out to be misguided, and it can be shot down freely by anyone who feels that way—but the suggestion that Bob is an idiot has no place in an effective committee environment. Likewise, defensive reactions to criticism of an idea (as if it were a personal attack) should be discouraged; Don should not act as if he has been called an idiot when all that has happened is that his idea has been questioned.

Leaders can lead by example on this, but everyone has a role to play in fostering the mutual respect in which effective cooperation thrives.

So committees should look at important questions from as wide a range of perspectives as possible and give room to consider doubts and identify potential weaknesses. It is good to exploit the benefits of diversity and foster mutual respect.

GETTING TO DECISIONSOf course, decisions still need to be reached. We’ve criticized the amygdala, but we have to concede that at least it drives us to action. The pre-frontal cortex, if it is to justify our rallying cry, must be able to do likewise. For that to happen, committees must be able to make decisions even without complete consensus. Votes should be taken, if necessary. This sounds obvious, but some institutional trustee boards and investment committees are so keen to avoid apparent disharmony that they will defer decisions when there is not unanimous agreement. Leaders must make sure that there are mechanisms for resolving disputed points, reaching decisions and moving on.

Indeed, the role of the group leader is central to achieving the effective committee environment for which we are aiming. A good leader can:

• Encourage the positive aspects of the group setting (drawing on the wide range of perspectives, letting each member act as a check and balance);

• Encourage committee members to think independently;

• Encourage genuine doubts to be expressed and considered;

• Frame issues appropriately with clear objectives;


• Be aware of the danger of groupthink;

• Ensure all committee members take ownership of decisions;

• Keep the focus on decisions that have most effect (i.e. either the ones that are biggest, or the ones which can be made most easily) and ensure that each decision is given appropriate consideration.

This task is made easier if each committee member recognizes these goals and sees his or her own role as improving the quality of the group decisions, not of competing with others or advancing a personal agenda.

One caveat we ought to add concerns the use of time. Is it really realistic to take the steps suggested above in a world in which time is already short, agendas are crowded and meetings frequently run over time?

We do not have a magic answer to that problem but would note that taking the time to test ideas and to address concerns is not the same thing as allowing tangents to be pursued or personal hobby horses to be aired repeatedly. Further, the approach to any decision should reflect the importance of that decision. Short cuts are fine if they are being taken consciously. Much of a good group leader’s work is done before a meeting takes place, structuring the agenda, creating the right focus and identifying key decision points. This makes it easier to keep a discussion on track in the meeting itself. Further, more time would be available if more decisions and activities were delegated.

SIGNS OF A COMMITTEE FUNCTIONING WELLSuccess is generally made easier if there is a clear picture of what it looks like. The list below provides one such picture of the signs that a committee is functioning as it should:

• Purpose of the committee is clear to all.

• Careful time control: length of meetings, as well as development of overall committee time path.

• Sensitivity to each other’s needs; good communication among all members.

• An informal relaxed atmosphere.

• Good preparation on the part of the chair and members.

• Interested, committed members.

• Minutes are complete and concise.

• Periodic self-assessment of the committee’s performance.

• Recognition and appreciation given to members so they feel they are really making a contribution.

• The work of the committee is accepted and makes a valuable contribution to the organization.5

Committees do indeed present challenges. But with a clear picture of the goals of the group and the right attitude on the part of all participants, they can offer a more effective—and less painful—experience than many do today.

5 Humphries, S. (1995) Effective Committees. Ontario Ministry of Agriculture and Food factsheet.


Human behavior and the asset allocation decision

Like any decision, strategic asset allocation rests on the establishment of the right goals at the start of the decision-making process. For an institutional investor, goals are complex—there are many facets of risk to be considered, and there is no easy way of deciding how much of each type of risk is acceptable. It is in the understanding of risk and investment goals that the greatest scope for improvement lies in the asset allocation process.

ASSIST THE PRE-FRONTAL CORTEX The establishment of strategic asset allocation policy is an area of investment where emotion tends to play a limited role and in which institutional investors tend to apply rigorous quantitative analysis. The statistical approach that is typically followed does have limitations, however, and if “resist the amygdala” is less important in this field than elsewhere, there is still a need to “assist the pre-frontal cortex”.

Some of the limitations in traditional asset allocation processes are:

• The models of asset class behavior on which this approach is based are only models, not reality.

• There is structural uncertainty: we cannot know the true probability of a stock market return in excess of 10% over the next twelve months, we can only estimate it.

• Investors’ true objectives are complex and hard to capture, the true nature of risk even more so.

• A close look at the output of an optimizer can lead to disillusionment: every diversified strategy seems to end up close to the efficient frontier and charts plotting the range of outcomes frequently do little more than establish that it is difficult to distinguish one strategy under consideration from another.

Allow for estimation error and the whole process might appear to have told us little.

This statistical analysis is not, however, an end in itself, and these limitations become less important when the object of the analysis is kept in mind. The analysis is a tool to aid the understanding of the impact on the investor’s objectives of the uncertainty inherent in capital markets, interest rates and so on. This impact can be difficult to grasp by intuition alone, and well-designed models can help to quantify the likely consequences of various choices.

GOAL-SETTINGThe decision process begins with setting goals. We have seen earlier in this paper that these are often inadequately considered. This observation seems well-founded in the case of many asset allocation decisions. Indeed, it seems to us justifiable to spend the majority of the time devoted to an asset allocation study on this first stage.

At a superficial level, asset allocation goals are easy to set: maximize expected return subject to acceptable risk. In reality, the quality of the final decision will rest largely on how well “risk” is defined, and how confidently the investor can decide what is “acceptable.”

The goals of an institutional investor are complex, and the goals can conflict with each other. There are many stakeholders to be considered. For example, in the case of a corporate defined benefit pension plan, the plan sponsor has certain objectives which may differ in some regards from the objectives driving the plan fiduciaries. Different cohorts of beneficiaries will have different goals: for older members of the plan, security of principal is the dominant goal (to ensure benefits can continue to be paid) while for younger members, capital appreciation is necessary to ensure that the plan continues to be viable and that


future benefits can continue to accrue. Another source of conflict is the fact that there are several accounting and reporting conventions and requirements to confuse matters.

In addition, investors all of types are influenced by their peers. Even though it is many years since it was generally considered acceptable to set a goal of “outperform the peer group”, most committees find it hard to reach decisions that deviate too far from the practice of others. In part, this instinct is reasonable: those who pay no regard to peers may be guilty of overconfidence. One of the tests of prudence that courts apply in practice is a comparison of how others in similar circumstances behaved. The onus is on those acting differently to prove that their actions are justified. Those who stick with the herd will rarely if ever be blamed for doing so.

Peer groups are rarely part of the formal goal-setting process and we are not arguing that they should once again become a major factor in asset allocation decisions. Where material differences from peers do arise, however, it is worth the extra effort of identifying why you have reached different conclusions, remembering our aim of always asking “why might we be wrong?”

These issues should not be regarded as complications, getting in the way of the “pure” asset allocation decision you’d like to make based on a single measure of success. More time should be spent on understanding the multiple goals of asset allocation policy, risk and what constitutes success and failure. That is the heart of the asset allocation decision.

RISK TOLERANCEHaving considered what risk is, an investor must then decide how much of it to take.

Defining risk tolerance has always been one of the least rigorous aspects of the strategic asset allocation process. Decades ago, in the early days of mean-variance analysis, risk tolerance was a simple number that was plugged into Markowitz’s formula. So if your risk tolerance was 100 (which was quite high), you were prepared to accept up to 100 units of additional variance for an additional unit of expected return. Depending on the exact asset class assumptions used, this might lead to an equity allocation of 90% or more. If your risk tolerance was 10 (quite low), you were less willing to take on additional variance6 in return for a higher expected return: you would accept only 10 units of additional variance for the same additional unit of expected return. The resulting equity allocation might be 30% or less. Most investors, strangely enough, turned out to have risk tolerances of something like 55 or 60. Most investors therefore ended up with equity allocations around 60%.

Of course, what was really happening most of the time was that investors were choosing the risk tolerance that produced the asset allocation that they wanted to adopt. The question of “how much additional variance are you prepared to take on for an additional unit of expected return?” is not a particularly intuitive question. Very few people thought in those terms until the question was asked.

The best asset allocation processes today avoid making risk tolerance an abstract input. It is a concept underlying much of the modeling, below the surface. But the investor’s risk tolerance is explored in more meaningful terms: by looking at the impact on funded status, return volatility and so on, of the various choices open to them. Indeed, it is clear today that risk has several dimensions, that these cannot simply be aggregated into one number and that the investor almost certainly tolerates some forms of risk more easily than others.

So whereas risk tolerance used to be expressed in terms that seem very well-suited to the pre-frontal cortex (“55”) and today it is not, today’s models are in fact allowing investors to be far more rigorous in this critical area.

6 Even though the natural trade-off to use would be standard deviation of returns versus expected return (because each is measured in the same unit: percent per annum), the mathematics are more tractable if variance (the square of standard deviation) rather than standard deviation is used, and the end result of the optimization process is the same.


ASSESSING POTENTIAL NEW INVESTMENT STRATEGIESOne area of asset allocation that is particularly prone to behavioral influences is the treatment of new strategies. From time to time, new opportunities for investment become feasible that were not previously available or easy to implement. These need to be considered, because one of the goals of a good investment program is to achieve diversification. Some old opportunities (like private equity) fall into this camp, too, because they tend to fall in and out of fashion.

The first (and most obvious) danger to be aware of is the strong tendency of investors to get in at the top of strategies. The problem is that the strategies that get considered are the ones that have performed well in the recent past. These are the strategies on which new products will be launched, new marketing efforts devised, and most new interest from investors will be focused. Often, these are sound strategies. But by the time they get onto the radars of most investors, it is too late. If the only time that you ever consider an allocation to private equity is after one of the years of 40%+ returns that occur every now and again, you are stacking the odds of long-term success against yourself.

It is also easy to judge new strategies by representativeness (“that is like portfolio insurance, therefore it is bad” or “that is like equities in the 1950s therefore it is good”). Or to ignore the fact that a number of conditions all need to hold true for a new strategy to make sense (for example: past data should not be misleading us; relationships that are expected to exist really must hold, and must continue to hold; the return advantage to the strategy should not have been not arbitraged away; there must be a cost-effective means of implementing the strategy, etc.).

INCONSISTENCIESSome investment policies contain inconsistencies. This is not a disaster; it is almost inevitable to some extent given the complex decision-making structure that exists at most institutions and the piecemeal way in which some aspects of the asset allocation decision are made. However, when inconsistencies are identified, they ought to be looked into.

For example, some investors have risk guidelines that restrict fixed income investments to investment grade issues (BBB or better). As a result, a fund may hold in its small cap portfolio the equity of a large number of companies whose debt its fixed income managers are not permitted to hold. It does not seem to make sense to hold the equity of a company but NOT allow its debt which, by its very nature, is a less risky investment.

Similarly, the treatment of active currency risk is sometimes inconsistent between portfolio guidelines for international equity and fixed income mandates. This can be the result of a view that a given currency risk is big in the context of a fixed income portfolio but not so big in the context of an equity portfolio, which contains many other, larger risks. But the impact of the given currency risk on the total portfolio is the same no matter whether it is taken in a fixed income portfolio or an equity portfolio. Unless there is reason to believe that the equity manager is better equipped to manage the risk than the fixed income manager, active currency risk should probably be treated the same wherever it sits.

Consistency should not be made a goal in itself, but inconsistencies such as these do serve as useful symptoms of room for improvement in investment policies.

SummaryWe know that the most frequent failings in decision making in general are inadequate definition of the right goals at the outset, and failure to tie the final decision back to those goals. In the case of asset allocation, these tendencies manifest themselves through inadequate treatment and over-simplification of what risk really is. In a world which is becoming ever more complex for institutional investors, the right asset allocation policy begins with a clear understanding of what it is trying to achieve.


Human behavior and the selection and monitoring of money managersManager selection and monitoring is one of the most difficult aspects of institutional investment. We argue that the primary reason it is so difficult is the lack of the right type of information. A lot of information exists but none of it tells us much about the manager’s future performance. However, human beings make decisions under conditions of uncertainty every day, so this is in fact a familiar challenge. It turns out that manager selection is a skill requiring a combination of natural aptitude, training and experience.

THE CHALLENGE OF SELECTING GOOD MONEY MANAGERSFridays show up with impressive and predictable regularity. Yet people seem frequently to be surprised by them. “Sorry I’m late—I got caught in the Friday traffic” can sound quite plausible. But there is not really any excuse for being surprised by Fridays and neither is there any excuse for being surprised by any occurrence that should have been anticipated in advance.

Manager underperformance is such an occurrence. Unfortunately, we do not generally know when a manager will hit a tough spell, but we know it will come eventually. The performance of all money managers varies over time. Whether compared to market indexes, to their peers or to any other benchmark, there will be ups and downs.

Selecting active money managers is difficult, no matter how you look at it. In 1999, one of the authors polled a number of leading UK pensions professionals about (a) the “importance to excellence” of fifteen factors involved in running a pension plan and (b) the “excellence rating” of their own organization with regard to those factors.7 The results showed that in most cases, the most important aspects were also those that were done best. There were two exceptions to this general observation, however. “Timely decisions” was rated as of above-average importance, but was rated one of the worst in terms of actual execution. And “selecting good investment managers” was rated as the single most important factor of all—yet participants gave it an “excellence rating” slightly below average.

The selection and monitoring of money managers is a very important and high profile activity, but results are hard to interpret because it is so difficult to separate skill from chance. When things are going smoothly, we tend to relax and assume we are doing a great job. When things start to unwind, we do not know what to do. One paper on the illusion of control captured this in its title “Heads I win, tails it’s chance.”8 Should we replace the underperformers and risk acting on the basis of noise, incurring costs and maybe ending up worse than where we started? Or should we try to ride it out, and risk finding that the underperformance persists, making us look weak for failing to act when things started to go wrong?

This is a difficult decision, but—like Friday—it is coming. It does not need to feel like a crisis. Dealing with underperformance, sometimes through decisive action, sometimes through restraint and patience, is part of the job.

ANALYZING THE DECISIONIt is informative to apply the six-stage decision-making framework to this area. The first step is setting goals, and this is fairly easy in this case: We want enough value added from the positions taken by the managers to recoup their fees and the trading costs incurred. In practice, there are also considerations of risk management, portfolio structure, complementarity and so on. But these are less important than finding the value added in the first place.

7 Ezra, D. D. (1999). Excellence in UK Pension Fund Management: A Perspective on Current Practices, Russell London Monograph.

8 Langer, E. & Roth, J. (1975). Heads I win, tails it’s chance: The illusion of control as a function of the sequence of outcomes in a purely chance task. Journal of Personality and Social Psychology, pp. 32, 951–955.


The second step is to gather information, and the third step is to evaluate that information. It is here that we encounter our main problem. Numerical data on the nature, style and past performance of money managers is widely available, but none of it tells us much about the manager’s future performance. “Past results are no guide to future performance” is not some empty small-print, it’s the truth. When you are buying a car, if you want to know how much luggage it will be able to carry, how fast it will go, or any other measure of performance, you can get it. When you are selecting a money manager, you want to know what the future performance will be, but you don’t know. You don’t even have a good proxy. In most decisions we make in life, past performance gives us a reasonable starting point in assessing future performance. How likely are the Toronto Blue Jays to beat the Boston Red Sox in a single game? Is Bill Smith likely to be a better line manager than Suzy Doe? These questions can usually be answered with some confidence if we have enough information about the past. In money management, the link between past performance and future performance is much weaker than what we are used to in almost any other area of life. This is partly because of chance and partly because of change.

Chance and change mean that any performance history long enough to be a robust measure of skill (as opposed to short-term “noise”) is probably no longer relevant to the firm and the markets as they are today.

Are there other proxies for future performance that are more reliable? It seems unlikely. At Russell, we have for many years looked closely for links between particular measures and performance. We have found that certain characteristics, for example clarity of objectives and stability, do tend to go with good performance. But that does not mean that manager selection can be reduced to a set of rules based on these characteristics.

The remaining three steps of the decision-making process are to ascribe utilities, trade off between goals and, finally, to make the decision. These would not be particularly challenging in the case of money manager selection and monitoring, if reliable information were available about likely future performance.

So a close look at the decision-making process simply confirms what most readers will already have suspected: information on managers abounds, but it is largely irrelevant, constantly changing and often misleading. It is this lack of relevant information that makes manager selection and monitoring so difficult.

THIS IS BY NO MEANS A UNIQUE PROBLEMThis is, in fact, familiar territory. In Section I, we reproduced a table generated by Gary Klein, which contrasted the circumstances in which experience-based (heuristic) decision making was dominant and the circumstances in which rational (analytical) decisions making was dominant.

The point here is that some decisions are better made by experienced individuals using judgment than by following detailed analytical procedures. Many investment decisions are essentially analytical decisions, but manager selection, it turns out, is not. Because of limited information and dynamic conditions, manager selection should not be regarded as a quantitative, rules-based process. Manager selection is a skill, like many other skills. Like most skills, it comes down to a mixture of natural aptitude, training and experience. Rigorous analysis can inform and support the process, but not replace it.

So selecting active money managers is difficult, but it is not impossible. Human beings, in fact, are well accustomed to making decisions when the information we are working with is incomplete or unreliable, and when conditions are dynamic. Our Stone Age brain can handle that. Making good decisions in those circumstances demands something different than making good decisions in an analytical environment. For a start, investors should not


9 Klein (1999) Op. cit.

aim for perfect decisions, but rather for good ones. Some decisions will turn out to be wrong: that’s OK. (Another case of Friday. Even the most skilled experts in every field sometimes make mistakes.) They should be wary of overconfidence: it is very human to believe, with little justification, that we possess skills innately when we do not.

Manager selection is not a skill that is easy to acquire. For a start, there are hundreds of investment firms to choose from, each of whom has professional sales staff and a good story to tell. Most investors make these decisions only occasionally. Feedback on whether decisions were good decisions or bad decisions is slow to arrive and frequently ambiguous (because luck means that good decisions can sometimes lead to bad outcomes and bad decisions to good outcomes). Investment is a complex subject. And, for the majority of investors, decisions are closely watched by others, creating a need to justify the choices that are made. All of these circumstances make the selection and monitoring of money managers a challenge.

To improve decision skills, a number of steps are possible. Klein lists, for example, the following observations about how experts actually learn:

• They engage in deliberate practice, so that each opportunity for practice has a goal and evaluation criteria.

• They compile an extensive experience bank.

• They obtain feedback that is accurate, diagnostic, and reasonably timely.

• They enrich their experiences by reviewing prior experiences to derive new insights and lessons from mistakes.9

ALTERNATIVE RESPONSES TO THE CHALLENGEThe steady growth of passive management can be seen as an alternative response to these challenges: if meeting the goal of generating return through active management is too high an aspiration, then saving on cost by employing low-cost passive managers is a much more achievable objective. Another rational response is to outsource the selection decision to specialists. There has been considerable growth in recent years in the outsourcing of manager selection through the use of multi-manager investment products. In the case of hedge funds, manager selection and the construction of well-diversified portfolios are so difficult that few institutions select hedge funds directly and Hedge Fund of Funds are the preferred investment vehicle for most.

SummaryWhile there is no place for emotion or panic in manager selection and monitoring, neither can it be reduced to a quantitative, rules-based process.

/ p 69

Can a defined benefit plan provide a competitive edge?

Even though a defined benefit (DB) pension plan is a trust that is held separately from the corporation that sponsors it, the way that it is funded means that, for the purposes of corporate accounting and planning, it can be regarded as a subsidiary operation. As such, a DB plan represents both a source of risk and a source of potential competitive edge for a corporation. A great deal of work has been carried out over the past 30 years on the economics of pension policy. The paper that follows was published in 2008 and places the pension plan into that corporate context.

Russell Investments // Risk management: A fiduciary’s guidebook // Can a DB plan provide a competitive edge?

/ p 70

Can a defined benefit plan provide a competitive edge? When a sponsor creates a defined benefit (DB) pension plan, it creates a financial operation that affects its profitability. This paper explores three aspects of that operation: How can one measure the size of the operation? In what ways can it be made a source of competitive financial advantage to a sponsoring company? And How would traditional corporate financial metrics be affected if those ways are pursued?

Periodically the world seems to revive its interest in looking at a DB plan in the context of an operation that affects its sponsor’s financial position. Later in the paper we cite the earliest references to this context, dating back to 1980. The “perfect storm” of 2000–2002, when falling equity prices reduced asset values at the same time that falling interest rates increased liability values, caused the notion to take on a reality it previously lacked. And the 2006 mark-to-market developments in U.S. legislation (the Pension Protection Act) and accounting regulations (Financial Accounting Standard 158) have kept the notion in the forefront of corporate thinking.

Our angle here is simply whether the DB plan, considered purely as a financial operation, can be made into a source of competitive advantage, to improve the sponsor’s earnings and share price in a way that competitors can’t match. We suspect that in most cases the answer will be “No—at any rate, not by much.” But in some cases the answer may be

“Yes,” and in all cases the discussion itself is instructive and can lead to some increased financial efficiency.

DB as a financial operationA DB sponsor underwrites benefits promised to its employees. It creates a sinking fund to pay off those promises—commonly called the pension trust. It contributes capital from time to time to that sinking fund—commonly called contributions. It invests the sinking fund. Over time the investment returns become even more significant than the capital contributions; by the time the promises are redeemed, much more than half the redemption comes from investment returns.1

Since investment returns and capital contributions together eventually pay off the benefits, the size of the investment returns affects the amount of capital that must eventually be contributed. In turn, that affects the amount of capital available to the sponsor’s mainline operations, or the terms on which that capital can be raised in financial markets. There follows, necessarily, an impact on the sponsor’s profitability and worth.

By: D. Don Ezra and George Oberhofer

February 2008

Russell Investments // Risk management: A fiduciary’s guidebook // Can a DB plan provide a competitive edge?

A competitive edge arises when a company is able to do something better than its competitors. Doing that something involves taking a risk. The DB operation is no different.

1 Keith P. Ambachtsheer and D. D. Don Ezra, Pension Fund Excellence (John Wiley & Sons, 1998).

/ p 71Russell Investments // Risk management: A fiduciary’s guidebook // Can a DB plan provide a competitive edge?

In other words, the pension fund is effectively a financial operation of the sponsor.2

In the USA, the fiduciaries of the pension trust cannot, under the Employee Benefit Security Act (ERISA), accept investment direction from the sponsor; they are charged with deciding investment policy in the sole interest of the beneficiaries.3 But ERISA’s regulations4 also point the fiduciaries in the direction of discussing with the sponsor the funding—that is, capital contribution—policy desired by the sponsor. Almost invariably, the sponsor’s funding policy anticipates—indeed, requires—an investment return higher than that likely to arise from a risk-minimizing approach. Otherwise the sponsor would not agree to support benefits at the level promised. That can place the fiduciaries in a difficult position.

This paper ignores that fiduciary issue. It considers only the perspective of the sponsor. Of course, that means in practice that the conclusions that benefit the sponsor most may not be capable of being carried out by the fiduciaries. But it does at least give clarity to the sponsor’s perspective, and therefore becomes a useful input when sponsor and fiduciaries have their discussions.

This paper also ignores the value of defined benefits in giving the sponsor a competitive edge in attracting and retaining staff. That used to be a very significant factor in the early days of DB, when a DB offering stood out competitively. DB lost its edge when it became a routine offering among large sponsors. As DB declines, it may regain that edge. But that aspect does not concern this paper. This paper focuses solely on the corporate financial aspects of a DB plan already in place.

How significant is the size of the DB operation? Hall (2007) cites the availability of many measures of the size of a DB plan and of its significance to the sponsor, and shows the ranges of some of the particularly interesting ones.5 He also demonstrates that, contrary to the headlines one often sees, in the U.S. there are relatively few large corporate sponsors for which the DB plans are of significant financial importance.

In practice we have found three measures nicely sum up the financial significance of the DB operation.

The first looks at the amount of capital required in the sinking fund (in other words, the size of the accrued pension liability relative to corporate equity). What the Financial Accounting Standards Board used to call the “accumulated benefit obligation,” based on service and salaries to date, and valued consistently with a bond yield curve, is the appropriate measure.6 The “projected benefit obligation” (which includes the impact of future salary projections) is typically the (imperfect but more frequently available) substitute.

Exhibit 1 shows this ratio for those US Russell 1000® Index companies that have defined benefit plans.

2 See Black (1980) and Tepper (1981) for the earliest examples and Hall (2007) for a more recent reference. 3 Employee Retirement Income Security Act [ERISA], Section 404(a)(1).4 Department of Labor, Code of Federal Regulations, Title 29, Chapter XXV, Part 2550 (29 C.F.R. ¤ 2550.404a-1(b)(2)).5 Hall, Michael (2007). “Financial Economics and Corporate Strategy: How Do Pension Plans Fit within the Broader

Context of the Corporations that Sponsor Them?” 6 This is also the “funding target” under the Pension Protection Act of 2006.


Exhibit 1: Ratio of PBO to Market Capitalization, Russell 1000 Companies

Percentile point of range 5% 25% 50% 75% 95%

Ratio of PBO to market cap 1% 3% 8% 20% 63%

Source: Hall, op. cit., data as of fiscal year end 2006—562 Companies

For the 562 large companies with DB plans, the median is a company for which its DB projected benefit obligations amount to 8% of its market capitalization—so the DB plan isn’t a significant operating division of the company.

But at the high end the ratio is very large. Indeed, the figure for the 95th percentile means that, for 28 of these companies, the pension fund may be the tail wagging the dog. That’s why some companies have been described as staff welfare plans that sell product or service X in order to finance themselves. We have heard of one newly-hired CFO who said, surprised, “I didn’t expect to find that running the pension fund would be my most significant role.” At that end of the spectrum, that CFO has got it right.

“Business as usual” means running the fund as a sideline, with no corporate purpose. Far better to define a conscious mission statement for the fund—a topic we address in the next section, on competitive advantage.

We have found that for many companies, a large sample such as the Russell 1000 group is actually far less valuable than the much smaller group constituting the sponsor’s main competitors. With this smaller group, one has a more precise focus on whether or not one wants one’s fund to adopt a particular role, and whether or not the fund is large enough to actually be a source of competitive advantage. The larger the ratio in Exhibit 1, the larger the impact of the pension fund can be. Indeed, the larger the ratio, the greater the importance of giving the pension fund a conscious mission statement.

The second useful measure is the ratio of the plan’s unfunded liability to its market cap. This measures the extent to which a portion of corporate equity is, in effect, earmarked for contribution to the sinking fund, even if no further liabilities accrue. Exhibit 2 shows this ratio for the Russell 1000 companies.

Exhibit 2: Ratio of Unfunded PBO to Market Capitalization, Russell 1000 Companies


Ratio of unfunded PBO to market cap –3% 0% 1% 2% 8%

Source: Hall, op. cit., data as of fiscal year end 2006—555 companies

Again, the larger this ratio, the more important it is to address the mission of the pension fund. And again, looking at one’s competitors is more interesting than looking at all the Russell 1000 companies.

These first two ratios look at the past: the accrued liabilities and how much is as yet unfunded. The third useful ratio looks at the future. Assuming that future capital transfers will preferably be made from operating profits rather than from newly raised capital, it becomes obvious to look at the ratio of each future year’s normal accrual of new benefits to some measure such as EBITDA (earnings before interest, tax, depreciation and amortization). Exhibit 3 shows this ratio for the Russell 1000 companies.


Exhibit 3: Ratio of Defined Benefit Service Cost to EBITDA, Russell 1000 Companies


Ratio of defined benefit service cost to EBITDA -3% 0% 1% 2% 8%

Source: The same data as underlies Hall, op. cit., data as of fiscal year end 2006—550 Companies

Again, the direct competitive picture is typically even more useful than this large sample.

A DB mission statement: What sort of competitive edge, from a corporate finance perspective? A competitive edge arises when a company is able to do something better than its competitors. Doing that something involves taking a risk. The DB operation is no different. It therefore has two possible ways of generating a competitive edge. One is to earn a higher return than the competitors do. The other is to expose the company to less risk than the competitors take.

Either of those aspects can generate a mission statement for the fund.

“Our DB fund will earn a higher return than our competitors, enabling us to EITHER offer higher benefits at the same cost, thus attracting and retaining a higher caliber of staff OR offer competitive benefits at a lower cost, thus enhancing corporate earnings.”

“Our DB fund will consciously take less risk than our competitors, thus EITHER making us a less risky company OR enhancing benefit security for our employees OR enabling us to take more risk in areas where we already have a competitive advantage.”

One factor in deciding whether either of these types of mission is appropriate is the size of the competitive edge that the fund can generate. If it turns out to be small, it may seem hardly worthwhile to pursue it. One possible approach in that case is: “We might at least use the pension fund to deny our competitors an edge.” This implies following the competitors’ strategy, so that, broadly, the fund will earn the same pattern of returns as the competitors.

Let’s now look at the investment implications of the three basic approaches outlined here.

How does a company earn a higher return than its competitors? In either or both of two ways: by adopting a more aggressive asset allocation, or by earning higher returns in each asset class through active management.

How does a company earn a higher return than its competitors? By adopting an asset allocation that more closely matches its projected liability cash flow. This attitude is one that lies behind much of the current talk about liability-driven investing.

How does a company follow its competitors’ strategy? By doing whatever they do, in terms of asset allocation, active management and risk management generally. Of course, if significant competitors do different things, the company has the problem of deciding which of them to mimic, or whether to mimic their average instead. And the same strategy could have markedly different effects on two sponsors with different corporate balance sheets, betas, and so on. So even this “follow the competitor” strategy needs thinking through.

It’s worth noting one natural consequence of adopting any of these three conscious approaches. Each approach has a natural sort of benchmark by which to evaluate progress. It may be the degree of aggression in the asset allocation, and its success; or the degree of matching the liabilities, and its success; or the degree of active management, and its success; or the degree of closeness to competitors’ strategy, and its success.


Only one of these measures (in connection with active management) relates to what everyone else is doing (which remains the traditional approach to judging success). Two of them (tracking, or being more aggressive than, competitors) relate to a very small group of others (the designated competitors). And one (reducing risk) relates not at all to competitors or others but exclusively to the company’s own benefit liabilities.

In general, then, adopting a conscious mission statement for a DB fund requires the sponsoring company to abandon the traditional performance reporting format and design a new report that measures progress relative to its new goals. Rather than “we’re above median,” the new mantra becomes “we’re on track.”

How would traditional corporate financial metrics be affected if those ways are pursued? In this section we’ll look at a set of four approaches in terms of their impact on the traditional corporate finance metrics of EPS, leverage and particularly beta. That’s because, if a company decides to reduce risk in the pension fund, that in turn affects the company’s beta. The company can respond in one of three ways: increase corporate leverage enough to preserve beta, accept the reduced beta, or take more risk in core operations enough to preserve beta. If a company decides to increase pension fund risk, a similar set of choices presents itself; of these choices, the one that is different in ultimate outcome is to accept the increased beta. Hence the following four approaches essentially cover the full range of outcomes:

A. Reduce fund risk by matching assets to liabilities, then preserve the company stock beta by increasing corporate leverage.

B. Reduce fund risk, and accept the reduced beta.

C. Reduce fund risk, then preserve the beta by taking more risk in core operations.

D. Increase fund risk, and accept the increased beta.

(A) REDUCE FUND RISK BY MATCHING ASSETS TO LIABILITIES, THEN PRESERVE THE COMPANY STOCK BETA BY INCREASING CORPORATE LEVERAGEOver the years, this approach is well documented in pension and corporate finance literature.7 But beware of the necessary imperfection in defining the cash flow pattern of the liabilities, which for corporate finance purposes is typically and conveniently, but inaccurately, assumed to be defined with precision and capable of being matched with precision.

The basic idea is that, in a regime where there is differential taxation between corporation and pension fund and between bonds and stocks, it pays to see where tax arbitrage applies. In the U.S., this approach would lead to the conclusion that it pays to hold as much fixed income as possible in the pension fund. Hence, sell stocks and buy bonds in the fund, and reverse the effect (after tax) in the corporation by issuing bonds and buying back stock to preserve company beta. This typically results in a net gain to the company’s cash flow (and hence, via P/E, to market cap).

For this to work, it is of course necessary that bond interest paid by the company is tax deductible, and that the company’s cost of debt is not far above the interest rate on the fixed income being purchased in the fund—conditions that may not apply in practice. This is similar to the familiar situation in which an individual pays down a mortgage by running up a credit card balance—not a good idea if the interest rate on the credit card balance noticeably exceeds the mortgage interest rate.

7 Black, op. cit.


A constant-beta strategy depends critically on shareholders being able to diversify idiosyncratic risk entirely away. Raising leverage not only restores beta to its original level but increases sponsor-specific risk. If for some reason, shareholders cannot completely diversify idiosyncratic risk, they will demand additional return to compensate them for taking it. Recent empirical studies suggest that this in fact may be the case.8 If so, a constant beta strategy may reduce shareholder wealth.

Lesson: it’s worth working through this exercise in practice rather than just assuming that it will automatically work.

(B) REDUCE FUND RISK, AND ACCEPT THE REDUCED BETAThis is also an approach well documented in corporate finance literature.9 The idea is that there is no need for a company to ramp up corporate risk by increasing its leverage, as approach (A) suggests. That should be a decision left to each shareholder. It is the shareholder’s risk budget that is important, not the company’s. If the shareholder wants to increase risk via leverage, he or she can do it personally, without the need for the company to do so. The company should stick to generating profits from areas in which it has a competitive advantage, since those are not conditions that a shareholder can find elsewhere.

It is worth working through some calculations, though. An example is shown in the Appendix. The example is useful, for two reasons. One is that it shows the mathematical relationship across earnings per share, share prices, leverage and beta (from a model, at any rate—real life is always less predictable). The other reason is that, in the example, it appears that the reduction in earnings per share (arising from a reduction in the fund’s expected return—a number built directly into the calculation of the accounting “expense” charged against earnings) is likely to be greater than the market would accept, given the likely reduction in beta.10 That is not an outcome the shareholder is likely to be happy with, regardless of his or her ability to increase personal leverage back to where it was before the company reduced its risk.

Again, the lesson is that markets and market conditions don’t always precisely match what theory says should happen, so it’s worth working through the numbers in one’s own situation.

(C) REDUCE FUND RISK, THEN PRESERVE THE BETA BY TAKING MORE RISK IN THE CORE OPERATIONSHere again, as in approach (A), the idea is that the company beta will remain unchanged. The expected return on the pension fund will fall, but it should be more than made up by the increased earnings from core operations as the pension risk budget is transferred to core operations.

The calculations involved in estimating the impact of this approach relate to the evaluation of the pension fund’s expected return and risk in the context of the hurdle rate and risk measures used by the company in evaluating projects in its core operations.

One might assume that for most companies its core projects ought to be more profitable than its pension fund. Yet this will not be an approach favored by all companies. For example, there are many mature companies that throw off cash flow from operations and prefer to use that cash flow to reduce debt or buy back equity. These companies are not seeking new risk capital and do not have a number of projects that fail to gain acceptance because of capital starvation. For them, ramping up risk in core operations is not desirable.

8 See Goyal and Santa-Clara (2002), and Malkiel and Yexiao (2000). 9 Tepper, op. cit. 10 Why this unexpected outcome? It may be, for example, that the company’s current situation isn’t consistent with its

historic beta, which is used in the calculations.


(D) INCREASE FUND RISK, AND ACCEPT THE INCREASED BETAAn increase in fund risk should increase the fund’s expected return. In turn, that should increase earnings per share as well as company beta.

Is the increase in earnings higher than the market would expect, given the increase in beta? Is ramping up pension fund risk the best way to use a company’s risk budget? These are the issues to be considered for this approach. Remember that Modigliani and Miller11 point out that beta is easy for shareholders to take on themselves, and that a company cannot increase its market value by doing something for shareholders that they could easily do themselves. Alpha-seeking is the only risk-taking that can add value for shareholders. Hence the particular skill sought here is alpha-seeking, not ramping up beta. (Of course, with some asset classes alpha is inseparable from beta.)

There is nothing new in the calculations involved in this approach.

Few funds, if any, seem to be pursuing this avenue. Most conclude that they have been taking too much risk or exactly the right amount of risk. But, at least in principle, there must be some funds that ought to consider increasing risk.

Related readingBlack, Fischer (1980). “The Tax Consequences of Long-Run Pension Policy.” Financial Analysts Journal, July-August.

Goyal, Amit and Pedro Santa-Clara (2002). “Idiosyncratic Risk Matters!” Discussion draft, Anderson Graduate School of Management, University of California at Los Angeles (March).

Hall, Michael (2007). “Financial Economics and Corporate Strategy: How Do Pension Plans Fit within the Broader Context of the Corporations that Sponsor Them?” Russell Viewpoint (October).

Malkiel, Burton G. and Yexiao Xu (2000). “Idiosyncratic Risk and Security Returns.” Discussion draft, University of Texas at Dallas (July 13).

Oberhofer, George (2003). “Boots PLC’s All-Bond Pension Portfolio: Will the World Follow?” Russell Viewpoint (March).

Tepper, Irwin (1981). “Taxation and Corporate Pension Funding Policy.” Journal of Finance (March).

11 Merton Miller and Franco Modigliani, The cost of capital, corporation finance and the theory of investment (American Economic Review, June 1958).


Appendix: A numerical example of the interaction of pension fund return expectations, company leverage, earnings per share, beta and share prices This is how one company approached the first-order calculations involved in the question of what would be equity market expectations if they were to defease (that is, exactly match) their DB liabilities. The exact numbers (which of course have been disguised to conceal the identity of the company) are not important; the mathematical approach is, for the purposes of this paper.

STEP 1: We must first construct the aggregate balance sheet of the company, adding its core operations and its pension fund together. We first show the outcome, and then state the assumptions used to bring it about.

Exhibit 4: Aggregate balance sheet, company operations plus pension fund

(All figures in millions of dollars)

Assets Liability

Operating 2600 Operating 600 Pension 250 Pension 250

Equity

Market Cap 2000

Total assets 2850 Total liabilities and equity 2850

We make three critical assumptions about the pension assets and liability in Exhibit 4. The first is that the plan is fully funded. The second is that plan assets not invested in equity effectively defease the pension liability, so that they are excluded from Exhibit 4 along with the portion of the liability they defease. The third assumption is that a dollar invested in the plan is the equivalent of $1 x (1 – tax rate) = $1 x (1 – 0.375) = $0.625 in operating assets.

The second and third assumptions may become clearer with more background numbers. The plan’s assets are $600 million, as are its liabilities. Of those assets, equities are $400 million and bonds $200 million. We offset the bonds against $200 million of the liabilities (they are effectively matched in length), and so, for this case study, we consider only the $400 million of equity assets and $400 million in (not yet defeased) liabilities. That’s the effect of the second assumption above.

The impact on the company of $400 million in the plan is the same as the impact of $250 million on the corporate balance sheet, because each $1 contributed by the company to the plan (resulting in $1 added to the plan’s assets) only costs the company $0.625 in earnings, since the $1 is tax-deductible to the company. Hence, for the aggregate balance sheet, plan assets and (not yet defeased) liabilities are shown at $400 x 0.625 = $250 million. That’s the effect of the third assumption above.

We took the market value of the operating liabilities to be the same as their book value. And the equity market cap = $20 (share price) times 100 million shares outstanding.

STEP 2: Suppose we were to move to an all-bond portfolio in the pension fund, to reduce risk (in the pension fund and therefore also in the company’s overall scope of activities). We can calculate the immediate effect on earnings per share (EPS).


We assume an “expected long-term return on assets” (ELTRA) of 9% p.a. for our allocation that includes approximately 65% in equities. This ELTRA is based on annual return assumptions of 5.5% for fixed income and 11% for equities, an equity premium of 5.5%. So, if we were to move to 100% fixed income, we would lose an expected 5.5% return on our current $400 million of equities, or $22 million.

This means an increase in pension expense of $22 million before tax. At a tax rate of 37.5%, this translates into a reduction in earnings of $13.75 million. With 100 million shares outstanding, EPS would fall by 13.75 cents per share. With current EPS at $1.50, this is a 9% reduction.

The company’s beta would also fall. The crucial question becomes: how much EPS reduction would the market accept in exchange for the reduction in beta?

STEP 3: The fact is that the company’s historical beta (regressed against the Russell 3000® Index over the past five years) has been 1.20. The current aggregate balance sheet, of company operations plus pension fund, is as shown in Table 4, below.

Assuming a beta of zero for the company’s debt: Beta unlevered = beta levered/ [1+ (1 – tax rate) x debt/equity] = 1.20 / [1+0.625*850/2000] = 0.95.

What would be the implicit beta of our operating assets only?

Beta of operating assets = [unlevered beta – pension assets/total assets] x [total assets/operating assets] = [0.95 – (250/2850)] x [2850 /2600] = 0.95.

So, finally, what would be the implicit beta of these operating assets, with our operating leverage built back in?

Beta levered = beta of operating assets x [1+ (1 – tax rate) x (debt/equity)] = 0.95 x [1+0.625*600/2000] = 1.13.

STEP 4: What reduction in EPS would the market accept if we reduced our beta from 1.20 to 1.13?

Let’s assume that systematic equity risk is estimated by the market at 18% per annum.

With a risk-free rate today of 5.25% and our expected long-term equity assumption of 11%, this implies that the capital market line has a slope of [.11 – .0525] / .18 = 0.319.

The systematic risk associated with our current beta of 1.20 is 1.20 x .18 = 21.6%. The systematic risk associated with our reduced beta of 1.13 is 1.13 x .18 = 20.3%.

Thus the reduction in systematic risk arising from an all-bonds pension fund portfolio is 21.6% – 20.3% = 1.3%.

Given the slope of the capital market line at 0.319, the reduction in required return accepted by the market in compensation for the reduction in beta is 0.319 x 1.3 = 0.4%.

With EPS at $1.50 and a share worth $20, EPS per share is 7.5%. A reduction of 0.4% cuts this to 7.1%.

In percentage terms this is a reduction in EPS of [7.5 – 7.1] / 7.5 = 5%.

STEP 5: So, how would the market react?

From Step 2, we would reduce reported EPS by 9%.

From Step 4, the market would accept a reduction of 5% as compensation for reduced beta.

CONCLUSION: The market would react unfavorably.

/ p 79

Risk Management

Plus ça change, the French are fond of saying, plus c’est la même chose: the more things change, the more they stay the same. Our next paper predates the bursting of the tech bubble and the bear market of 2000–2002, it predates the Pension Protection Act, it predates the credit crisis of 2008 and the resulting global financial crisis. In 1999, when this paper was written, annual contract volume on the Chicago Board Options Exchange (CBOE) was 22% of what it is today. The IntercontinentalExchange (ICE) did not even exist. Yet the opening lines sound eerily familiar: “In recent months, many of our clients have asked about new quantitative tools to evaluate risk. This interest is especially keen as the result of recent derivative scares, the surprising downside volatility of U.S. equities, and the crash of Asian and emerging market economies”.

If only they’d known what was to come! Yet even in 1999, the fundamental message was the same as it is throughout the more recent papers in this guidebook: sound governance, multiple perspectives on risk and careful attention to structure.

Russell Investments // Risk management: A fiduciary’s guidebook // Risk management

/ p 80

Risk management In recent months, many of our clients have asked about new quantitative tools to evaluate risk. This interest is especially keen as the result of recent derivative scares, the surprising downside volatility of us equities, and the crash of Asian and emerging market economies.

While the new tools are useful, their singular focus on investment risk excludes other risks that should be addressed in a comprehensive risk management program. Clients often consider bets taken by active managers to be the most crucial area of risk. In our experience, other areas can cost a fund more: the custodian making erroneous transactions, outdated policies leaving investment roles unclear, plan managers purchasing unauthorized derivatives, an asset allocation inconsistent with client’s tolerance for risk. We have also found that to properly manage investment risk, and make the best use of new tools such as Value-at-Risk (VaR), it is beneficial to define and measure different types of investment risk.

This article presents our thoughts on how risk should be categorized, measured and ultimately minimized in pension funds or any other large pool of money with liabilities attached. We will first present a risk management framework, then move on to describe where VaR fits within this larger risk management process.

Russell Investments // Risk management: A fiduciary’s guidebook // Risk management

By: Nathan Dudley, Chris Hensel and Jennie Tyndall1

February 1999

1 Many people contributed to the content of this paper. We want to thank Janine Baldridge, Paul Bouchey, Monica Butler, Howard Crane, Kelley Fairbank, Steve Fox, Grant Gardner, Michael Hall, Catherine Higgins and Gloria Reeg for helpful discussions and comments on earlier drafts and Cate Goethals for her insightful editing.

/ p 81Russell Investments // Risk management: A fiduciary’s guidebook // Risk management

Exhibit 1: Risk management process

The risk management process The risk management process should follow a progression similar to that of other sound fiduciary practices. Risks must be first identified. Then, to the extent possible, it must be understood and measured. Next, it must be reduced to an acceptable level. Finally, the results of the process must be assessed.

In order to fully address the multitude of risks faced by any large investor, we believe fiduciaries must regularly focus on the five categories of risk shown in Exhibit 1. These categories are loosely defined—some risks could fall into more than one—but they provide a working risk management framework. The first two categories represent the areas most often left out of risk management programs; neither is directly quantifiable. The final three cover different types of investment risk; while measurable, each must be quantified differently.

FIDUCIARY RISK Fiduciary Risk encompasses those risks funds take by not having effective decision-making policies and procedures in place. This risk is listed as the first category because all the other categories contain components of Fiduciary Risk and rely on sound fiduciary management. To prevent risk, fiduciaries should develop a solid decision-making framework focused on long-term plan objectives rather than on short-term decisions. Such short-term focus can cause underperformance relative to policy benchmarks or shortfall relative to liabilities.

One area crucial to managing Fiduciary Risk is to clearly define in writing the responsibilities of all levels of fiduciaries in decision making. Without this written delegation, decisions take longer to be made and opportunities will be missed.

OPERATIONAL RISK While Fiduciary Risk relates to not having good policies and procedures, Operational Risk results from not following the adopted policies and procedures. To continue with the above example, not defining in writing the roles for various levels of fiduciaries is a Fiduciary Risk; not following the roles as defined in writing is an Operational Risk. Coupled together, the two risk categories could involve substantial costs. A Department

R I S K M A N A G E M E N T P R O C E S S

IdentifyUnderstand

MeasureManageAssess

FiduciaryRisk

OperationalRisk

Asset/LiabilityRisk

Asset Allocation and

Capital MarketsRisk

ImplementationRisk


of Labor audit leading to fines for improper administrative structure or poor governance, for example, can be expensive, disruptive, and embarrassing.

Operational Risk generally arises from time constraints, staffing constraints, or a lack of awareness. This risk can be caused internally or externally. Internal causes can include staff actions, manager hiring/firing, and transitions. Externally-driven examples include Year 2000, manager melt downs, natural disasters, and other unforeseen events. Following well-developed internal policies and procedures can mitigate most internal risks; externally-caused operations risks require awareness and vigilance.

INVESTMENT RISKS Because investments require multiple actions taken by multiple fiduciaries with different time horizons, we believe the associated risks should be split into three categories representing successive decision levels. The first covers the risk associated with a broad asset allocation as the result of an asset/liability study. The second represents the passive implementation of that allocation in the capital markets including structural tilts within the broader asset classes. The third takes into account the risk associated with active implementation of the chosen structure.

ASSET/LIABILITY RISK The first and most important area of investment risk is Asset/Liability Risk—the overall risk to the current and future surplus of the fund and therefore to maintaining benefit security for plan participants. Asset/ Liability Risk is normally addressed in an asset/liability study. Most funds undertake such a study at least every three to five years, depending upon major plan, company, or market changes.

Our asset/liability studies use broad asset class forecasts for assets and market-like discount rates for liabilities to examine the potential volatility of surplus, returns, contributions, pension expense, assets, and liabilities. This information helps the plan fiduciaries develop asset allocation policies that adhere to the long-term objectives and risk objectives for the plan.

For assets, the relative allocation between stocks and bonds is by far the greatest ongoing source of portfolio return variability. For liabilities, the greatest risk is a sudden drop in interest rates and the concomitant increase in liabilities. These risks are therefore crucial to manage by finding the allocation consistent with the risk tolerance of the fiduciary decision-makers.

Immunization—investing assets to match the cash flows of liabilities—can minimize Asset/Liability Risk, but at the expense of return potential: it limits the fund’s ability to grow surplus. By investing in equities, the fund expects the financial markets to finance future contributions.

ASSET ALLOCATION AND CAPITAL MARKETS RISK The next category, Asset Allocation and Capital Markets Risk, reflects the risks inherent in the passive implementation of the asset allocation policy developed from an asset/liability study. While the study determines allocation to broad asset classes, the policy itself may include structural tilts to small cap, high yield, or emerging markets that were not broken out in the initial allocation (which more accurately reflect inherent beliefs of fiduciaries). For example, the non-US equity allocation specified by the study may be split into 80% developed markets and 20% emerging markets, diverging from market- cap weightings.


IMPLEMENTATION RISK Implementation Risk is associated with the active implementation of the investment structure reflected in the previous level. Since active managers are expected to invest in different securities or weightings than the index, their choices add residual risk relative to the index and thus to the Asset Allocation Risk discussed earlier. For example, a deliberate small cap tilt (included in the Asset Allocation Risk) might be implemented with small cap value managers. This value bias adds additional risk. Of course, Implementation Risk assumes that managers are operating within fund guidelines for achieving the intended structure.

Summary All the risk categories are interrelated, and the investment risk categories represent successive potential deviations from one another. For example, not rebalancing non-US equity when outside of policy ranges represents an Asset/Liability Risk. Not rebalancing to the 80/20 developed/emerging market structure described above would reflect an Asset Allocation and Capital Markets Risk. Not rebalancing to manager weights periodically would be an Implementation Risk.

Of the three investment risk categories, we believe the first two—Asset/Liability and Asset Allocation—should be the focus for most institutional investors. Conducting timely asset/liability studies and allocating assets in a manner consistent with the studies are of primary importance to risk management.

Many investors, on the other hand, focus on what we define as Implementation Risk and dedicate their energy to monitoring the risk of active management. In our experience, this risk is less significant. In general, our clients take informed bets and their managers don’t deviate significantly from the asset structure they have so carefully developed. Once the first two levels of risk are adequately addressed, the risk in the third category is reduced.

The new measurement tools Within a larger risk management program, quantitative tools can be used to help determine the risk/return trade-off involved in the various types of investment risk. While many measures are available, including standard deviation, downside risk, tracking error, and information ratio, we will focus here on Value-at-Risk (VaR) because it is relatively new, less widely recognized than others, and is appearing frequently in the financial press.

VaR measures were developed for trading desks of large financial institutions. They have enabled traders to quickly calculate the downside risk exposures to a portfolio in a single dollar figure over a specified period of time. Traders can determine, for example, that a portfolio has a one in 20 chance of losing $1 million or more over the next week: a one-week 5% VaR of $1 million.

The intuitive appeal of VaR has been that it provides a single-dollar figure for risk. It may be measured using a variety of statistical methods. Rather than focus on different methods of calculating VaR, however, we will focus here on the different VaRS that could be calculated for the three categories of investment risk defined above.2 Each category has a different time horizon and requires different inputs to determine its VaR.

2 Risk is most often measured in statistical probabilities and we can seldom be totally confident of the accuracy of our estimates. Estimation error affects the standard deviation of the probability distributions, altering the risk estimates. Because accurately projecting risk and return assumptions in a financial model is difficult at best, it is not surprising that different companies selling VaR systems have developed different methodologies. A current shortcoming of VaR is that there is no industry standard for its calculation and differences between methodologies can be significant. Another shortcoming is that many VaR models are historically based. If the future is radically different than the past, these VaR measures will be judged less useful.


Exhibit 2 lists the information necessary for measuring the VaR for each category. For clarity, we will refer to the VaR for Asset/Liability Risk as VaRA/L; for Asset Allocation Risk (structure), VaRS; and for Implementation Risk, VaRM.

Exhibit 2: Information requirements

VARA/L FOR ASSET/LIABILITY For Russell clients, the appropriate level of Asset/Liability Risk is determined through asset/liability studies that use a risk model with consistent risk measures over multiple time horizons. These studies take into account liabilities and use rebalancing assumptions. We believe it essential to incorporate liabilities into these studies because they drive plan sponsors’ risk exposure decisions. In addition, liability changes can dramatically affect the funded status of a pension fund and therefore should be included in any calculation reflecting the fund’s health.3

While asset/liability studies enable fiduciaries to create the broad outlines of their portfolios based upon their comfort level with risk, VaRA/L is one way to measure the downside risk to the fund over long time periods (three to five years). It includes the risk of assets falling and/or liabilities increasing, combining these two into a single measure. Determining the fund’s VaRA/L requires broad asset class forecasts as asset inputs and liability inputs from an actuary or from a liability model.

VARS FOR ASSET ALLOCATION (STRUCTURE) While VaRA/L focuses on a few broad capitalization weighted indexes, VaRS attempts to measure additional risk resulting from structural deviations from these capitalization weights. If a plan sponsor creates an asset class structure that intentionally tilts away from capitalization weights, VaRS would measure the aggregate risk from this structure. Tilts could include size, style, or sector. VaRS measures structural risk using passive benchmarks. This use of passive benchmarks is one of the main differences between VaRS and VaRM. VaRS measures structural risk using passive benchmarks. This use of passive benchmarks is one of the main differences between VaRS and VaRM.

Asset/ liability risk

Asset allocation and capital markets risk

Implementation risk

Risk measure VaRA/L VaRS VaRM

Asset allocation used

Policy asset allocation

Structural policy allocation

Actual investment allocation

Level of asset information

Few broad asset classes

Indexes necessary to capture structure

Security level or index proxies

Liabilities included Yes Optional No

Level of liability information

Detailed actuarial information

Duration matched bonds as proxies

N/A

3 This is highlighted in Russell’s October 1998 Conversation Piece, “Flat Markets and Falling Interest Rates: Painful Pension Fund Accounting.”


Exhibit 3: Comparison of VaR

In addition to asset data, VaRS can include liability data. Liabilities can be divided between those for active employees and inactive employees. Durations can be calculated separately for the two groups. Negative weights are given to the zero-coupon bonds matching the durations of active and inactive liabilities. Thus VaRS can include both assets and liabilities and can be calculated for different time horizons; it is most useful, however, for those between one and three years.

VARМ FOR IMPLEMENTATIONThe risk of a sponsor’s actual implementation using active managers is captured in the third measure, VaRM. This measure accounts for the actual investments and allocations of the managers and can be calculated in two ways. The more complex method builds up from security level information, estimating the relationship (covariances) between all securities held based upon historical asset return relationships or from simulations drawn from historical covariance distributions. This method is very expensive and data intensive.

In the second and simpler method, a small number of indexes are used as proxies for a large number of securities. All managers are mapped to their benchmarks so their allocations are represented by a set of indexes. Alternatively, index weightings could be obtained from a style classification system. Once the manager’s index weightings have been determined for each asset class, the covariances of the indexes is all the information required. This simplified method—using index information as a proxy for individual securities—dramatically reduces the time and data necessary to calculate VaRM. For more plans who elect to use VaRM, we believe the security level data approach adds little value over the proxy approach.4

VaRM is a shorter-term risk measure than VaRA/L or VaRS since it focuses on the managers and excludes liabilities. It should be used for time horizons no longer than a year. Exhibit 3 lists the data required and time horizon for each of the three VaR measures.

SUMMARYPrudent risk management is a process that should take into account multiple categories of risk rather than a single aspect. Many sponsors focus their efforts primarily on monitoring the risk of active managers.

Within a risk management program, VaR is a measurement tool that can provide good information about the downside risk of pension funds. While it traditionally has focused only on assets, the impact of liabilities is also important to managing risk, especially over longer time horizons. Thus, the VaR measures we have proposed take liabilities into account. Of those measures, the most useful is VaRA/L, which captures risk at the

Risk category Data Time horizon Accounts for

VaRA/L Asset/liability risk Broad indexes Long Assets and liabilities

VaRS Asset allocation and capital markets

Specific index weightings

Intermediate Assets and liabilities

VaRM Implementation Security weightings

Short Assets only

4 Exceptions include plans with significant internal asset management or plans using VaR-adjusted performance evaluation. In both cases, sponsors may benefit from the precision of security-level information.


broad asset class level and is essentially an asset/liability study. In our experience, using VaR to capture the successive levels of risks—Asset Allocation Risk and Implementation Risk—provides decreasing value.

The most important aspects of risk management are developing and following sound policies and procedures. These should evolve over time. Good review processes keep policies and procedures current and ensure that a fund’s risk management program continues to reflect its structure and priorities.

In the area of investment, we believe broader risks deserve priority. Plan sponsors should pay the most attention to Asset/Liability Risk-risk at the broad asset class level. Next in priority should be Asset Allocation Risk, the additional risk created by asset class structure deviating from cap weightings. The next level of risk, Implementation Risk—that resulting from additional deviation by active managers within the policy weights is less important.

/ p 87

Narrowing the knowing-doing gap in investments through effective fund governance

The next few papers in our guidebook address the question of fund governance. This is a subject inextricably linked to risk management, because without good decision making, there will never be good risk management. Behind poor outcomes generally lies poor decision making.

The challenge of good decision making is one that is not restricted to investment. As a result, there are lessons that can be drawn from broader fields. This next paper looks to the fields of organizational behavior and corporate governance, and in particular to the book The Knowing-Doing Gap, for such lessons. It examines the role played by organizational design, the interacting roles of different parties, delegation and the need for ongoing evaluation and draws a picture of what best practices look like.

This article was first published as a Russell Canada Monograph in 2006.

Russell Investments // Risk management: A fiduciary’s guidebook // Narrowing the knowing-doing gap in investments

/ p 88

Narrowing the knowing-doing gap in investments through effective fund governance Over the last 15 years there have been a number of institutional investment research studies that have been conducted on the barriers to investment excellence or the problems of implementation slippage. These research studies have noted a host of issues which appear to have led to poor performance in investment funds. The common theme of the research is that investment underperformance is a symptom of ineffective decision making or poor governance practices.

Yet despite the fact that these research papers are common knowledge, that many good solutions have been offered up by the pundits/consultants and that investment fiduciaries know what should be done, poor governance practices largely persist in the institutional investment industry. The net result is that the average institutional investment fund in Canada generally pays active management fees for passive performance or underperforms. When (investment) industry solutions are known but most participants (fiduciaries) fail to respond, it’s often a good time to search for answers in other disciplines. Is this type of ineffective behaviour a common phenomenon in most industries or is it only common to the institutional investment industry? To answer that question and various others, this paper sought guidance in the two most relevant disciplines, organizational behaviour and corporate governance.

Introduction In 1999 Jeffrey Pfeffer, a professor of organizational behaviour at Stanford Graduate School of Business and Robert I. Sutton, a professor of organizational behaviour at Stanford’s School of Engineering coauthored a book entitled The Knowing-Doing Gap. In short, they noted that: “Time after time people understand the issues, understand what needs to happen to affect performance, but don’t do the things they know they should.”1 The problem was deemed not solely one of analysis (a lack of knowledge) but rather one of poor implementation (a lack of discipline or appropriate process). Based on these introductory comments, it would appear that the institutional investment industry’s failure to deal with its governance issues is therefore not unique!

This paper will attempt to apply the precepts of this insightful text on organizational behaviour to the world of investment management, in the quest for practical solutions to the governance problems that have plagued institutional funds for years. In short, it will attempt to provide fiduciaries with a competitive advantage for their fund by recommending positive action.

By: Bruce Curwood

October 2006

1 Pfeffer, Jeffrey and Sutton, Robert I., The Knowing-Doing Gap, Harvard Business School Press, 1999 p. 11.

Russell Investments // Risk management: A fiduciary’s guidebook // Narrowing the knowing-doing gap in investments

/ p 89Russell Investments // Risk management: A fiduciary’s guidebook // Narrowing the knowing-doing gap in investments

The knowing-doing paradox Pfeffer and Sutton’s text explores one of the great mysteries in organizational management: “Why knowledge of what needs to be done frequently fails to result in action or behaviour consistent with that knowledge? They call this the knowing-doing problem— the challenge of turning knowledge about how to enhance organizational performance into action consistent with that knowledge.”2 In the remainder of this section, I will be quoting and paraphrasing extensively from the above co-authors in an attempt to summarize their key thoughts.

Even though companies pour billions of dollars into training programs, consultants and executive education, the so called knowledge advantage is often a fallacy. The reason is not that knowledge isn’t important. It’s that most companies know, or can know, the same things. Moreover, even as companies talk about the importance of learning, intellectual capital, and knowledge management, they frequently fail to take the vital next step of transforming that knowledge into action. The Knowing-Doing Gap confronts the paradox of companies that know too much and do too little.

“Research demonstrates that the success of most interventions designed to improve organizational performance depends largely on implementing what is already known rather than from adopting new or previously unknown ways of doing things. The so called, best practice, hints were actually well known practices, with the extra dimension that they were reinforced and carried out reliably in better performing organizations. They are in fact common sense. Yet it is interesting how uncommon common sense is in its implementation.”3

What is important is not so much what to do, but why and how it is done (the underlying philosophy and views of the people/business that provide a foundation for the practices). Attempting to copy just what is done—the explicit practices and policy—without holding the underlying philosophy is a far more difficult task and is less likely to be successful. Competitive advantage comes from being able to do something that others don’t do. Anyone can read a book or attend a seminar. The trick is to turning that acquired knowledge into organizational action. If you and your colleagues learn from your own actions and behaviour, then there won’t be much of a knowing-doing gap because you will be “knowing” on the basis of your “doing” and implementing that knowledge will be substantially easier.

Pfeffer and Sutton outline numerous barriers to performance in most organizations. The five main barriers noted in the text are:

1. When talk substitutes for action;

2. When memory is a substitute for thinking;

3. When fear prevents acting on knowledge;

4. When measurement obstructs good judgment; and

5. When internal competition turns friends into enemies.

Barriers to action in investments Many of the issues outlined in The Knowing-Doing Gap seem relevant to the investment industry. Having determined that the problems facing institutional investment funds may not be unique, the balance of this paper will address whether the barriers to performance

2 Pfeffer, Jeffrey and Sutton, Robert I., The Knowing-Doing Gap, Harvard Business School Press, 1999 p. 4.3 ibid p. 15.


outlined above, can be applied to the investment industry: If so, can these barriers then be eliminated or at least reduced? What does the specific, factual evidence in the investment fund industry show? If it can be demonstrated that similar barriers exist in the investment industry, then perhaps, the same guidelines for action outlined in The Knowing-Doing Gap can also be applied to limit their influence.

Therefore let’s first examine each of the barriers noted above, to see if we find evidence of their occurrence in the investment industry:

1. Talk not action There is no end to the amount of talk in the investment industry, but little time for decision-makers to strategize, innovate or act proactively. Most governing trustees (boards and investment committee members) have separate, full-time careers in unrelated endeavours, who can devote only part of their time to investments at regular quarterly meetings and occasional industry conferences. There is a widespread failure to delegate to trusted full time staff, while retaining proper oversight. As such, meeting agendas are generally long and cluttered with administrative issues and by manager presentations focused on the past instead of the future.

2. Memory substitutes for thinking The text was quick to point out that “when people are unsure about how they or their organizations should act, they automatically imitate what others do.”4 Investment funds are particularly likely to imitate people or organizations, deemed similar to them, while ignoring the economies of scale or differing circumstance:

• Ontario Teachers Pension Plan is regularly used as the standard of excellence in Canada, yet most funds lack their fund size ($95B), resources (120 full time staff), governance approach (incentives tied to risk budgets) and extensive in-house research.

• The Yale and Harvard endowment funds are similarly touted in the U.S. Yet research has shown that size, resources and early innovation have been some of their key success factors. Most funds of smaller size have significantly underperformed these larger counterparts.

3. Fear and mistrust Mistrust or misalignment amongst trustees, key staff (CIO, treasurers, fund sponsors, etc.), consultants and investment providers, pervades the investment industry. Although control and proper oversight will always be important, better due diligence and greater delegation to those with the most expert knowledge, time and resources, are essential to create successful partnerships based on collaboration, transparency and open communication. The appearance of seeming incompetent or less than proactive may also cause some trustees to terminate investment managers based on poor short-term results, to ease pressures from their constituents.

4. Measurement obstructs good judgment A focus on short-term performance, in this long-term investment game, is a recipe for disaster for institutional funds. Fear of not meeting multiple and potentially conflicting objectives may lead to inappropriate, short-term decision making. As the text clearly states the time scale of the measurement (a quarterly corporate approach) or how often the fund assesses results helps to establish the time horizons that tend to govern behaviour in the organization.

5. Internal competition High turnover of busy committee members can be a detriment to performance, as

4 Pfeffer, Jeffrey and Sutton, Robert I., The Knowing-Doing Gap, Harvard Business School Press, 1999 p. 73.


the new members may not have the opportunity to learn from the funds’ past mistakes. With competing demands on their time they often fail to fully understand the investment issues and decisions they face, thinking it is old hat or they can make better use of their limited time.

The aforementioned are just a few of the many examples of detrimental investment decision practices. A more complete listing evidencing the five main barriers to performance in investments is contained in Appendix A.

Best practice to narrow the knowing-doing gap in investments Based on the above, there is little doubt that the knowing-doing gap exists in the investment fund industry. The previous section demonstrated numerous behavioural and organizational barriers that most investment funds face. Pfeffer and Sutton noted that there were no simple answers or an easy solution. Nonetheless, they described some recurring themes that were evident and outlined eight guidelines for action, as follows:

• Why before how: philosophy is important;

• Knowing comes from doing and teaching others;

• Action counts more than elegant plans and concepts;

• There is no doing without mistakes;

• Drive out fear;

• Beware of false analogies;

• Measure what matters;

• Leaders know how to spend time and allocate resources.

They concluded that organizations that can turn knowledge into action, avoid the “smart talk trap.” Executives must use plans, analysis, meetings and presentations to inspire deeds, not as substitutes for action. In short, our co-authors had realized the importance of governance, but had arrived at that same solution as the investment pundits/consultants from a somewhat different perspective, organizational behaviour. This paper will therefore put forth a number of practical actions or processes to limit those barriers to action and narrow the knowing-doing gap in investments through better governance.

Governance is about assigning the authority to make decisions to ensure the organization is run effectively. Although regulatory compliance isn’t always easy, the real challenge is performance. Effectiveness is a function of the quality of the process rather than the time devoted to it. Few would debate that a properly governed investment fund has a much better chance of delivering the investment promise. John McCallum, professor of finance at University of Manitoba points out that more probably goes wrong in the design phase than anywhere else. “A badly flawed investment mandate is often an accident waiting to happen.”5 The investment mandate drives the standards that fund sponsors use to judge success. Each individual within the organization must have focus and understand the fund’s overriding investment objective and their role in its achievement. Action is required when performance continually and meaningfully appears to be falling short of that objective in the long term.

D. Don Ezra, director of strategic advice at Russell Investment Group, identified four broad areas that need to be addressed to implement governance best practices in the industry.

5 John S. McCallum, Governing the Investment of Pension Fund Assets, Ivey Business Journal, (January/February 2003) p. 1.


1. Governance must be an explicit rational focus;

2. Fiduciaries need to know what is important;

3. They need to work as a team;

4. Use a long-term multi-dimensional framework for evaluation; and

All four, although distinct in their own right, are often interrelated or linked.

Exhibit 1: The chair and CIO play a vital role in effective fund governance

Source: Based on the “Proposed Board Effectiveness Model” of Richard Leblanc, PhD.

Let’s delve into each of these broad areas and offer some practical observations and time tested advice. For governance to be an explicit rational focus there must be unambiguous accountability, where each party knows their responsibilities and constraints, but does not encroach on each other’s territory. The investment board and/or investment committee (trustees) must recognize their primary role of higher level decision making is generally to set the strategic direction and objectives, determine broad risk management policy and perform fund oversight to track progress in achieving the primary investment goal. As part time members of the organization they must acknowledge the limitations of time and recognize the need to delegate extensively to full-time staff and third party providers with greater investment expertise and a full-time focus on investments. Effective delegation relates in particular to day-to-day operational issues but also in many cases to formulating strategies and tactics. All too often these higher level trustees underestimate the resources required for successful strategic analysis and implementation. A logical framework will evaluate each and every issue based on the skills required and determine whether those requisite skills are in existence

RESOURCES

TIMELY

DECIS

IONSRESEARCH / BELIEFS

A COMMONVISION

BOARD COMMITTEEMEMBERSHIP

Tenure / Recruitment / OrientationBehaviorExpertise

ExperienceTraining / Development / Appraisal

GOVERNANCEPROCESS

Proper business planAgendas and documentation

Investment educationand innovation oversite

Prioritize by importance

GOVERNANCESTRUCTURE

Unambiguous accountabilityFocust on what matters• Primary objective• Risk management• Strategy

Delegation to Professionals

BOARD EFFECTIVENESS

A culture of collaboration through partnership and rigorous self-assessment


or need to be acquired (internally or externally). Insiders tend not to trust outsiders (as well as other insiders in some cases) and can fail to provide them the necessary power to facilitate change, but without fresh eyes and outside experience, insiders may not be able to achieve success in certain areas on their own.

In addition, trustees need training, development and regular appraisal. Trustee performance demands a learning approach that continuously refines competencies which will add fund value. The mandates that govern trustees must be rigid and specific enough to keep them from meddling in operational issues but flexible enough so they challenge but not stifle creative strategic thinking. A culture in which direct, serious and important questions are asked and plainly answered is a good way to ensure that there is trust between the trustees and staff. In short, mandates must be clear and focused with all fiduciaries (including staff) understanding their role. In particular the role of the board, the chair and the chief investment officer should be clearly defined. If the trustees and senior managers are going to trust each other, it is crucial that each knows the limits of their roles. The role of the chair, as defined by Larry Tapp, former dean at the Richard Ivey School of Business at University of Western Ontario is to:

• “Build a strong, effective, well-balanced and representative board and committees;

• Set the board agenda and schedule meetings;

• Manage the affairs of the board;

• Ensure all committees are working effectively;

• Head the evaluation of the CIO and review of performance;

• Act as advisor and sounding board to the CIO;

• Provide a link between management, the board and stakeholders;

• Ensure the board and committees are receiving timely and appropriate information, before during and after meetings;

• Appraise the board’s and committees’ performance”.

Leadership comes from the chair, who must cultivate a common vision for the fund with the CIO and mentor committee members to clearly articulate and realize that vision. Effective leadership by the chair is vital to the success of the fund, as is developing a competency and behavioural-based matrix for recruitment, continuing education, development and tenure, while encouraging a climate of constructive challenge. These functions are well captured in the exhibit on page five, which is largely based on a board effectiveness model, put forth by Richard LeBlanc, PhD at the Corporate Governance Program, Schulich School of Business.

This design phase is therefore one of the keys to success and should be well thought out and planned. It encompasses several sub-segments including: member orientation; functional behaviour; a proper organization; and succession planning. An organized orientation program is of great assistance to all members. Not only does it outline the priorities, by-laws, regulation, structure, policies and constraints, but it should also document the current investment beliefs and process rationale, acting as a reference guide for past and future decisions. Best practice includes having a trustees manual that provides background material that is always current and available to the new trustee, as well as one-on-one orientation sessions with both the chair and CIO and perhaps additional mentoring by a senior trustee. All forms of communication must be in plain understandable language, so misinterpretation is limited. Good organization requires a well thought out agenda where items are ordered and prioritized by importance, with sufficient time for a complete discussion of important issues and decision making.


Agendas must limit presentation time and maximize discussion time for trustees to make timely decisions. If the members are to bring a critical eye to the debate they must be well versed and receive suitable material in a timely fashion well in advance of the meeting. They must be prepared to do their pre-reading homework and understand the technical issues. Routine matters should be noted and tabled without undue discussion unless there is an exception to policy. This process allows for adequate discussion of the key issues. Good investment research will help to discern those key issues.

Minutes should abide by the four C’s by being correct, concise, clear and consistent. They should be extensive enough to record decisions taken and the important considerations, while outstanding action items should be noted along with the individual responsible and due date for completion. Follow-up on the progress of these action items should occur at the next meeting and until full implementation.

Member assessment and succession planning assumes that trustees have assessed the necessary expertise and experience required on the investment board/committee. The Rotman International Centre for Pension Management at the University of Toronto confirmed that board of trustee selection and board effectiveness evaluation processes continue to be problematic for many of the globe’s largest funds.6

To summarize, the trustees’ role is one of oversight but they must be decisive and prepared to act when action is required. There should be an annual business plan which allows sufficient time, primarily for risk definition/management, as well as education/innovation and secondarily for the monitoring, implementation and tracking of performance to objectives. Everything on the agenda should have a clear purpose and be logically ordered, with time allotted by importance. It has often been observed that the two greatest problems in investment governance are that trustees:

1. Don’t understand their role and tend to meddle (micromanage) in areas beyond their mandate; and

2. Spend a disproportionate amount of time on routine administrative and reporting aspects to the detriment of these higher level or emerging issues.

At the end of the day this reflects a failure to sufficiently delegate to investment professionals and set up an effective monitoring process, which reports on administrative matters but only discusses them, based on exception. This naturally leads us to the second area of governance best practices: Namely that trustees need to understand what is important in investment management. Eighty percent of directors reported that their boards spend more time on monitoring than strategizing.7 Trustees need to spend more of their time engaged in policy and strategy—not owning the strategic process but providing their expertise and guidance. They need to understand their primary investment objective (be it risk adjusted return maximization or surplus risk management etc), stay focused to that long-term goal and not succumb to short-term, behavioural thinking or multiple, secondary and often competing, objectives. Ultimately they need to manage risk in a disciplined fashion, by spending more time on understanding innovative, new strategies, the asset allocation decision relative to the liabilities and overseeing risk within the fund. Each member of the committee need not be an investment expert, but there is no substitute for investment training, know-how and experience. The board/committee must be able to distinguish noise and nonsense from those items that are meaningful and material. Hindsight is not insight unless we can learn from past mistakes. One of the most common mistakes in investment management is not recognizing

6 Keith Ambachtsheer, Peter Drucker’s Pension Legacy: A Vision of What Could Be, Ambachtsheer Letter, (November 2005) p. 3.

7 Robert Colman, The Better Board, Corporate Governance Quarterly (Spring 2006) p. 8.


that assessing money managers is a specialized skill best left to full-time, dedicated, investment professionals who can do adequate due diligence and generally is not a good use of trustees time. After agreeing on policy and strategy, money manager and portfolio due diligence is best delegated to full-time staff or outsourced, due to time constraints. However oversight must be retained on both process and performance.

There is also no substitute for good research! In addressing investment proposals, reasoning begins with research, such that there is enough relevant information to make sound decisions. The availability of these resources is highly important. Having good in-house research or partnering with providers with best in class research, will limit errors in judgment. Also recognizing that there are no sure things in this industry (only probabilities) should force trustees to ensure they can live with the downside of any particular investment strategy. Risk analysis is essential! To determine what they need to know, trustees should insist that they have access to the best advice available and the information they need to make informed decisions. Those practitioners with implementation skills should be valued highly, as they have knowledge by doing. The chair plays a key role in assessing the expertise and experience of the board/committee members, while determining their collective knowledge as a group. Expertise entails not just theoretical investment wisdom but must be based on industry experience, which ensures patience, discipline and self control. Members need to know when to act, when to consent and when to hold the course. The chair must regularly assess the requisite business savvy of the group in regards to its deliberations and upgrade as necessary. Understanding the trustees’ relative strengths and weaknesses will lead to the right additions to membership and continuous improvement in decision making. There is no doubt that an effective committee must be composed of members with the diverse competencies (experience and expertise) to fulfill its strategies. In addition, however, the members must be able to work collaboratively together, to come to effective decisions. Partnering with professionals who have relevant, robust research and demonstrated implementation skills, will ensure better strategies, better outcomes and timely decision making by trustees.

Working as a team was the third broad area required to achieve better investment governance. After putting in place the right structure and the right membership, the key to better governance, lies in the working relationships between trustees, staff and providers, as well as the social interaction within and between groups. This process determines how decisions are made and is often based on the individual traits or behaviours of the members themselves. In short we are talking about interdependence not independence. Now it is true that independence of mind and character are valued member traits as it can lead to challenging the status quo, when change is imperative. Nonetheless it’s just as important that criticism come in an environment of mutual respect and positive feedback for good, new ideas. Investment funds are made up of a diverse group of individuals who behave differently. How individuals interact with each other has a great deal to do with determining effectiveness. The members need to have read the investment materials, be prepared, and to have invested the time to understand the complexities of the interdependent issues. With ineffective groups more members sit silent, hoping their collective incompetence isn’t noticed. The process, previously discussed, of assessing individual members will end the practice of retaining ineffective and dysfunctional individuals, simply because it is inconvenient to remove them. Group and individual assessment by members can assist the chair in fulfilling these responsibilities. Fostering trust, transparency and a culture of collaboration, while recognizing we can learn from our mistakes, ensures early detection of investment issues. This is particularly important in dealings with staff. Caution, prudence and careful investigation are the cornerstone of investment trustee behaviour, but it shouldn’t come at the expense of ridicule or putting forth unattainable goals, that stifle creativity. Trustees that are kept better informed of developing issues are no doubt, better qualified to exercise sound judgment in a proactive manner. All participants in the investment


process must therefore feel a part of the solution, that they have input and that their participation is valued. This should include staff and third party providers, such as consultants and investment managers. If the trustees have ensured proper and adequate due diligence in the selection process (including an analysis of potential conflicts of interest), they should empower trusted and committed partners, whose understanding is recognized and input is valued. Proper documentation, accountability and a system of checks and balances ensure responsibilities are delegated but nevertheless continually monitored. This should promote long standing relationships where effective communication and transparency are ever present and promoted.

The latter point brings us to the final area or the requirement for long-term, multi-dimensional evaluation. Successful investing is not about short-term gratification. It is about measuring what matters and not dwelling on the recent past. If trustees focus more on inconsistencies in the qualitative factors (people, processes, policy and philosophy) and risk management as opposed to over reliance on short-term, often random, quantitative performance, they will be better served. Intelligent investors are disciplined and establish a consistent, risk managed approach, focused on understanding and progress to the primary investment objective. Patience and self control are tempered by independent thinking and an eagerness to understand new concepts through research and education. Diversification is one of the primary tools at an investor’s disposal for risk reduction. More concentrated bets should only be applied to higher conviction strategies, which have been supported by sound research.

The aforementioned recommendations have been summarized in the attached table (Appendix A) as suggested positive actions to address many of the barriers and issues to effective investment decision making.

Concluding thoughts In conclusion, applying the aforementioned best practices in investment governance is likely to lead to a narrowing in the knowing-doing gap. It is a rare fund that would not benefit from a periodic review of its investment fund governance practices. But best practices are often specific to the fund’s circumstance and cannot be implemented by checklists and codes. Rather a positive climate of assessment, analysis, innovation and education must be fostered, to promote collaboration and challenge the status quo. These are the governance challenges facing most investment trustees.

I will leave the last word to Bob Garratt, professor of corporate governance at the University of London, who boldly stated that trustees “should be selected, trained, developed and appraised on their abilities to:

• Formulate policy and so give foresight;

• Think strategically;

• Collect accurate external and internal data;

• Generate imaginative ideas;

• Be capable of open questioning and critical review ;

• Be rigorous in risk assessments;

• Be collegial in strategic decision making;

• Learn systematically from its strengths and weaknesses;

• Ensure organizational capabilities for successful strategic implementation; and

• Ensure rapid feedback for honest information on the implementation of the strategy.


Pension governance has always been difficult, but given the extensive changes to the market that are expected to take place over the next several years (due to the separation of alpha and beta, greater use of liability driven investing and countless other, new investment approaches that many seem to point to as an inflexion point in the industry), the fundamental principles outlined in this paper should become even more relevant for fiduciaries today. In short, effective decision making has never been easy, and it’s getting harder; but it’s never been more important than it is today.

References Curwood, Bruce B., Deriving a Competitive Advantage in Investments, Russell Viewpoint, (September 2004).

Curwood Bruce B., Total Fund Performance: How Much Value is Enough?, Russell Viewpoint, (April 2004).

Pfeffer, Jeffrey and Sutton, Robert I., The Knowing-Doing Gap, Harvard Business School Press, 1999.

McCallum John S., Governing the Investment of Pension Fund Assets, Ivey Business Journal, (January/February 2003).

Ezra, Don D., Effective Fund Governance, Presentation (2005).

Fairbank, Kelley, The Good Fiduciary, Presentation (2004).

Ezra, Don D., Excellence in UK Pension Fund Management: A Perspective on Current Practices, Russell London Monograph (October 1999).

Tapp, Larry, Building Trust between Managers and Directors: Seven Steps to Better Governance, Ivey Business Journal, (September/October 2002).

Nadler, David H., Building Better Boards, Harvard Business Review (May 2004).

Ambachtsheer, Keith P,. Peter Drucker’s Pension Legacy: A Vision of What Could Be, The Ambachtsheer Letter (November 2005).

Ross, Robert, Myners, Principles for Institutional Decision-making: Review of Progress, Russell London Monograph (May 2005).

Gillies, John, The Challenge of New Ideas—Avoid Carelessness not Risk, Russell London Monograph (November 2004).

Leblanc, Richard W., New Corporate Governance Guidelines: What Corporate Secretaries Need to Know, ICSA The Administrator (June 2005).

O’Donovan, Gabrielle, Change Management—A Board Culture of Corporate Governance, Articles (2005).

Garratt, Bob, The Future for Boards: Professionalisation or Incarceration?, Keeping Good Companies (October 2003).

Leblanc, Richard W., Preventing future Hollingers, Ivey Business Journal (September/October 2004).

LeBlanc, Richard and Gillies, James, The Coming Revolution in Corporate Governance, Ivey Business Journal (September/October 2003).

Reiter, Barry J. and Rosenberg, Nicole, Meeting the Information Needs of Independent Directors, Ivey Business Journal (January/February 2003.

Hall, Gordon M., 20 Questions Directors Should Ask about their Role in Pension Governance, The Canadian Institute of Chartered Accountants/Institute of Corporate Directors (2003).

Cantor, Paul, Board Performance: A Three—Legged Stool, Ivey Business Journal (November/December 2005).

Kelly, Sorca, A Model Agenda for an Investment Committee, Russell London Monograph (April 2004).

Colman, Robert, The Better Board, Corporate Governance Quarterly (Spring 2006).

Laverance, Phillip A., The Art of Writing Minutes, Chartered Secretaries Canada 2000.


Nissing, Nick, Breaking the Barriers, CMA Management (June/July 2005).

Bruno, Mark, The Bigger They Are, the Better They Perform, Pensions & Investments (January 23, 2006).

Appendix A: A summary of issues and suggested positive actions

Barriers Issues Actions

1. Talk a) Too little time at infrequent quarterly meetings

Set an annual business plan;

Prioritize agenda by importance;

Delegate to qualified experts to make timely decisions, within established risk parameters;

Minimize presentation time and maximize discussion time of key topics;

b) Part time not devoted to investments

Improve members with investment expertise through succession planning;

c) Long agendas Focus on what matters or high level decisions (risk management, asset allocation, strategic thinking, creativity and oversight);

Expedite administrative matters by report and delegate manager due diligence/evaluation to staff or outsource;

Review staff process for reasonability and potential conflicts of interest;

d) Jargon Be clear and use plain language;

e) Overly critical Proper oversight of process;

Ensure adequate checks and balances;

Develop partnerships with implementation experts;

Adequate due diligence by staff;

Key chair role to coach proper conduct and functional behaviour;

f) Beliefs discussed but not truly implemented

Clearly understand your primary objective relative to the liabilities and focus on attaining that goal;

Draft statement of beliefs with strategies based on strong, sound research that is continually referenced in reviewing creative solutions;

2. Memory a) Following leaders

b) Imitating larger funds

Ensure have sufficient resources, the beliefs, the risk tolerance and the processes to attain the results;

c) Unwilling to vary from past ideas

Make time for strategic thinking and new ideas while constantly ratifying, reviewing and refining past approaches;

d) Misguided assumptions

Foster an environment to respectfully challenge past practices through sound research and regular review;

e) A checklist approach

Make more time for creative thinking and innovation;

Review the adequacy of your resources to successfully implement new and existing approaches;

Don’t be penny wise and pound foolish;



3. Fear a) Micromanagement Engage expert investment staff and providers performing sound due diligence in hiring, with on-going evaluation;

Establish clear lines of authority and accountability;

Ensure trustees focus on higher level decisions and staff focus on implementation;

Be flexible so as not to stifle creative thinking;

Know when to act decisively and when not to act at all;

Review process/risk analysis not minutia;

b) Fear of job loss Promote an environment of collaboration, transparency and clear communication;

Work as a team and promote team skills;

Create a partnership environment with skilled staff and providers, who have implementation expertise, to develop a solutions based approach;

Learn from past mistakes;

c) Short-term decision making

d) Multiple competing objectives

Focus on the long-term overriding goal;

e) Herd instinct/fear of being different

Members require training, development and regular appraisal;

Clearly articulate strategies based on sound research and focus on how it will attain the overriding goal;

4. Measurement a) Short-term focus Focus on what matters and don’t respond to short-term random noise;

A risk managed approach with long-term multi-dimensional evaluation;

b) Quarterly performance reviews

Delegate to staff, with regular review of asset class strategy and structures incorporating risk tolerance;

c) Multiple quantitative objectives

Focus on the future and achieving the primary objective;

Be proactive and follow up action items regularly;

d) Hard measures drive out soft

Ensure ongoing qualitative assessment of managers and providers by staff;

Review and enhance process;

Track progress to attaining the business plan and the primary goal;

Appraise board and committee performance;

e) Compensation tied to inappropriate measures or time horizons

Rewards based on contribution to the overall goal and attainment of the business plan



5. Competition a) High member turnover Orientation, mentoring, succession planning,

Statement of Investment Policies and Beliefs and current trusteemanual (regulation, statute, etc.) are all required;

Proper documentation is a must;

Follow the 4 C’s of minutes (correct, concise, clear and consistent);

b) Overburdened members

Delegate and ensure full-time, expert resources adequate;

c) Few rewards/considerable career risk

Delegate to qualified investment resources and compensate accordingly;

d) Antagonistic behaviour

Develop partnerships with a solutions orientation;

Promote cordial social interaction to encourage effective working relationships;

Recognize the value of independent thought but interdependence of actions;

Chair needs to reprimand or remove dysfunctional members.

/ p 101

Investment governance: A pragmatic update

In 2010, as the attention of the world is focused on risk management, it is our belief that institutional investors will never get risk management right until they get fund governance right.

D. Don Ezra has been writing about investment governance for decades. This next paper gives his perspective on the state of play in 2010: how is the industry doing? He finds that some things are constant: the desire to avoid blame; the shortfalls in training and resources. He finds that decision-making roles remain unclear in too many cases. He lists the areas where funds most commonly fall short. He concludes that “[n]ot nearly enough time is spent… on studying decision making as a process that can be improved by the application of principles.”

Russell Investments // Risk management: A fiduciary’s guidebook // Investment governance: A pragmatic update

/ p 102

Investment governance: A pragmatic update In a sense, there is nothing new to report on investment governance. It is part of the subject of governance in the wider sense, with the principles applied specifically to the governance of institutional investment portfolios. The key principles are well-established, and many relevant lessons have been drawn.

But the principles are not all well known, perhaps not even widely followed. And decision making is never easy, because as humans we are subject to several well-documented behavioral flaws. So periodically it is worthwhile to remind ourselves of the principles, remember the lessons we have forgotten, and review the latest thinking about improving the processes involved in making decisions, so that those decisions can themselves be better.

That is the purpose of this paper.

We start with an explanation of what governance is, and why it is important. Then we move to an anecdotal reminder of some of the lessons that studies have taught us. From there we outline some principles for creating a good governance structure. We expand on practical aspects that, in our experience, greatly facilitate the adoption of improved practices. And finally we review a very recent study, conducted in the UK, on the current state of the art, with results that confirm that progress towards best practices is painfully slow.

Section One: Governance defined, and why it is important It is actually easier to start by saying what governance is not. It is not management. Management is about running an enterprise. Governance is about seeing that it is run well. So management is judged by results; in contrast, governance is about process.

In practice, therefore, a simple definition of governance is that it is the decision and oversight structure established in any enterprise.

Dictionaries can provide abstract definitions, but we have found it much more practical to think immediately of governance as a structure that establishes the so-called “decision rights” in an organization: who does what.

The contrast between management and governance comes through in another way. Process and outcomes are not the same. Processes can be good or bad; so can outcomes. There is no guarantee that good processes will lead inevitably to good outcomes. In other words, good processes can still lead to bad outcomes; and just because an outcome is bad, it doesn’t mean that the decision was based on a poor process. Luck plays a part too,

Russell Investments // Risk management: A fiduciary’s guidebook // Investment governance: A pragmatic update

By: D. Don Ezra

January 2010

Periodically it is worthwhile to remind ourselves of the principles, remember the lessons we have forgotten, and review the latest thinking about improving the processes involved in making decisions, so that those decisions can themselves be better.

/ p 103Russell Investments // Risk management: A fiduciary’s guidebook // Investment governance: A pragmatic update

because in investment we are typically placing a bet on the future—and not just on the future, but also on how human beings will see the future. That is difficult to predict at all, and impossible to predict precisely; hence the chance of error, no matter how careful and sensible the process of assembling inputs and making the decision.

Nevertheless, it’s common sense that a better process has a better chance of leading to a good outcome than a poorly designed process.

Some studies suggest that good governance is worth money: for example, Ambachtsheer and Ezra (1998) and Ambachtsheer (2006). But in general these are small studies, and limited to clearly defined outcomes, such as the excess return from active management. It is much more difficult to generalize broadly. Indeed, one of the gurus on the behavior of boards of directors says that it is fruitless to look for academic proof of the value of good governance. Ram Charan, in Boards that Deliver, states: “No matter how sophisticated the math, such research misses how directors actually interact, work together, and contribute.” In other words, behavioral factors are difficult to quantify; and the variables that are relatively easy to isolate and define (such as “is the chairman also the CEO?”) lead only to correlations, not to a clear connection between cause and effect.

As we will see, improving the process of making decisions is still an art, not a science, even though we can enunciate principles.

Section Two: Lessons remind us that the eternal verities don’t change Studies of investment governance began in 1992. Within ten years, most of the “big picture” problems had been identified. Let’s take a look at four of the lessons, before we look at the principles that underlie the search for better processes. Obviously we will grossly simplify the studies and lessons involved, in order to make the points clearly. Interested readers can pursue the references at the end of this paper.

LESSON #1: THE MOTIVATION TO AVOID BLAME IS PERVASIVE In 1992 social anthropologists O’Barr and Conley published a book called, provocatively, Fortune and Folly. After studying nine large US pension funds for two years, they concluded that the most pervasive cultural theme was the need to manage and deflect blame. They struck a nerve. The community said that anthropologists don’t know enough about the intricacies of pension fund management to make any judgments about its quality; but they admitted that their decision-making processes left something to be desired.

As we will see, decision matrices today still often lack clarity.

LESSON #2: DECISION-MAKING PROCESSES ARE OFTEN POOR Two years later this was confirmed in a survey of fifty of North America’s most senior pension fund managers. When asked—to their surprise: they thought they had come to a symposium on excellence in pension fund management—their spontaneous responses were telling. Listing barriers to excellence, 98% cited “poor process”, referring to decision structures, poor communication and inertia. No other factor reached the 50% mark.

The issue has not gone away. As recently as November 2, 2009, an editorial in Pensions & Investments, entitled “Not too big to fail”, cited this very study in discussing pay-to-play and similar governance-related failures in very large public sector pension funds, concluding that “better governance would help avoid such problems”.

LESSON #3: BACKGROUND TRAINING IS OFTEN INADEQUATE In 2001 the Myners Review of Institutional Investment in the UK quantified the qualifications of trustees of the UK’s pension funds. Some had been selected because

Nevertheless, it’s common sense that a better process has a better chance of leading to a good outcome than a poorly designed process.


they were outstanding in other fields, some to represent constituencies, but not surprisingly it turned out that few had any background training for the roles they were expected to play.

Myners pointed out two things. One is that decisions should be taken only by persons or organizations with the skills, information and resources necessary to take them effectively. The second is more subtle: that decision-makers do not need to be experts; rather, they need to know enough to be able to assess the advice they are given by experts, and to have the experience and skill to mount a sophisticated challenge to the advice received. This notion is that trustees should be aware of what they need to know, and that what they need to know does not amount to being an investment expert.

The lesson has not yet been broadly learned. Boards of trustees often feel that they should rubber-stamp recommendations made to them by experts.

LESSON #4: RESOURCES ARE OFTEN INADEQUATE The 1994 survey mentioned earlier cited, as its next most important barrier to excellence, inadequate resources. A Dutch study (Risseeuw and ter Wengel, 2001) took this a step further.

The academics involved divided Dutch pension funds into three sizes: large, medium and small. They measured the ambitions of the funds by the funds’ stated objectives and by the complexity of their investment structures. And they measured the resources devoted to the decision-making operations of the funds. By and large, they found that the large funds had high ambitions and large resources, and the small funds had low ambitions and small resources. But the mid-sized funds had a mismatch: they had high ambitions, like the large funds, but devoted small resources to meeting them, like the small funds. The researchers emphasized that the mismatch was crucial, and that it was mid-sized funds in which the gap between ambitions and resources showed up most noticeably.

That the resourcing angle is of fundamental importance shows up, for example, in Clark and Urwin (2008), in which the authors use the dedicated resources as their measure of the governance of an institutional investment fund.

Section Three: Some principles for a good governance structure These lessons will all find an echo as we look at the principles of good governance for institutional investment funds.

For a start, we observe that all enterprises have the same basic structure, so there is no need to reinvent the wheel when we are thinking purely of institutional investment. But there is no “one size fits all” solution to designing a governance matrix for a particular enterprise. What is necessary is to apply principles to a specific situation. And then periodically reassess the structure and its effectiveness.

THREE ROLES IN ANY ENTERPRISE No matter what the enterprise, it is easy to identify three different types of roles: to govern, to manage and to do. The discussion here expands on the outline in Ilkiw (1996).

To govern means to enunciate the mission, create a plan, and ultimately review progress from time to time. Typically this is initially performed by the founder. As an enterprise grows, it may be performed by an expanded group. And in very large enterprises, there is usually a board of directors elected or appointed to take on these tasks. In the field of investing, the role of governing can be played by an individual or a family, when considering personal investments; in institutional investing, involving trusts or life insurance companies, for example, a board takes on this role. It does not require day-to-day attention.


To do is obvious: it is to carry out the work of the enterprise, that is, to create the products or services embodied in the mission, every day. In a small business, this might also be done by the founder, or by a few employees.

As businesses grow the gap between the founder (or the board) and the employees grows. That creates the need to manage, that is, to have a person or a layer of people (or multiple layers!) available every day for the intermediate role of hiring and organizing the workers in a way that facilitates the achievement of the mission, creating the environment for them to work efficiently, and smoothing away the inevitable day-to-day frictions that arise.

In an investment organization, the end product consists of investment decisions. This can in principle be done by anyone, but bearing Myners in mind, it is invariably done by investment professionals. The professionals need to be hired (and possibly fired) and organized into a coherent structure; this is the intermediate role of “management”.

Of course it is possible that one person or group plays more than one role. That may work well in a given situation. But it is more useful to think of the three roles as different, and to think of people as wearing multiple hats when they play multiple roles, rather than simply to think of the people involved, because the latter creates confusion in job definition. This will become more apparent when we look at an example of an investment governance matrix.

And, as we will see, it is important to recognize that in any enterprise—whether it is a corporation, a charitable foundation, a community trust, an investment group, or whatever—the board works at the governing level.

AN EXAMPLE HELPS TO FIX THE IDEAS Rather than dealing in abstractions, let’s use an example of a governance matrix to illustrate the principles involved in its structure.

Exhibit 1, which is based on Ilkiw (1998), shows a commonly accepted sequential list of types of issues that characterize an investment fund, and the natural level of fiduciary responsibility associated with each issue.

Exhibit 1: The asset management decision process

Governing fiduciaries Objective setting Asset allocation policy investment strategy

Managing fiduciaries Manager structure Manager selection

Operating fiduciaries Investment decisions

Based on the earlier discussion, it will be apparent that the governing fiduciaries set the objective (decide the mission, in effect), select the basic asset allocation policy (essentially defining the enterprise’s risk tolerance) and declare, in their investment strategy, the beliefs that will underlie the types of investment decisions that constitute the output of this investment enterprise (for example, why they explicitly select active or passive exposure to some asset class, or deliberately over-weight or under-weight some sector). These are calendar-time functions.

The operating fiduciaries actually make the investment decisions. This is a real-time function.


The managing fiduciaries hire the operating fiduciaries. They select them as being fit for the purpose of making the relevant investment decisions, and structure them in such a way that, when the full complement of operating fiduciaries has been hired, the investment decisions in the aggregate reflect both the asset allocation policy and the underlying beliefs enunciated by the governing fiduciaries. As we will see, this responsibility also requires an oversight and evaluation of the operating fiduciaries, so the managing fiduciaries carry out real-time functions.

Now to an example of an actual governance matrix, a portion of which is shown in Exhibit 2 (the exhibit does not show the fund’s permanent staff which perform some of the investment functions). Readers who work with institutional funds will recognize many of its elements.

Exhibit 2: Example of a governance matrix

Vertically, along the left hand side, are the parties involved. They are given the right to make decisions, so they are fiduciaries. They are governing fiduciaries, managing fiduciaries or operating fiduciaries, depending on which of the three types of role they play. Each party can play different roles for different types of decisions.

Horizontally, across the top, are some of the types of investment-oriented decisions this enterprise has decided to tackle. And each column potentially shows three crucial aspects of the decision process: who provides input, who makes the decision, and who oversees and evaluates its outcome.

PRINCIPLE #1: EACH TYPE OF DECISION SHOULD HAVE CLARITY AS TO WHERE THE THREE WORDS “INPUTS”, “DECIDES” AND “OVERSEES” GO (In this example the precise form of input is specified.)

A corollary to this principle is that confusion results if “decides” occurs more than once, because shared decision making creates an environment for deflection of blame: remember O’Barr and Conley. Similar confusion results if “decides” does not occur at all—and that happens frequently, typically with words like “recommends” and “approves” being used. These words do not automatically imply accountability for, or ownership of, the decision, so they lead easily to deflection of blame when things go wrong, as in investing they are bound to from time to time. If for some reason the word “decides” is frowned upon, there are ways around it. “Recommends with enthusiasm” or “approves with enthusiasm” (but not both in the same column!) will imply accountability, when the time comes.

Setting Strategy Implementing Operating Educating

Decision

Fiduciary Group

Asset allocation policy

Rebalancing Manager selection, structure and benchmarks

Portfolio holdings

Day-to-day operations

Performance reports

Investment committee education

Investment committee

Decides and oversees

Establishes ranges

Oversees Oversees Reviews Participates

External consultant

Prepares studies

Researches and recommends

Oversees Provides some input

Provides some input

Group of investment managers

Decides Provides some input


In this example, relating to a corporate defined benefit pension fund, the corporation’s board of directors has appointed an investment committee (IC) to be its “named fiduciary” for the purposes of the law; in other words, the IC is essentially the board of directors of this pension fund.

Having set an objective (related to assets exceeding liabilities, and consistent with the sponsor’s funding policy), they call for studies of how different feasible asset allocation policies help to achieve the objective, and simultaneously expose them to the risk of failure. In this case the studies are prepared by external consultants, who work closely with the fund’s permanent staff—yes, you can guess this is a large fund, to be able to afford permanent internal staff—to prepare a recommendation for the IC. The IC doesn’t simply rubber-stamp the recommendation. They discuss it and its range of possible outcomes, compare it with other feasible policies, challenge the consultant on the basis for the underlying projections, and so on. The idea is that, when the IC selects an asset allocation policy, it clearly owns the decision. In this case, it also knows that it needs to oversee how the decision is carried out and evaluate whether or not it helps keep them on track to achieve their mission.

Here the IC evidently believes (this should be a belief that is explicitly stated) that from time to time the risk/reward characteristics of different asset classes depart from what they were when the policy was selected, giving rise to opportunities to succeed in doing better than expected. They have therefore permitted the staff to decide to depart from the basic policy, within established ranges. Note, however, that they have failed to set up a mechanism for evaluating whether this leeway adds to or subtracts from the fund’s value.

The asset allocation policy is implemented by selecting a number of investment managers: here, external rather than internal managers.

They are selected, in this case, by the staff, on the basis of research and advice from an external consultant. The consultant is in day-to-day touch with the investment community and has an up-to-date set of files on the capabilities and availabilities of candidate managers. The staff doesn’t have this information, but they are sufficiently familiar with the workings of the investment management community to be able to interview the candidates, after being briefed about them, and to select them. In addition, the candidates are so selected that they reflect the beliefs enunciated by the IC, and the normal portfolios they are expected to hold will add up, in the aggregate, to the selected asset allocation policy—that’s the structural element mentioned in the matrix.

The consultant monitors the actual portfolios, as the staff doesn’t have the budget to locate this capability in-house. If there are departures from the mandates given to the managers, or other concerns arise, the consultant reports to the staff, who review whether to retain or replace the manager.

Clearly these are real-time functions1, and the IC does not get involved. The IC does, however, oversee and evaluate the results of the staff’s decisions.

This example describes a structure as it existed two years ago. Today, after the global financial crisis, there would probably be an additional column for the function of risk management. Fiduciaries have learned that the risk environment changes; it does not stay constant from one asset allocation study to the next. Monitoring current conditions, and being prepared to take action (whatever that action may be—typically it would involve reducing risk exposure) is now seen to be an extremely important

1 Aficionados may note that “real time” here has two subtly different implications. The consultant’s monitoring is almost continuous: possibly daily, certainly very frequently. Staff will get involved as required, to consider recommendations made by the consultant; because the need arises spontaneously and should be dealt with promptly, the staff too are considered to perform a real-time function, even though this function is only required occasionally.


function. It would probably have a decision structure similar to that for asset allocation and rebalancing policy, with the governing fiduciary setting policy and accountable for oversight, and the managing fiduciary responsible for implementation.

This governance structure is based on Myners’ conclusion that the best process is to delegate decisions to those best qualified to make them. And in determining who is actually best qualified for a particular type of decision, the principle can be explained as follows:

PRINCIPLE #2: FOR EACH DECISION, ASK THREE QUESTIONS1. What are the knowledge requirements?

2. What are the time requirements? In turn, the time requirements are of two types:

• How much time is required?

• Can this decision be made in calendar time, or is it a real-time decision? (In other words, does it require instant action, or can it be deferred until the next scheduled meeting of the decision makers?)

3. Do we have someone, or a group, that meets these qualifications? If not, depending on the culture of the enterprise, filling the spot offers a further choice:

• Should we hire someone?

• Should we outsource the function?

It is clear that the answers will vary with the composition of the existing fiduciaries and the culture of the enterprise.

The final three columns in this example show how this IC looked at other issues, but we won’t comment on them because they don’t add to the two fundamental principles outlined in this section.

Section Four: Where do many funds fall short of best practice? In this section we deal with four broad ways in which funds tend to fall short of best practice: governance slippage, poor processes, over-elaborate oversight reports, and wasted time at board meetings. Each aspect itself is multi-faceted.

BROAD CATEGORY #1: GOVERNANCE SLIPPAGE Two particular aspects show up repeatedly in practice.

The first is that the managing fiduciary role isn’t properly clarified. All too often a board over-reaches and attempts to play this role as well as its own governing fiduciary role, not recognizing that the managing fiduciary’s tasks require real-time commitment and can’t be done on a calendar time basis—or at any rate that it’s simply bad process to confine it to calendar time exposure. Another occasion when a board sometimes overreaches into managing fiduciary territory is when investment managers are selected, and the board gets briefed by the external consultant and interviews and selects the investment managers. Very often the board members haven’t enough background exposure to know what to look for in a candidate: here inadequate knowledge is the issue. They aren’t aware that a quick briefing followed by a candidate’s well-honed presentation isn’t enough for them to ask the sophisticated questions that properly challenge the candidate and help the managing fiduciary to distinguish among candidates on criteria other than appearance, track record and the polish of the presentation.


Solutions to this aspect of the problem are to use a selection sub-committee that does have the knowledge, or to delegate the task to an internal or external group (typically staff or a consultant) that has both the time and the knowledge required.

The second aspect is that inadequate resources are devoted to governance as a whole. As we have seen, this is particularly likely to happen with mid-sized organizations, where there isn’t enough money available to hire someone internally. A solution that has sprung up is variously called “implemented consulting” or “fiduciary management”, in which an external organization is hired to do the job, the idea being that this external organization is dedicated to doing this full-time (satisfying the time and knowledge requirements), and its scale enables it to charge something less than the cost of hiring one’s own full-time staff.

BROAD CATEGORY #2: POOR PROCESSES Three particular aspects show up repeatedly in practice.

The first is that there is no explicit statement of investment beliefs to guide plans and actions. The idea is that beliefs, policies and actions should be consistent with each other, and complete. It is typically a fruitful exercise to look at what is done (the actions), and infer the beliefs that might give rise to those actions. For example, to justify active management in an asset class requires the twin beliefs that there are superior managers and that a process has been put in place to identify them. Again, seeing wide ranges in the asset allocation policy statement, or observing wide differences in the exposure of an asset class from time to time, might cause one to infer that there is skill in timing asset class exposures and that that skill has been brought to bear on the case in question. Or a policy of over-exposure to domestic rather than foreign equities might suggest that domestic equities as an asset class have some quality (which ought to be explicitly identified) of greater significance to this fund than to funds elsewhere. The beliefs typically aren’t capable of being proved right or wrong; but good process requires them to be defensible, as well as explicit. Check with the governing fiduciaries whether they have actually considered the implied beliefs; and when finally beliefs and policies and actions are in sync, test (over time) whether the beliefs are turning out to be justified.

The second aspect is that often manager assignments don’t follow fit-for-purpose criteria. That’s a mouthful! What it means is that managers have been selected for a role that they might not have been given, if explicit criteria for that role had been identified in advance and applied. This can happen, for example, when a small fund finds a manager who is good at one thing and gives the manager broad discretion to make all sorts of other decisions too: perhaps an equity specialist being given a balanced fund mandate, for convenience—not a reason that meets best practice. Another rationale for such a structure might be that it’s cheaper to do so—in which case it might indeed be defensible, if the cost of specialization exceeds the benefit. Either way, the process needs to be defensible.

The third aspect can be called, in general, implementation slippage. Decisions are made but are not, or are sloppily, implemented. Or the beautiful consistency across beliefs, policies and actions that existed the day after the structure was examined fades as time passes, and it’s never re-examined. The obvious solution is a periodic review of the governance process and how it is being implemented: a procedural audit.

BROAD CATEGORY #3: OVER-ELABORATE OVERSIGHT REPORTS When an oversight report runs to dozens of pages filled with numbers and bullet points, it constitutes a data dump rather than information. All too often, this is what overseers receive. Indeed, this is a prime reason why boards over-reach and do not delegate.

It’s worth a tangential diversion to explain two reasons why delegating is so difficult in practice. One reason is that boards haven’t thought about the difference between


ultimate responsibility and immediate responsibility. Since they are ultimately responsible for everything, and they can’t delegate that responsibility away, they may feel that they are responsible for actually carrying out everything by themselves. But in fact they can delegate the immediate responsibility to carry out a task, while retaining the ultimate responsibility. Another way of putting it is that they can retain the ultimate responsibility while delegating the authority to carry out a task. A second reason for withholding delegation is a simple reluctance to delegate, because they feel they have lost control: they no longer know what’s going on. And that feeling of loss of control is likely to be reinforced by a meaningless data dump instead of a carefully crafted and customized oversight report.

Those with the delegated authority to carry out a task typically have much more detailed knowledge (because they need much more detailed knowledge) about their area than the overseer does. That’s how it should be. The problem arises when decision-makers (with pride, and therefore for all the right reasons) present the overseer with all the details that they themselves consider. An overseer shouldn’t have to notice the tenth number on the seventh page to find out that something is going wrong: the fact ought to be presented in a way that jumps out. So it’s worth taking the time to custom-design an oversight report, because not only does it save the overseer’s time, it also helps the overseer to delegate with confidence.

It’s an art to design such a report, and it needs to be customized. Here we outline the broad contents and the characteristics of a good oversight report.

First, the contents. What does the overseer need to know? Four things, essentially:

• (Looking forward in time) Are we on track to achieve our mission?

• (Looking at today) What is our current stance?

• (Looking backward) Where have we created or lost value?

• Are we being prudent?

It should be possible to answer these questions on a sufficiently small amount of paper that the report can be comfortably held in a purse or a jacket pocket. At any rate, that’s the ideal.

The characteristics of the report are:

• It is brief. That means it focuses on the big picture. Yes, it should also be possible for the overseer to drill down; but where the drill goes isn’t part of the main report. And drilling down is most likely to take place in areas that are going wrong; all the details of areas that are going right just constitute clutter.

• It sets out the current situation. Very often this means it’s an exception report. If the overseer knows what the policy is, and only a few things are away from policy, it’s only those few things that need to be reported.

• It helps an evaluation of whether the delegated task is being done well. The traffic light protocol (red, yellow, green) can be very helpful. It’s the yellow stuff that needs a drilling-down and a discussion; and the overseer should probably have been alerted to the red stuff before the calendar-time meeting.

• It is available promptly and regularly. Ah, there’s the rub! The greater the degree of customization, the more difficult it is to produce the report promptly. We have no universal solution to this problem!


BROAD CATEGORY #4: WASTED TIME AT BOARD MEETINGS When a board over-reaches, or isn’t sure what it needs to do, it typically fills its time with activity. But the activity doesn’t help its decisions or its oversight unless it is specifically geared to that purpose.

In fact, it ought to be easy to decide how to structure the agenda for a meeting at any fiduciary level. Look across the row describing the responsibilities of that fiduciary level. Does the word “inputs” (or its equivalent) occur somewhere? If so, place on the agenda any decision for which input needs to be created, and arrange for the assembly and discussion of materials leading to the input. Does the word “decides” occur anywhere? If so, place on the agenda the issue to be decided, and arrange for the delivery and discussion of all the inputs. Does the word “oversees” occur somewhere? If so, place on the agenda any aspect that needs oversight, and arrange for the delivery and discussion of the relevant oversight report. Anything else is overkill and probably a waste of time.

Let’s apply this, in particular, to board meetings. Most of the time all that is necessary is a calendar-time focus on the oversight of the big picture. If there are yellow or red lights attached to something, time for an in-depth discussion should be added; otherwise not. Periodically (once a year, perhaps) spend time on an in-depth review of everything, even the things that are going right; it isn’t necessary to do all of that in a single meeting—the in-depth reviews can be spread over the year. And one of the aspects that should be subject to an in-depth review is the governance process itself.

In a board self-assessment, here are some items that can be useful:

• Ask each board member to identify areas in which they feel they have knowledge and skills, and areas in which they feel education would particularly help them.

• Ask them to make “exception report” comments (“particularly good” or “can be improved”) on their leader, the other board members, the staff, regarding the composition of each group, their effectiveness, the team feeling, the way meetings are conducted, and the reports received.

• Ask them to comment on the external service providers.

• Ask them how they feel about their own contributions: whether they feel valued, whether there are aspects in which they feel under-utilized.

Depending on the interaction and the culture, these statements by the board members might go to their chair or to an external party to assemble them and shape them into a form that makes for a useful discussion. There’s no right or wrong way that fits all circumstances: the intricacies of personal interaction become all-important.

Section Five: A recent case study—investment governance in UK pension schemes We asserted earlier that the same issues keep recurring, over and over again, around the world. In November 2009 a survey of the investment decision-making practices of UK defined benefit pension schemes (Kelly-Scholte and Kothare, 2009) was published, dramatically illustrating our assertion.

Here are some of the findings reported in the study, conducted for Russell by IFF Research, one of the largest independent research companies in the UK:

• Lack of clear accountability: Approximately one in three respondents believe that multiple groups are responsible for each of the different investment decisions. For each of several types of decisions (investment strategy, liability-driven investing, portfolio structure, active versus passive, manager selection), the average number of groups perceived to be responsible exceeded one.


• Insufficient delegation: The bulk of decision making on each of the investment issues cited continues to be retained at the level of the board of trustees.

• Need for greater education: 75% of respondents have no knowledge of the concept of fiduciary management (outsourced decision making at the level of managing fiduciaries).

• Too much board time on investment matters: Trustees of schemes that have an investment committee (IC) spend much the same proportion of their time on investment matters as trustees of schemes that don’t have an IC. That suggests time poorly spent, with efforts unnecessarily duplicated. Quite simply, if you have an IC, a full board should spend less time on these matters than if you don’t have an IC.

• Over-confidence in the governance structure: Despite evidence of non-conformity with agreed principles of best practice, 86% of respondents are confident that the scheme’s current decision-making processes are suitable to meet future challenges. However, 40% are not confident in their ability to respond quickly to new situations.

And a final observation The thoughtful reader will have noticed the omission of a significant aspect of governance. We have dealt with the decision structure, but we have said nothing about how to improve the decision process itself. That is deliberate. The subject is far too extensive to deal with in this paper. It has to do with the role of the leader; the interaction of the members of each fiduciary level; how the independent views of each member are expressed and assembled and contribute to the ultimate decision; how to avoid the leader, or the most knowledgeable or vocal person on an issue, from effectively becoming a one-person sub-committee whose views are rubber-stamped; and so on.

Not nearly enough time is spent, either in business courses or at any other level, on studying decision making as a process that can be improved by the application of principles. In this paper we simply refer the interested reader to a recent book that explains many of these principles: the Mauboussin book cited in the references.

Related readingAmbachtsheer, Keith P. (2006), How much is good governance worth?, The Ambachtsheer Letter #245, June 2006.

Ambachtsheer, Keith P. and D. D. Don Ezra (1998), Pension Fund Excellence, John Wiley.

Charan, Ram (2005), Boards That Deliver, Jossey-Bass.

Clark, Gordon L. and Roger Urwin (2008), Leadership, Collective Decision-Making, and Pension Fund Governance. Available at SSRN: http://ssrn.com/abstract=1133015.

Ilkiw, John H (1996), Prudence is Process, not Performance, Russell Canada Monograph.

Ilkiw, John H (1998), Handbook on Asset-Liability Management: A Guide for U.S. Fiduciaries, Russell Investments.

Kelly-Scholte, Sorca and Shashank Kothare (2009), Mind the ‘Governance Gap’, Russell Investments.

Mauboussin, Michael (2009), Think Twice, Harvard Business Press.

Myners, Paul (2001), Institutional Investment in the United Kingdom: a Review, HM Treasury.

O’Barr, William M. and John M. Conley (1992), Fortune and Folly: the Wealth and Power of Institutional Investing, Business One Irwin.

Pensions & Investments, “Not too big to fail”, November 2, 2009, http://www.pionline.com.

Risseeuw, Peter A and Jan ter Wengel (2001), Fund Governance in the Netherlands, Vrije Universiteit, Amsterdam (written in Dutch).

/ p 113

Roadmap for fiduciary risk management: the importance of an effective investment management program

Frameworks make management easier. They provide an organizational structure and aid clear thinking. The framework that Russell has used for many years in consulting to large institutions is based on a decision flow process—commonly referred to as a fiduciary ladder—developed in the 1990s and described in detail in the following piece written in 1999. This structure has proved so durable that it remains in use, largely unchanged, more than a decade after it was first captured.

Russell Investments // Risk management: A fiduciary’s guidebook // Roadmap for fiduciary risk management

/ p 114

Roadmap for fiduciary risk management The importance of an effective investment management program

Two key risks pervade every aspect of an investment management program: the risk that good policies and procedures are not in place; and the risk that sound policies and procedures are ineffectively implemented.1

If policies and procedures are not effective and executed consistently, the consequences may be severe. Both individual fiduciaries and the institution may be subject to increased liability and embarrassing public disclosure. More likely, inefficient policies or poorly executed procedures may result in increased contributions and earnings charges, lower asset returns, or lower benefit payments.

Fortunately, by observing good practice across our industry, it is possible to draw a roadmap to guide us reliably to our goal of effective policy, procedures, and implementation.

Exhibit 1: Asset management decision flow process

Russell Investments // Risk management: A fiduciary’s guidebook // Roadmap for fiduciary risk management

By: Janine Baldridge

October 1999

Execution

Implement

Asset class strategy

Portfolio structure

Manager selection/monitoring

Asset allocation

Plan

Governance

Objective setting

Performance measurement and evaluationReview

1 Nathan Dudley, Chris Hensel, Jennie Tyndall, “Risk Management,” Viewpoint, February 1999.

/ p 115Russell Investments // Risk management: A fiduciary’s guidebook // Roadmap for fiduciary risk management

The decision levels defined The labels used to describe the levels of decision making depicted by Exhibit 1 are not completely self-explanatory. A brief definition of each is provided below:

Governance is not about running the funds; it’s about seeing that the fund is run properly. Its focus is to ensure that decision-making authority is sensibly assigned to the most appropriate person or group. It also involves establishing systems for compliance of legal and regulatory requirements, and assuring compliance with stated guidelines.

Objective setting follows governance and is perhaps the most important step in running an investment program. The objective of an overall investment program is the investment result the sponsor aims to achieve. The objective must be realistic, reconcile competing sub-objectives by defining suitable trade-offs, and distinguish the objective from various constraints. After the next three decision steps have been completed, objective setting must be extended downward from the overall program to the level of individual asset classes and funds.

Asset allocation determines what proportion of total assets will be allocated to each broadly defined class of assets—e.g., equities, bonds, or real estate. Driving the asset allocation decision are the objectives and constraints defined in the previous decision step, and forecasts of the behavior of the asset classes being considered.

Asset class strategy takes the generic allocation to a given asset class and defines in detail how the allocation will be invested across sectors within the class. The steps involved include analyzing the appropriateness of investment styles, biases and tilts versus benchmarks, determining active vs. passive management allocations, and establishing specific performance benchmarks and targets for asset and sub-asset classes (i.e., U.S. equity and style allocations).

Portfolio structure matches specific investment managers (previously identified as suitable) to the chosen asset class strategy. This step is critical in reviewing and changing asset class and manager structure.

Manager selection and monitoring involves research to identify the best available candidate investment advisors for the fund and then to determine which of the advisors identified are appropriate to hire, given the fund’s specific asset allocation and desired structure.

Execution focuses on the steps necessary to efficiently implement the investment program. This includes developing investment manager guidelines and implementing changes across and within asset classes (portfolio transitions, asset class rebalancing, vendor changes, etc.).

Performance measurement and evaluation closes the loop on the entire process. Measurement generally involves a comprehensive analysis of total fund, asset class and manager performance vs. benchmarks and peer groups. The evaluation process tests the measurements against their objectives, and identifies any changes that may be necessary.

REQUIRED ASSET MANAGEMENT DECISIONS Exhibit 1 illustrates the various levels of decisions required to run a fund. The figure organizes these decision levels vertically by the impact that each has on the ultimate return, and horizontally by the typical frequency with which the derivatives must be revisited2. A successful approach to achieving fund goals and objectives gives attention to each level of the decision process and manages the risks inherent to them.

2 John Ilkiw, “Handbook on Asset-Liability Management: A Guide for US Fiduciaries,” Research Commentary, December 1998.


LEVELS OF DECISION MAKING VS. CYCLE OF ACTIVITY The levels of decision making discussed above are an approximate, but not exact, mirror of the cycle of activity a sponsor should undertake to manage a fund. When initially setting up an investment program, the sponsor will probably choose to use the levels as a sequence of action steps, at least to the point of developing portfolio structure, where manager selection and the research it entails are prerequisite. In an ongoing investment program, daily activity iterates through the bottom levels at most times. Occasional performance or manager monitoring developments may kick activity up one or more levels, to portfolio structure or asset class strategy. Only infrequently will performance events or manager circumstances trigger a review of strategic asset allocation or more rarely, a reexamination of objectives. As a matter of discipline, the top three levels are probably best to address only at scheduled, but infrequent, intervals—but not in reaction to market events.

WHO’S RESPONSIBLE FOR MAKING IT HAPPEN? Having defined what steps need to take place to efficiently manage an investment process, the next logical question is, who should be responsible for the necessary work? There are different levels of fiduciaries within each organization, with differing levels of expertise and resources. These levels tend to be functional in nature, but are often distinct groups as well. The functions can be characterized as follows:

• Governing fiduciaries generally consist of the Board of Directors/Trustees who establish broad governance policies and objectives, and provide a top-level review of results and operations. Governing fiduciaries are ultimately accountable for achieving long-term objectives.

• Managing fiduciaries are generally the internal management group that are accountable for strategic decisions (asset allocation and investment structure).

• Operating-staff fiduciaries are members of the internal staff who are accountable for day-to-day responsibilities, investment expertise, investment manager recommendations and compliance.

• Operating-investment manager fiduciaries are external organizations or individuals to whom internal fiduciaries delegate responsibility for creation of portfolios that satisfy specific mandates and comply with policy guidelines. They execute the plan’s chosen strategies in the market.

In addition, consultants, actuaries and other external providers often have key roles advising on the management of the plan.

For many US corporations, the Board of Directors has delegated the role of named fiduciary (an ERISA requirement) to an internal management group. While the Board does retain full fiduciary responsibility, the internal group assumes both longer-term strategic responsibilities and the nearer term policy development role for asset allocation and investment structure. In these cases, the Governing and Managing fiduciary functions become intertwined.

A NOTE ON “POLICY” AND “STRATEGY” One’s first attempt to distinguish policy from strategy typically focuses on time frame, with policy’s time frame being very long term and strategy’s being somewhat shorter (but still longer than for “tactics”). On a moment’s reflection, it becomes apparent that this contradicts common industry use. We commonly refer to both “asset allocation strategy” and the “policy portfolio” that results from it. Asset class strategy defines manager structures, but the managers populating these structures receive “investment policy statements.”


Definitions of policy, strategy, and tactic that better fit the everyday use of the terms would be as follows:

• Decisions that are exogenously imposed on a fiduciary are policies. Coincidentally, policies often last for many activity cycles of those on whom they are imposed.

• Longer-term decisions controlled by a fiduciary are strategies. Strategies typically persist for at least one activity cycle—e.g., a market cycle.

• Short-term decisions controlled by a fiduciary are tactics. Tactics are variations in approach or activity that persist for less than a full activity cycle.

This means that strategies determined by one level of fiduciary may be regarded as policy by lower-level fiduciaries. Perhaps the only policy a governing fiduciary must follow is to meet its obligations as fiduciary and as a legal person subject to applicable regulations. From the governing fiduciary’s perspective, all other decisions are strategic. Conversely, a managing fiduciary would take the governing fiduciary’s decisions as policy, but regard asset class strategy and subsidiary decisions as strategic.

Keeping in mind this overview of the decision steps and who takes responsibility for them, we can turn to a more detailed discussion of each.

Fund governance

GOALSThe goal of a governance structure is to maximize the probability of meeting financial goals, and limit fiduciary risks. To achieve these goals, an effective governance structure establishes efficient decision-making processes, clearly defined fiduciary roles and responsibilities, and effective policies and procedures. Good decision-making processes improve the potential to increase investment returns, reduce opportunity and explicit costs, and help avoid classic investment mistakes (selling at the bottom, etc.).

ISSUESIn the US, ERISA’s “prudent expert rule” requires fiduciaries to decide carefully how and when to delegate authority to experts. It is absolutely essential to document a decision to delegate. Each level of fiduciary within and outside the organization needs to have clearly defined roles and responsibilities that are well documented, measurable and achievable within the available resources.

In addition, decisions should be made by those with the appropriate knowledge, accountability and resources. Exhibit 2 presents a fiduciary structure designed to promote effective decision-making practices.

Policies and procedures should be consistent with the fiduciary structure and established on the basis that they will be fully implemented and monitored, given available resources (internal and external).

Execution

Implement


Portfolio structure


Asset allocation

Plan

Governance

Objective setting



Exhibit 1: Decision-making structure example

Compliance reporting should also be consistent with fiduciary structure and resources, allowing each fiduciary level to affirm the success of delegation procedures.

TASKS/DELIVERABLES1. Make clear and explicit agreements at each fiduciary level on roles and

responsibilities:

• Determine who is responsible for oversight, approval, recommendations, and monitoring of each fund management element (Exhibit 1). Assure all levels are aware of decision-making accountabilities (Exhibit 2).

2. Have a written investment policy for the fund that is approved at the appropriate fiduciary levels. The policy should include:

• Fund and investment (asset class) objectives

• Asset allocation policy and ranges

• Rebalancing and cash management policies

• Compliance procedures

• Roles and responsibilities (including general delegation authority)

3. Develop written investment manager guidelines that define sponsor objectives, investment constraints, performance expectations, time horizon and delegated responsibilities.

4. Establish reporting procedures for each fiduciary level to affirm compliance activities and review relevant fund information.

5. Periodically evaluate compliance with procedures and policies.

Planning Implementation and review

Fiduciary level

PolicyFund objectives asset allocation internal/external

StrategyAsset class strategies and structure

ManagementManager meetings and implementation

ExecutionSecurity selection

ControlReview for compliance vs. objectives

“Governing” board

Decides Oversees Oversees Oversees Review investment committee decisions

“Managing” investment committee

Recommends Decides Oversees Oversees Review staff decisions

“Operating” investment staff

Recommends Recommends Decides Oversees Review manager decisions

“Operating” investment managers

Consulted Consulted Implements Compliance vs. guidelines


Objective setting

GOALSThe goal of formal objective setting is to think through and document what a fund must accomplish. It is more difficult than it sounds, as the stakeholders in a fund have several, occasionally competing, objectives for the fund. A formal statement of objectives reconciles or prioritizes the interests of stakeholders. This in turn provides the targets for a long-term investment strategy.

ISSUESThe ultimate goal of any investment program is to ensure that obligations to the beneficiaries are met. In service of this ultimate objective, reasonable goals may include maximizing total return, maximizing surplus, stabilizing surplus, minimizing contribution costs, or minimizing administrative complexity.

Which comes first? The objective-setting process forces all fiduciaries to focus, communicate and reach consensus on prioritizing key fund and investment objectives. Since objectives are often in conflict with each other, a process for fiduciaries to formally prioritize them provides continuity and discipline through difficult periods (or when fiduciaries change). This helps lower the probability for ad-hoc revisions to well-developed long-term policies and strategies.

Objectives should be developed with a view of both assets and liabilities, and focusing on assessing and tolerating shortfalls. In setting objectives, fiduciaries should consider the expected range of outcomes, and appropriate time horizon and risk tolerance. Policies should be developed that seek to achieve outcomes within acceptable ranges.

It is important to document only those policies that the fund intends to follow consistently, as there may be more risk associated from not following a written policy statement, than by having one that is less comprehensive.

Objectives should acknowledge and endorse regulatory mandates (ex: ERISA for qualified US funds) in the policy.

TASKS/DELIVERABLES1. Formalize key objectives with the appropriate fiduciary level.

• Seek to prioritize conflicting objectives and assign risk/return expectations and acceptable ranges of outcomes.

• Objectives tend to be broad in scope, and include investment, operational and fiduciary considerations.

• Governing fiduciaries should typically focus on funding policy (i.e., contribution timing, ERISA funding policies), and financial reporting (i.e., FASB impact, targeted funded status). Managing fiduciaries should consider fund and asset class investment structure and results relative to objectives. Operating fiduciaries should generally focus on objectives for implementing investment structure.

Execution

Implement


Portfolio structure


Asset allocation

Plan

Governance

Objective setting



2. Document key objectives in the Investment Policy Statement and include only objectives within the control of fund fiduciaries (i.e., investment, fiduciary and operational) and for the sole benefit of the participants.

3. Review/affirm objectives on a scheduled basis. Fund objectives should be evaluated versus benchmarks at least annually with the appropriate fiduciaries. Governing/managing fiduciaries typically tend to look at more summary funding and investment information on a quarterly basis or less. Operating fiduciaries tend to focus on detailed results on a more frequent basis.

Asset allocation

GOALSWhile objective setting establishes what is to be achieved, asset allocation establishes at the broadest level how objectives are to be achieved. Asset allocation encompasses two stages. The first stage asset allocation decision determines the broad asset classes a fund will invest in, and in what proportions, in order to achieve fund objectives. Asset allocation policy incorporates underlying liabilities, and embodies fund-specific risk tolerance and return expectations, and is owned by the key fiduciary decision-makers. The second stage determines, within an asset class, the major sectors in which to invest and the normal proportion of funds allocated to the asset class that will be invested in each selected sector.

ISSUESAsset allocation policy is based on evaluating the impact of various asset mixes relative to fund objectives. Governing/managing fiduciaries must agree on key elements including acceptable asset classes, risk and return expectations, and time horizon, in order to ultimately “own” the decision and limit the potential for making ad hoc changes in light of negative near-term results. Key to this process is assessing the relationship of asset mixes versus the underlying liabilities and objectives for the fund.

Stage 1 asset allocation generally considers core asset classes (US equities, non-US equities, fixed income, real estate and cash) as they are conceptually distinct and have the largest impact to long-term fund performance. The primary expression of risk tolerance is the policy percentage allocated to equities. Thus, it is critical for governing/managing fiduciaries to be able to clearly articulate the rationale behind the strategic asset mix policy relative to objectives.

Stage 2 asset allocation considers the amount of funds to invest across distinguishable sectors within asset classes. Forecasts of sector performance tend to be less reliable than forecasts of broad asset class return distributions; therefore, stage 2 asset allocation tends to be heuristic rather than model-driven, and takes into account such issues as liquidity, and materiality of potential impact versus administrative costs.

Rebalancing policies and ranges create a disciplined approach for adhering to long-term strategies, avoiding near term market timing and allowing operating fiduciaries to implement policies in a timely manner.

Execution

Implement


Portfolio structure


Asset allocation

Plan

Governance

Objective setting



TASKS/DELIVERABLES1. Formalize the asset allocation policy and acceptable ranges with the appropriate

fiduciaries (governing/managing fiduciaries).

• Complete an asset/liability study every three–five years, or more frequently with material changes to plan/liability structure or fiduciaries.

• Reach consensus agreement on risk and return expectations, acceptable time horizon for judging success, and acceptable asset class exposure.

• Understand potential range of outcomes given capital market and liability expectations.

• Review/affirm on a periodic basis.

2. Formalize objectives for each asset class. Adopt long-run return expectations for each asset class.

3. Establish effective rebalancing procedures to adhere to long-term asset mix policies.

4. Document asset mix policy, asset class objectives, return expectations, rebalancing procedures and acceptable tracking error in Investment Policy Statement.

Asset class strategy and portfolio structure

GOALSThe process of devising asset class strategies and associated portfolio structures moves the fiduciary from planning to implementation. The fundamental decision of asset class strategy is whether to make active bets or to remain passive.

Active bets include tilts away from the policy portfolio, whether implemented by the managing fiduciary or delegated to an operating fiduciary, and the choice to use active managers. The goal at this stage is to establish an investment structure within and across asset classes that is at least consistent with fund and investment objectives, including risk and return expectations, and at best creates a meaningful possibility of incremental returns.

Portfolio structure takes the roles defined by asset class strategy and matches them to the names of the best available managers.

ISSUESAsset class strategy may include determining the optimal number of managers, active/passive allocation, alpha expectations for active strategies, tilts against benchmarks and review points for judging success. It is critical to align the ultimate structure with available resources to ensure effective implementation of monitoring and compliance obligations. Where resources (both internal and external) are limited, passive strategies may be more appropriate than multiple active products.

The proper context for sub-asset class bets (style tilts, private equities, and high yield bonds) is the investment structure within each asset class. While governing/managing

Execution

Implement


Portfolio structure


Asset allocation

Plan

Governance

Objective setting



fiduciaries should affirm and understand material structural bets from the benchmarks, operating fiduciaries are responsible for recommending and implementing these bets.

Fiduciaries should understand and agree on the range of anticipated outcomes with the investment structure. Understanding the potential downside of a strategy helps minimize the probability of frequent strategy revisions that increase transaction costs and erode returns.

The asset class structure decision specifies the roles to be played by external operating fiduciaries. Portfolio structure matches the roles with specific managers. There may not be suitable managers to fill all roles defined by asset class strategy. Asset class strategy and portfolio structure thus strongly influence each other; several iterations between the two may be needed to create a suitable asset class structure populated with appropriate managers.

TASKS/DELIVERABLES1. Affirm structural bets within or across asset classes with operating fiduciaries:

• Ensure consistency with investment objectives, risk tolerance and time horizon policies.

• Identify expected tracking error expectations and minimum return hurdles for asset classes based on fund-specific structural objectives.

• Establish clear and measurable benchmarks with the appropriate time horizon for each asset class.

2. Document asset class objectives, including desired tilts and benchmark biases, in the Investment Policy Statement. Review annually to ensure continued applicability.

3. Monitor the structure of the aggregate asset class portfolio relative to objectives, focusing on the key risk drivers within each asset class. It is important to control the undesired structural bets. Examples include:

• US equities: capitalization, valuation characteristics, economic sectors, cash.

• Non-US equities: country, currency, capitalization, valuation characteristics, economic sectors, cash.

• Fixed income: interest rate and yield curve sensitivity, economic sectors, country, currency.

• Real estate: geographical location, industries, property type, private/public markets.

4. Establish formal review procedures to evaluate strategies falling below minimum expectations.

Manager selection and monitoring

GOALSManager selection and monitoring identifies the opportunity set of investment managers that might play a role in the chosen portfolio structure. The desired result is to ensure

Execution

Implement


Portfolio structure


Asset allocation

Plan

Governance

Objective setting



the fund utilizes appropriate investment advisors and products likely to achieve stated investment objectives (minimum alpha hurdles, risk controls, etc.).

ISSUESAsset class strategy and portfolio structure presuppose broad familiarity with styles of management and key managers providing it. Staff operating fiduciaries should develop selection criteria in advance of the search process, formalize the criteria as firms/products are hired, and review the criteria on an ongoing basis to ensure continued compliance with the hiring rationale. They should assess, verify and document deviations from original objectives.

Manager selection should be based on compatibility with fund objectives, return and risk expectations and structure. The fund’s time horizon for judging success should be consistent with investment product expectations and processes.

Benchmarks must be clearly stated and compatible with both fund and manager expectations (example: avoid comparing mid/small cap products with large cap benchmarks over short time periods).

Investment manager guidelines form the basis for communicating sponsor expectations for returns relative to benchmarks, diversification, constraints and fiduciary practices. Guidelines should also be considered a “living” document, with periodic discussions of the continued applicability for both the fund sponsor and manager.

TASKS/DELIVERABLES1. Hire/retain managers on the basis of the sponsor’s high confidence that they will

meet stated objectives. Ensure internal and external resources are consistent with investment and manager structure. In other words, the number and types of products or firms employed should be based on ability of sponsor to maintain necessary expertise and monitoring capabilities.

2. Document sponsor expectations for firm/product in Investment Manager Guidelines (and Investment Manager Contract). The documentation should include minimum risk/return expectations, investment constraints, eligible securities and strategies, and delegation responsibilities, and reviewed annually with the manager.

3. Monitor firms/products through direct meetings, performance analysis, structural evaluation and external reports (consultants, custodian). Process should include:

• Assess performance versus stated benchmarks.

• Assure compliance versus investment guidelines, including adherence to delegated duties (i.e., proxy voting, performance reporting).

• Review consultant analysis of investment products and firms.

• Focus on critical changes in the organization: people, process, portfolio characteristics, and performance.

• Utilize measurement and analytical tools to evaluate strategies.

4. Establish formal processes to review underperforming strategies relative to guidelines.


Execution

GOALSThe goal of all tasks that collectively comprise execution is to ensure that previously-established policies and strategies are implemented completely and cost-effectively.

ISSUESImportant aspects of execution include portfolio transitions, cash management, trustee/custodian services, and investment manager guidelines. It is critical to establish effective procedures for managing these activities in order to limit explicit and opportunity costs.

Effective execution is the front line of defense against operational and administrative risk—the risk that returns will be diluted by sluggish implementation or more dramatically, that misdirection or theft of assets will not be quickly detected because of flawed procedures or lax implementation. Success in warding off such dangers results from a grinding focus on detail and relentless discipline.

Implementation activities generally increase expenses to the plan because they generate additional trading, custody and opportunity costs. The sponsor must ensure that the innumerable implementation steps they take are prudent and further the best interests of the plan participants.

Procedures for managing portfolio transitions have changed significantly over the last few years as markets and systems have become more efficient. Sponsors should pay careful attention to minimizing uninvested positions as old portfolios are liquidated and new manager’s portfolios established. Investment vehicles such as futures contracts or daily index commingled funds permit full investment in appropriate asset classes and sectors, even during a transition period. This is especially critical in the more volatile asset classes and countries. Global crossing networks minimize market impact and lower the cost of trading. Transition specialists effectively limit trading errors and provide external due diligence and fiduciary support.

TASKS/DELIVERABLES1. Develop investment manager guidelines that incorporate fund objectives, establish

prudent limits, and provide a communication protocol to help the sponsor and manager to verify compliance and evaluate proposed modifications. Negotiate with managers and jointly implement procedures for guideline compliance verification.

2. Develop effective transition strategies for managing portfolio changes and rebalancing activities. Transition plans should include minimizing implicit and explicit trading costs, remaining fully invested relative to the asset class policy (i.e., using futures to equitize cash), minimizing trading errors, and managing all vendor activities (i.e., trustee/custodian, managers, transition specialists).

3. Establish compliance procedures to evaluate trustee/custodian activities including securities management, cash management, securities lending, participant record keeping, proxy voting and benefit payment procedures.

Execution

Implement


Portfolio structure


Asset allocation

Plan

Governance

Objective setting



Performance measurement and evaluation

GOALSMeasure performance of the actual fund, asset class and investment product returns against benchmarks, allowing fiduciaries to evaluate results against expectations and make adjustments as necessary.

ISSUESThe process should lead to early identification of problems with investment structure, products and/or firms, and should provide a framework for quickly resolving unanticipated issues. It should also affirm compliance with objectives and fulfill fiduciary obligations.

Reporting should be targeted towards the needs of the different fiduciary levels to enable effective analysis and enhance the decision-making process. Reporting might address the following key questions:

Governing fiduciary

• Are assets managed in compliance with policies and procedures?

• Are agents in compliance with their mandates?

Managing fiduciary

• Are goals and objectives sensible from the perspective of benefit security?

• Do due diligence monitoring processes ensure that decisions made in the past are still appropriate today?

Operating fiduciary

• Does the structure remain consistent with goals and objectives?

• Are due diligence procedures implemented properly?

• Should additional procedures be developed?

TASKS/DELIVERABLES1. Generate reports analyzing investment performance:

• Measure results of total fund, asset and sub-asset classes, and investment managers/products.

• Utilize multiple evaluation tools, such as investment returns (absolute and relative to benchmarks/peers), attribution, and profiles, to gain full understanding of results.

2. Review Investment Policy Statement for compliance to guidelines and policies, including: appropriate asset classes, diversified asset class structure, avoiding potential and actual conflict of interest, appropriate trustee/custodian asset valuations, and appropriate implementation of securities lending and proxy voting programs.

3. Review actual asset mix versus policy mix:

• Evaluate asset allocation slippage versus policy; monitor for compliance within acceptable ranges.

Execution

Implement


Portfolio structure


Asset allocation

Plan

Governance

Objective setting



• Reaffirm and document rationale for holding asset classes falling below minimum return hurdles.

4. Develop relevant reporting for each fiduciary level. Examples include:

Governing fiduciary

• Compare plan asset allocation and results to policies

• Review summary of compliance activities versus policies

Managing fiduciary

• Review results of near-term activities

• Reaffirm policies and decisions

• Review compliance activities versus policies

• Receive regular compliance reports

• Conduct a procedural audit

Operating fiduciary

• Review external vendor reporting versus guidelines

• Ensure implementation activities are properly completed by internal/external providers

5. Establish review sessions with each fiduciary level to ensure understanding of results and ongoing evaluation of goals and compliance activities. Generally, meeting frequency is one to two times annually for governing fiduciaries, three to five times per year for managing fiduciaries, and as needed for operating fiduciaries.

SummaryKey to an effective risk management program is the adoption of a comprehensive approach. The program should be customized relative to the fund’s specific objectives and resources to ensure successful implementation and results. A successful program can help increase the return potential of the fund while reducing fiduciary liability, both important objectives toward meeting the financial goals of the investment fund and its participants.

AcknowledgementsMany people have contributed to developing this fiduciary “framework,” particularly John Ilkiw and D. Don Ezra. The author would also like to thank Nathan Dudley, Chris Hensel, George Oberhofer, and Gloria Reeg for their insights and assistance with this paper.

/ p 127

Mind the ‘governance gap’

In 2009, Russell’s UK team conducted a survey of governance practices among DB plans. They gathered responses from more than 200 individuals involved in running investment programs at these plans, exploring such issues as the process and effectiveness of decision making, the constitution of trustee boards, delegation, and the allocation of time, among other subjects.

While they found some areas of strength, they also observed room for improvement in some regards, such as ineffective delegation to the appropriate level of capability and a lack of clear accountability.

The survey includes a section devoted to the question of fiduciary management (a structure in which a specialist third party is charged with designing portfolio structure, selecting money managers, monitoring and reporting on fund progress.) An appendix sets a number of principles of good fund governance, and some lessons that can be drawn from the wider field of corporate governance.

Russell Investments // Risk management: A fiduciary’s guidebook // Mind the ‘governance gap’

/ p 128

Mind the ‘governance gap’ A survey of the investment decision-making practices of UK DB pension schemes

Executive summary

BACKGROUND TO THE SURVEY Over the last decade the management of pension assets has become ever more sophisticated and demanding. Trustees are faced with increasingly challenging decisions on investment strategy, not least much greater complexity and a vastly wider array of choice in terms of both asset class configuration and product selection.

Accordingly, the range of investment decision-making structures available today has developed beyond the simple trustee/manager model prevalent in the UK 10–15 years ago. The choices available include the use of Investment Committees, internal CIOs or Executive Offices, and more recently, access to fiduciary managers.

Many schemes already incorporate some of these options, such as Investment Committees, into their governance structure. However, given the fast pace of change in the industry recently—with increased focus on liability-hedging and greater interest in alternatives and alpha-orientated approaches—we suspect that many trustees need to adapt their decision-making structures to help them address the new challenges more effectively. In particular, we believe that trustees now need to spend more time on strategy and less time on implementation, which can be effectively delegated, and there are strong arguments in support of having a more ‘real-time’ oversight beyond the traditional quarterly meeting cycle.

The cost of not adapting can be significant. According to some estimates, poor governance can cost pension schemes up to 3% pa1. While there is no perfect metric tying good processes and returns, the results from the academic studies fuel our doubts on the effectiveness of current decision-making structures.

Our suspicions were largely based on anecdotal evidence so we wanted to test whether our views bear scrutiny. There are many surveys on general pension fund governance, but there is a paucity of surveys that focus specifically on investment decision making. Therefore, we commissioned IFF Research (IFF) to look into current decision-making structures of UK defined benefit (DB) pension schemes.

The main aims of the survey were to:

• Identify the range of investment governance strategies currently being used by UK-defined benefit schemes;

• Compare these structures to best practice principles; and

• Understand awareness of and attitudes to fiduciary management.

Russell Investments // Risk management: A fiduciary’s guidebook // Mind the ‘governance gap’

By: Sorca Kelly-Scholte and Shashank Kothare

November 2009

1 Some studies in the US (O’Barr and Conley, 1992; Cost Effective Measurement, 1998), UK (Ezra, 1999) and Australia (Coopers & Lybrand, 1997; William M. Mercer, 1998) estimate the opportunity cost of poor governance to be 50–100 basis points per annum while others estimate it to be as much as 300 basis points (Ambachtsheer, 2006). While it is difficult to come up with a definitive number, the key conclusion is that studies in this area point to a large opportunity cost.

/ p 129Russell Investments // Risk management: A fiduciary’s guidebook // Mind the ‘governance gap’

Trustees are faced with increasingly challenging decisions on investment strategy, not least much greater complexity and a vastly wider array of choice in terms of both asset class configuration and product selection.

We focused on Fiduciary Management as an example of an alternative structure that pension schemes can adopt given the wide press coverage on this subject and increased interest in our own fiduciary management services. In the Netherlands, the Fiduciary Management model has been widely embraced as one potential solution to address the so-called ‘governance gap’. Many providers in the UK are now offering similar sets of services, under the banner of Fiduciary Management or Implemented Consulting, IFF carried out 200 interviews with pensions managers, internal CIOs and trustee chairs spanning different scheme sizes. They looked at the following areas in particular: membership of Trustee Boards, the use of investment sub-committees or other delegates, and the degree of delegation. They also polled each participant’s views on the effectiveness of their current structures and their knowledge and thoughts on Fiduciary Management.

KEY SURVEY FINDINGS Many observations from the survey were not unexpected and indicate good governance practices. However, some of the findings surprised us and point to ineffective delegation to the appropriate level of capability, specifically:

1. Inappropriate focus of trustee time Almost half of respondents indicated that they spend more than 25% of their trustee meeting time on investment matters. Arguably this is too much and may not leave enough time for other Board issues. The observation suggests that trustees may not be delegating appropriately. Other results from the survey support this view with trustees continuing to be involved in the bulk of investment decisions.

Perhaps more surprising is the result that there seems to be little difference in time spent on investment matters by trustees between those schemes that do have an Investment Committee and those schemes that do not. Are trustees actually delegating decisions to Investment Committees or are they effectively re-doing the work of the Investment Committee at the trustee level? In some cases, trustees will need to spend time getting comfortable with the issues considered by their Investment Committees but this should be reserved for the decisions that have a large impact on performance, such as setting strategy, not for every decision.

2. Lack of clear accountability Approximately one in three respondents believe that multiple groups are responsible for each of the different investment decisions. This is particularly surprising when looking at who is responsible for setting strategy, one of the most important investment decisions, where there should be no ambiguity on who is responsible. This result indicates lack of clarity, and ultimately, accountability.

16% of respondents indicate that their adviser or actuary is jointly responsible for setting investment strategy. While these parties would certainly be expected to provide input on the strategy, they should not be the decision-makers.

3. Inappropriate structures for responsive decision making 86% of respondents believe that their current decision-making structure is effective, despite evidence of non-conformity with agreed principles of best practice. This is not surprising given that we know that a ‘confidence bias’ is evident in most behavioural finance studies. However, 40% of respondents are not confident in their ability to respond quickly to new situations, which given recent market events seems a critical characteristic of any effective decision-making system.

Linked to the overconfidence observed, the challenge of how to appraise the performance of trustees remains. The overwhelmingly positive self-assessments given in response to this survey perhaps serve as evidence of the need to address this challenge.


Many observations from the survey were not unexpected… However, some of the findings surprised us and point to ineffective delegation to the appropriate level of capability.

If governance structures are to evolve, decision-makers need a good understanding of the options available to them. Given widespread coverage of Fiduciary Management, we were surprised to learn that 75% of respondents did not feel that they had a meaningful understanding of this subject, saying that they had ‘little’ or ‘no knowledge’ of this approach. This stark mismatch between provider hype and investor knowledge highlights the need for greater education on alternative decision-making models to help Trustee Boards give more informed consideration to the choices available to them.

IDENTIFYING THE RIGHT DECISION-MAKING STRUCTURE FOR YOU There are many structures available to pension schemes with no single structure being optimal for all schemes. Nevertheless, every scheme has available a solution that satisfies best practice, through the tools of delegation and outsourcing. The most effective arrangement depends on each scheme’s access to experience and in-house resource. However, access to in-house expertise does not preclude the delegation of some tasks. Importantly, in all instances, the need for continued engagement between those delegating and those delegated to is the key to success.

In addition, Trustee Boards should be well versed in sound principles of good governance. Considerable work has been done on identifying the key features of sound practice and experience suggests that the following practices foster successful outcomes:

• Clearly defining the different decision-making roles;

• Isolating the different investment decisions;

• Delegation of decisions to the appropriate level of expertise and resource; and

• Appropriate frequency and detail in reporting.

CONCLUSION It seems reasonable that sound decision-making structures and practices should improve the chances of investment success2. However, there is evidence to suggest that many schemes have not kept pace with the speed of change in the investment landscape over the last few years. Consequently, we believe there is room for enhancement in the decision-making structures of many UK defined benefit pension schemes.

With the increased complexity of investment, more schemes closing (drawing the final payment date closer) and increased pressure on scheme funding, trustees must consider carefully how to allocate responsibilities to maximise the use of scarce resource and ensure that the trustee body can remain focused on the key strategic issues around funding, investment and risk. We believe that trustees will benefit from greater delegation, allowing them to apply greater focus on top-level decisions such as setting strategy and would welcome more education to help them understand the different options available to them.

If trustees are to change to their decision-making models to deal with the extra intricacies of investment arrangements today, they need a clear understanding of all the options available. The results of our survey also suggest that the industry needs to offer more education to help decision-makers make more informed decisions.

2 It does not guarantee it though. A good process can sometimes lead to unfortunate outcomes, and a poor process can sometimes lead to favourable results. On average, however, a sound process should improve the chances of successful outcomes, and it will certainly provide a clear record of decisions and implementation. Moreover, the Prudent Man concept, introduced in an 1830 Harvard College decision, dictates that trustees will not be in breach of their fiduciary responsibilities if they follow sound investment principles. See Ilkiw (2007) for further details. For those interested, a more contemporary reference to the prudence standard can be found in Prudent Investor Rule or the Uniform Prudent Investor Act in the US (both derived from the Prudent Man concept).


Going forward, we plan to repeat this survey bi-annually to track how approaches to investment decision-making practices are changing. This would also allow us to explore some of the issues uncovered from this study in more detail.

Section 1: Introduction It is getting increasingly challenging for trustees to manage pension fund investments. Schemes have suffered large deficits, driven by volatile investment markets, a sustained reduction in long-term interest rates and longevity. This feature, together with an increased focus on liability-hedging and greater interest in alternatives and alpha-orientated approaches, have led to trustees implementing more complicated fund structures over the past decade. However, decision-making models have not materially changed for many schemes over the last decade. Investment committees remain in the minority, powers of delegation are not being exercised, and the quarterly meeting cycle remains the dominant structure for managing all decisions. Therefore, a ‘governance gap’ has been created.

This ‘governance gap’ can cost pension schemes up to 3% pa according to some estimates3. While these studies cannot draw a clear link between good processes and returns, the observations fuel our suspicions.

Are our suspicions justified? There are many surveys on general pension fund governance, but very few that focus specifically on investment decision making. Therefore, we commissioned IFF Research (IFF) to look into current decision-making structures of UK defined benefit (DB) pension schemes. Section 2 summarises the make-up of the participants we surveyed.

The questions we asked investigated the current decision-making structures of each scheme: the membership of Trustee Boards, the use of Investment Committees or other delegates, and the degree of delegation (section 3). We also polled the respondents’ views on the effectiveness of their current structures (section 4) and their knowledge/views on Fiduciary Management (section 5), an investment decision-making structure that first became popular in the Netherlands to deal with the increasingly complicated investment needs of institutional investors.

Based on the results of our survey, we believe that there is sufficient evidence to indicate that many decision-making structures have not kept pace with increased fund intricacies. However, before trustees can evaluate alternative configurations, they need a clearer understanding of the different structures available and best practice governance principles.

Based on our experience of working with some of the largest investment funds globally over the last 40 years, we have observed some principles that we believe lead to effective decision making. In the Appendix we set out these principles, looking at other sources of inspiration too, and describe how to embody them into possible investment structures that trustees can implement.

In all instances, the need for continued engagement between those delegating and those delegated to is the key to success.

3 Estimates vary greatly. See O’Barr and Conley (1992), Cost Effective Measurement (1998), Ezra (1999), Coopers & Lybrand (1997), William M. Mercer (1998) and Ambachtsheer (2006).


Section 2: Survey participants

SCHEME SIZE Over June and July 2009, IFF carried out telephone interviews with representatives from 200 UK DB pension schemes of varying sizes selected at random.

Exhibit 1: Participants by size of scheme

Almost half of the respondents represent what we label ‘small schemes’ for this study (<£100m). For the purpose of the survey, we have grouped the remaining respondents together as ‘large schemes’.

PRIMARY ROLES OF INTERVIEWEES

Exhibit 2: Primary role of respondents

Approximately two-thirds of the interviewees were internal pensions managers, with chairs of trustees making up the majority of remaining respondents (mainly from smaller schemes). This was deliberate as we targeted participants that were fairly well informed, particularly about their own scheme.

The survey covered a range of different sized schemes and targeted individuals that were fairly well informed, particularly about their own schemes.

Exhibit 1 – Participants by size of scheme

£100-£499m23%

£500-£999m10%

Over £1bn12%

Unspecified9%

Up to £99m46%

Exhibit 2 – Primary role of respondents

Internal CIO10%

Part of investment committee

7%

Pensions manager

66%

Chair of trustees17%


Section 3: Survey results on decision-making structures

TRUSTEE BOARD COMPOSITION Q: How many trustees:

• make up the full trustee board?

• are professional trustees, member-nominated trustees or other (e.g. employer-nominated trustees)?

• have a financial/investment background?

Exhibit 3: Composition of trustee boards (number of people)

OBSERVATIONS • On average, bigger schemes have more trustees.

• The split of Trustee Board by trustee type does not vary significantly by scheme size, with a third of Trustee Boards comprised of individuals with some finance or investment expertise, on average.

It is not surprising that bigger schemes have more trustees, but the fact that the proportion of professional trustees and those with a finance or investment background is similar across scheme size means that larger schemes have access to more expertise.

Message: There is a “governance gap” between large and small schemes.

Average number of trustees0 2 4 6 8 10 12

Total

Largest (>£500m)

Large (>£100m)

Small (<£100m)

Professional MNTs Other

2.3 with finance or investment background

3.2

1.8

3.1


USE OF INVESTMENT COMMITTEES Q: Does your scheme have an Investment Committee?

Exhibit 4a: All schemes

Exhibit 4b: Large schemes (>£100m)

Exhibit 4c: Small schemes (<£100m)

Exhibit 4a – All schemes

Have an IC38%

Don't have an IC59%

No clear answer3%

Exhibit 4b – Large schemes (>£100m)

Don't have an IC42%

Have an IC58%

Exhibit 4c – Small schemes (<£100m)

Don't have an IC77%

Have an IC20%

No clear answer3%

60% of schemes do not have an Investment Committee, despite this feature being one of the key recommendations from the Myners Review.


OBSERVATIONS • 60% of schemes do not have an Investment Committee, despite this feature being

one of the key recommendations from the Myners Review (updated in 2008), which continues to be stated best practice today.

• While the issue is most acute for smaller schemes (<£100m), with 4/5ths not having one, 2 in 5 large schemes do not have an Investment Committee according to our research.

It is perhaps unsurprising that the majority of smaller schemes do not have an Investment Committee given the smaller Trustee Board size. This means that smaller schemes have access to less expertise than larger schemes and have to rely on other experts, such as consultants and investment managers, for this knowledge. In these cases, sufficient engagement is needed between the trustees and the advisor. However, size alone should not preclude delegation to an Investment Committee or other internal resource; this decision should be determined by the level of scheme complexity.

Message: There is a “governance gap” between large and small schemes.

INVESTMENT COMMITTEE COMPOSITION Q: If you have an Investment Committee, how many members:

• make up the Investment Committee?

• are professional trustees, non-trustees/company representatives, member-nominated trustees or other (e.g. employer-nominated trustees)?

• have a financial/investment background?

Exhibit 5: Composition of Investment Committees (number of people)

OBSERVATIONS • The split of Investment Committee by member type does not vary significantly by

scheme size, with roughly two-thirds of members having some finance or investment expertise, on average.

Exhibit 5 – Composition of Investment Committees (no. of people) Average number of members

0 1 2 3 4 5 6

Total

Largest (>£500m)

Large (>£100m)

Small (<£100m)

Professional Trustees Non-trustees/Company representatives MNTs Other

3.3

2.7 with finance or investment background

2.9

2.4


• The average proportion of members with financial expertise in an Investment Committee is double that of Trustee Boards.

• In general, where there is an Investment Committee, it includes whatever professional resource and investment expertise as exists at the trustee level, and also co-opts on other expertise in some instances.

The increase in the proportion of members with investment expertise in an Investment Committee relative to that of a Trustee Board is encouraging.

Member Nominated Trustees (MNTs) are still being included in most Investment Committees. While we do not have sufficient information from this survey to judge whether MNTs have been included for their expertise or principally for representation, our experience informs us that inclusion of expert MNTs onto Investment Committees is the exception rather than norm. This suggests delegation to representation rather than delegation to expertise.

Message: Where an investment committee exists, the composition suggests delegation to expertise, i.e. appropriate delegation, but delegation to representation is still evident.

TIME SPENT ON INVESTMENT MATTERS Q: In general, what proportion of the full Trustee Board’s meeting time is taken up by investment matters?

Exhibit 6: Trustee board time spent on investment matters

The ‘total’ figure in this chart excludes participants who did not give a clean response on whether their scheme has an investment committee.

OBSERVATIONS • Almost half of respondents indicated that they spend more than 25% of their trustee

meeting time on investment matters.

• Perhaps worryingly, there seems to be little difference in this statistic between those schemes that do have an Investment Committee and those schemes that do not.

Exhibit 6 – Trustee Board time spent on investment matters

0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

50%

Up to 10% 11%-25% More than 25%Trustee board time spent

% o

f res

pond

ents

Total With IC Without IC


There is no optimal amount of time that Trustee Boards should spend on investment matters. However, spending more than 25% of Board time on investment matters may not leave sufficient time to cover non-investment decisions and, in our view, could indicate that Trustee Boards are focusing on the wrong decisions, e.g. manager selection and monitoring.

The second observation suggests that trustees could effectively be re-doing the work done by Investment Committees. This could be because they do not trust the recommendations of their Investment Committee or because they feel that they need to cover the subject in the same detail to satisfy their responsibility. For the latter point, they should only do this if the decision will have significant impact on risk and returns at the total portfolio level.

There may be a correlation between the existence of an Investment Committee and the degree of complication in the investment arrangements, but we believe that this only partly explains the result, as the examination of delegation structures evidences (see below).

Message: Trustee boards may not be delegating adequately, even where they do have an investment committee.

DELEGATION STRUCTURES Q: Who within the scheme decides on various investment decisions (different decisions were listed out)?

• We show both the perceived number of decision-making groups for each decision and more detail of who these decision makers are in the exhibits below.

All schemes

Exhibit 7a: Perceived number of decision-making groups (all schemes)

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

Investmentstrategy

LDI policy

% o

f res

pond

ents

Portfoliostructure

Active vs passive

Manager selection

1 2+


Exhibit 7b: Groups responsible for various investment decisions (all schemes)

In the exhibit above, if the sum of the bars for each decision exceeds 100%, i.e. extends to the right of the black dotted line, respondents have indicated that multiple decision-making groups are responsible for that decision.

As you move across the list from left to right in exhibit 7a (or down in exhibit 7b), the decisions move from strategy-oriented to implementation-oriented. Based on best practice principles, the decisions nearer the left (or top) of the list should be retained by the Trustee Board and those near the right can be delegated. There is little evidence of this occurring in practice.

OBSERVATIONS • On average, approximately one in three respondents believe that multiple groups

are responsible for each of the different investment decisions.

• 16% of respondents indicate that their adviser or actuary is responsible for deciding the investment strategy, with three of these indicating that these groups were solely responsible.

While multiple groups can help those responsible for each decision, there should only be one group responsible for each decision. This is particularly important when setting strategy, one of the most important investment decisions, where there should be no ambiguity on who is responsible. The results imply that not all respondents are clear on who is ultimately responsible for certain decisions.

Also, while advisors would certainly be expected to provide input on the strategy, they should not be the decision-makers. It is worrying that this decision is not retained 100% by the trustees and alarming that some respondents believed that their actuary or advisor was solely responsible for this decision on their behalf.

Despite the deliberate choice of words in the question, it is possible that the results were skewed by respondents interpreting the questions as “who is involved in” rather than “who is responsible for” each decision.

Message: There may be confusion on who is responsible for certain decisions. This can indicate a lack of clarity, and ultimately, accountability.

Exhibit 7b – Groups responsible for various investment decisions (all schemes)

% of respondents

Investment strategy

LDI policy

Portfolio structure

Active vs passive

Manager selection

Trustees Investment Committee In-house team Advisor/Actuary Other external

0% 20% 40% 60% 80% 100% 120% 140% 160% 180%


SCHEMES WITH AN INVESTMENT COMMITTEE

Exhibit 7c: Perceived number of decision-making groups (schemes with an investment committee)

Exhibit 7d: Groups responsible for various investment decisions (schemes with an investment committee)

OBSERVATIONS • There is little difference in the results for those schemes that have an Investment

Committee and those that do not, with approximately one in three respondents believing that multiple groups are responsible for each of the different investment decisions, on average. Moreover, 2 in 5 respondents from schemes that have Investment Committees indicated that multiple groups were responsible for setting strategy.

• The bulk of decision making continues to be retained at the trustee level.

It would appear from these observations that trustees are not delegating the actual decision-making function to Investment Committees, but getting these sub committees to do the leg work.

Message: Trustees may not be delegating adequately.

Exhibit 7c – Perceived number of decision-making groups (schemes with an Investment Committee)

0%

10%

20%

30%

40%

50%

60%

70%

80%

% o

f res

pond

ents

Investmentstrategy

LDI policy Portfoliostructure

Active vs passive

Manager selection

1 2+Exhibit 7d – Groups responsible for various investment decisions (schemes with an Investment Committee)

% of respondents

Investment strategy

LDI policy

Portfolio structure

Active vs passive

Manager selection

Trustees IC In-house team Advisor/Actuary Other external

0% 20% 40% 60% 80% 100% 120% 140% 160% 180%


Section 4: Self assessment survey results

CONFIDENCE Q: Overall, how confident are you that the scheme’s current investment decision-making processes are suitable to meet future challenges?

Result and observations (no exhibit)

• 86% of respondents stated that they were “confident”, despite evidence of non-conformity with agreed principles of best practice in governance and decision-making (delegation to expertise, clarity of accountability etc).

This observation is not surprising and demonstrates the classic human behavioural trait of overconfidence4. No-one likes to admit (or admit loud enough to instigate change!) that their decision-making structure needs to be improved.

The challenge of how to appraise the performance of trustees is one that the industry has yet to solve, and the overwhelmingly positive self-assessment here perhaps serves as evidence of the need to develop clear guidance on how best to address this challenge.

Message: Overconfidence can be a barrier to the improvement of investment decision-making structures.

EFFECTIVENESS Q: How would you rate the scheme’s current effectiveness regarding investment decision making based on different criteria?

Exhibit 8: Self-assessment of investment decision-making

OBSERVATIONS • Unsurprisingly, respondents are generally confident in their scheme’s investment-

decision process (the majority responded ‘Very good’ or ‘Fairly good’ when answering this question).

Exhibit 8 – Self-assessment of investment decision-making

% of respondents

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Trustee's time being spentefficiently on investment matters

Minimising the burdenon trustees

The ability to respond quickly tonew situations/opportunities

The ability to makedecisions generally

Obtaining specialistexpertise when required

Very good Fairly good Neither Don't know Fairly poor Very poor

4 According to academic studies (Svenson, O. (1981)), 90% of people asked whether they are a better than average driver are likely to say yes. How can this be true? By definition, only half of respondents can be above average drivers and the other half must be below average. This is known as “overconfidence,” a typical human trait, and can cause costly mistakes when it comes to investing.


• We gain more value looking at relative levels of confidence. Interestingly, 40% of respondents are not confident in their ability to respond quickly to new situations.

Given the high market volatility experienced over the last 18 months and the desire by many pension funds to adopt more dynamic strategies, such as liability-responsive asset allocation or other de-risking strategies, this observation points to an obvious ‘governance gap’ within the structure of many schemes.

Message: Many schemes may not have the right structure in place to deal with a need for responsive decision making or ‘real-time’ oversight.

FURTHER OBSERVATION • Over 90% of respondents indicate that they are good (‘very’ or ‘fairly’’) at obtaining

investment advice.

While it is important that trustees rely on experts to help them make more informed decisions, it is also important that trustees are in a position to challenge the advice they receive. The very high confidence levels in the ability to obtain advice may indicate over-reliance in some instances, particularly when coupled with the observation that in a significant number of cases respondents indicated that advisers were decision-makers.

Message: Trustees may be over-relying on advisers in driving decisions.

Section 5: Responses on fiduciary management How much do you know about Fiduciary Management? If you do not know much, you are not alone according to the results of our survey.

With increased investment complexity over the last few years, the range of available investment decision-making structures has expanded recently. Do investors fully understand all of the options available to them? We tested respondents’ knowledge of one such structure, Fiduciary Management, also commonly referred to as Implemented Consulting in the UK.

For the questions relating to exhibits 9-11, we deliberately did not read out any definition of Fiduciary Management. Prior to asking respondents’ views on this approach as a concept (exhibit 12), we provided a definition, which is given later in this section.

KNOWLEDGE OF FIDUCIARY MANAGEMENT Q: What is your level of knowledge of Fiduciary Management, which can also be referred to as Implemented Consulting?

Exhibit 9: Knowledge of Fiduciary Management Exhibit 9 – Knowledge of Fiduciary Management

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

All schemes

Large schemes(>£100m)

Small schemes(<£100m)

Good knowledge Fair knowledge Little knowledge Heard the term Never heard of it


OBSERVATIONS • Only 25% of respondents indicated some knowledge of the concept of Fiduciary

Management, i.e. 3 in 4 respondents have little or no knowledge on the subject.

We believe that the majority of respondents do not have sufficient knowledge about Fiduciary Management to make an informed decision on the subject.

Message: If governance structures are to evolve, the industry needs to offer more education on new models so that trustees can make more informed consideration to the alternative structures available to them.

CURRENT USE OF FIDUCIARY MANAGEMENT Q: Do you currently use Fiduciary Management—this can also be referred to as Implemented Consulting?

Exhibit 10: The current use of Fiduciary Management

OBSERVATIONS • Very few schemes currently use Fiduciary Management.

This observation is not surprising given that very few schemes understand what it is.

Exhibit 10 – The current use Fiduciary Management

% of respondents

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

All schemes

Large schemes(>£100m)

Small schemes(<£100m)

Yes No Don't know


PERCEIVED BENEFITS AND DRAWBACKS Q: What are the perceived benefits and drawbacks of this concept? (Asked to those aware of Fiduciary Management.)

Exhibit 11: Perceived views on Fiduciary Management

The answers were unprompted to avoid biasing opinions. Also, there were many more answers with only the most common answers shown in the chart above.

OBSERVATIONS The perceived ‘losing touch with investment decision making’ answer suggests that schemes are unwilling to delegate certain decisions, believe that they will receive inadequate reporting from their fiduciary manager or are fundamentally opposed to the concept.

The ‘prohibitive fees’ view suggests that the industry must yet demonstrate the value of Fiduciary Management to justify the fees.

21 out of the 25 schemes that currently use Fiduciary Management are ‘very’ or fairly’ likely to continue to use it. But of those that don’t, only 5% of schemes are likely to adopt Fiduciary Management over the next 2 years, despite any perceived attraction. The low anticipated take up could also be because investors are waiting for a track record to build up before they commit to new structures.

Message: If governance structures are to evolve, further education is needed on new models to help Trustee Boards give more informed consideration to the different structures available to them.

21 out of the 25 schemes that use Fiduciary Management are likely to continue to use it.

Exhibit 11 – Perceived views on Fiduciary Management

0%

5%

10%

15%

20%

25%

30%

Access tospecialistexpertise

Reduces theburden ontrustees

More responsiveto changesin markets

Trustees maylose touch withinvestmentdecision making

Fees tooexpensive

% o

f res

pond

ents

Advantages Disadvantages


WHAT IS FIDUCIARY MANAGEMENT? Articulated by one of the founders of this concept, Anton Van Nunen5:

“Fiduciary Management is an alternative investment decision-making scheme that could be used by pension funds. Fiduciary managers are third party suppliers who advise in terms of assets and liability matching; they construct an efficient investment portfolio; they select the managers to run that portfolio; they monitor these managers; and they provide regular reports on each of these tasks.”

ATTRACTION OF FIDUCIARY MANAGEMENT Q: Overall, how attractive is Fiduciary Management as a concept? (This followed a brief description of Fiduciary Management as cited above.)

Exhibit 12: Relative attraction of Fiduciary Management

OBSERVATIONS • Around a quarter of respondents indicated that they see Fiduciary Management as

being an attractive option worth pursuing, with respondents specifying the three main advantages highlighted in Exhibit 11.

• Equally many were opposed to the concept, primarily due to a concern that they would lose control of investment decision making.

• About half of the respondents were undecided or insufficiently informed to form an opinion. This suggests that decision makers need to be better informed.

Message: If governance structures are to evolve, further education is needed on new models to help Trustee Boards give more informed consideration to the alternative structures available to them.

Exhibit 12 – Relative attraction of Fiduciary Management

% of respondents

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

All schemes

Large schemes

Small schemes

Very unattractive Fairly unattractive Don't know Neither Fairly attractive Very attractive

5 Anton van Nunen, PhD is head of Van Nunen & Partners, a consulting firm serving both institutional and individual investors since1998. Credited by many in the Dutch pension fund market as pioneering the concepts of Fiduciary Management, he has implemented them at such institutions as VGZ-IZA, a major Dutch health insurance company, the Campina Pension Fund, and the Yarden Insurance Company.


Section 6: Conclusion The results of our recent survey suggest that there is room for enhancement in the decision-making structures and practices of many UK defined benefit pension schemes. With the increased complexity of investment, more schemes closing (drawing the final payment date closer) and increased pressure on scheme funding, trustees must carefully consider how to allocate responsibilities to maximise the use of scarce resource and ensure that the trustee body can remain focused on the key strategic issues around funding, investment and risk.

If governance structures are to evolve, decision-makers need a good understanding of the wide array of options available today. Our survey also indicates that the industry needs to offer more education to help Trustee Boards to consider the choices available. There is no single structure that suits all schemes. Nevertheless, there is a configuration for each scheme which conforms to best practice with the most effective arrangement depending on each scheme’s individual circumstances. Importantly, in all instances, the need for continued engagement between those delegating and those delegated to is the key to success. In addition, Trustee Boards should be well versed in sound principles of good governance.

We hope that this insight helps trustees to understand the principles and options a little better so that they can make more informed decisions, and improve their chance of investment success. Going forward, we plan to repeat this survey annually to track how approaches to investment decision-making practices are changing. This will also allow us to explore some of the issues raised in this survey in more detail.

Related readingW. O’Barr and J. Conley (1992), Fortune and Folley: the Wealth and Power of Institutional Investing, Business One Irwin, Homewood, IL.

Cost Effective Management (1998), a study of 240 funds over the five years ending 1997—Toronto.

D. D. Don Ezra (1999), Excellence in UK Pension Fund Management—a perspective on current practices, Russell London Monograph 11.

Coopers and Lybrand (1997), Do superannuation funds add value through sector specialization?, Australia.

William M. Mercer (1998), Capturing Good Investment Performance, Lend Lease Services, Australia.

K. Ambachtsheer (2006), How much is good governance worth?, The Ambachtsheer Letter #245, June 2006.

John H. Ilkiw (1997), The importance of being prudent, The Portable Pension Fiduciary: a handbook for better pension fund management.

Svenson, O. (1981), Acta Psychologica (International Journal of Psychonomics).

Altis Investment Management AG (2009), Dutch Investment Manager Survey.

If governance structures are to evolve, decision-makers need a good understanding of the wide array of options available today.


Appendix: Sound governance principles and potential decision-making structures

PRINCIPLES FOR EFFECTIVE DECISION MAKING Russell has worked with a wide range of institutional investors across the globe over the last 40 years. This experience, and the lessons learnt from looking at other governance practices (see next section for lessons from corporate governance), suggests that the following practices foster successful outcomes:

• clearly defining the different decision-making (fiduciary) roles;

• isolating the different investments decisions;

• delegation of decisions to the appropriate level of expertise and resource; and

• appropriate frequency and detail in reporting.

DECISION-MAKING ROLES A ‘fiduciary’ is a person or body that stands in a position of trust to another person (the beneficiary). The fiduciary has scope to exercise some discretion and its actions will affect the beneficiary’s legal or practical interest. There are three main fiduciary roles:

Governing fiduciaries—members of corporate boards and boards of trustees—are ultimately responsible for ensuring that pension assets are prudently6 invested. They usually employ managing fiduciaries—often in the form of an Investment Committee or internal CIO—to provide advice and oversee policy implementation, who in turn, employ a number of operating fiduciaries with specialised knowledge and skills to implement and manage investment policy on an ongoing basis. A summary of the different roles is given in the table below.

Exhibit 1: The different fiduciary roles

Name Role Pension fund example*

Governing fiduciary Plan: Set policy Trustees

Managing fiduciary Manage: Oversee Investment committee or CIO policy implementation

Operating fiduciary Implement: Manage Internal or external money day to day activities managers

* Some groups can perform multiple roles.

INVESTMENT DECISION ISOLATION AND EFFECTIVE DELEGATION Identification of the distinct stages of investment decision making helps with the measurement of the success of different decisions and is helpful for accountability and effective governance.

In the mistaken belief that they are lessening their exposure to liability for breach of duty—or because these tasks are more interesting and easy to understand—many governing fiduciaries often retain decisions that are best delegated to managing and operating

A fiduciary has unilateral discretion over other people’s assets and is required by law to act prudently.

6 ‘Prudence’ is the exercise of good judgment to avoid decisions that result in undesired consequences and can be interpreted as ‘ensuring proper process’.


fiduciaries, such as manager selection and monitoring. In most cases, liability exposure is actually increased because they do not have the time, relevant knowledge or skills required to make timely and sufficiently informed decisions. Governing fiduciaries are more likely to satisfy their duties if they learn to delegate.

For effective decision making, governing fiduciaries should retain responsibility for those issues they are best positioned to address and delegate, to managing and operating fiduciaries, those issues which these groups are better equipped to handle. In general, the smaller the impact a given decision has on the fund’s results, the further down the governance ladder it should be delegated. Therefore, governing fiduciaries should decide a fund’s long-term asset mix strategic policy, which is the principle determinant of long-term performance, but delegate to operating fiduciaries security selection and asset mix timing decisions. These have a meaningful but nonetheless secondary impact on long-term performance.

The following figure shows the potential roles played by different fiduciary groups under our ideal of best-practice decision-making structures (and, of course, the same body can take responsibility for multiple roles).

Exhibit 2: The governance ladder: an effective decision-making framework

REPORTING AND MONITORING Informative and timely reporting of compliance and performance is the feedback mechanism for verifying and evaluating investment activities and their contribution to achieving established fund objectives. “Informative” means a package of relevant information designed to allow a supervising fiduciary to verify compliance with established investment policies and understand the impact of key investment decisions.

“Timely” means a reporting frequency that corresponds to the period of time which decisions can reasonably be monitored and evaluated.

In general, reporting detail and frequency should increase as you move down the chain of delegation. Historically, governing fiduciaries (trustees) only needed to meet once or twice a year to verify compliance and evaluate performance using total-level performance reporting formats. Managing fiduciaries usually met monthly or quarterly and utilised more detailed reports. Operating fiduciaries were continuously engaged and used multiple detailed data inputs into their decision-making structure.

Figure 2 – The governance ladder: an effective decision-making framework

Risk budgetting

Governance

Objective setting

Investment strategy

Governing Fiduciary

Manager selection

Manager structure

Manager researchManaging Fiduciary

Security selection

Custody & administration

Portfolio controlOperating Fiduciary

Implement

Plan

Manage

Risk budgeting

Governance

Objective setting

Investment strategy

Governing Fiduciary

Manager selection

Manager structure

Manager researchManaging Fiduciary

Security selection

Custody & administration

Portfolio controlOperating Fiduciary

Implement

Plan

Manage

The smaller the impact a given decision has on the fund’s results, the further down the governance ladder it should be delegated.


This was feasible when investment strategy was reasonably static, included little by way of alternatives and overlays, and was implemented via a small set of products.

However, with the increased complexity, trustees now need to spend more time on strategy and less time on implementation, which can be effectively delegated. Moreover, given the increased complexity at the implementation stage, there are strong arguments in support of having a more ‘real-time’ managing fiduciary beyond the traditional quarterly meeting cycle.

LESSONS FROM CORPORATE GOVERNANCE PRACTICES There is abundant literature and evidence on governance principles that come from looking at corporations. Consider the evolution of a corporation. It may start in a small way with a single founder or founding family or small group of colleagues with an idea for a product or service. It may stay small for a while, with the founders playing roles at various levels either intuitively or with an informal structure. They may act simultaneously as board, management and employees, doing everything from establishing a direction to producing and delivering the product or service and keeping records.

If they expand successfully, they may find that they need to hire others. These others can have any sort of role. The founders may need more employees directly involved with the product or service. They may need professional management to help hire the employees and to implement the policies that the informal board establishes. They may want independent directors to broaden the base of experience, knowledge and judgment that is applied to setting the direction and evaluating what is working well and what isn’t. Depending on the background and culture of the founders, their new additions may be specialists or generalists. Ultimately, for reason of cost or specialisation, some functions may be outsourced.

In essence, there are three broad governance levels, as depicted in the figure below. At the top there is the board, responsible for setting the direction and goals, and ultimately for all oversight and evaluation. At the bottom is the layer of employees, responsible for producing the product or service. In the middle is management, responsible for implementing the board’s policies, hiring and deploying employees and co-ordinating the efforts of employees.

Exhibit 3: The governance structure or corporations

Set business strategyBoard of Directors

Manage resources to deliver the strategy

Management

Produce the product / Deliver the service

Employees

Implement

Plan

Manage

Set business strategyBoard of Directors

Manage resources to deliver the strategyManagement

Produce the product / Deliver the serviceEmployees

Implement

Plan

Manage

There is abundant literature and evidence on governance principles that come from looking at corporations.


The more complex the organization, the more necessary it becomes to formalise the governance structure. A multi-national corporation and small family-run business look quite different. The same principles can be applied to the governance structure of pension schemes.

POTENTIAL DECISION-MAKING STRUCTURES The figure below illustrates some of the options available to institutional investors. There is no single solution that is suitable for every scheme and these structures are not mutually exclusive. Nevertheless, there is a structure relevant for every scheme that satisfies best practice. The most effective structure depends on each scheme’s investment complexity, and access to in-house resource and experience. However, the key to success is the continued engagement between those delegating and those delegated to.

Exhibit 4: Possible decision-making structures (not mutually exclusive)

SIMPLE TRUSTEE/MANAGER STRUCTURE (FIRST COLUMN) When the investments of a pension fund are very simple, the trustees can outsource implementation to asset managers. The trustees, with the help of advisors and in house resources (e.g. a pensions manager or executive team), should set strategy and manage the investments.

This was the model used by most UK pension schemes up to the mid-1990s, with implementation usually outsourced to a single balanced manager. However, the historic implementation this popular model diverges from best practice in two key ways:

• All investment decisions were usually delegated, including setting strategy. Best practice dictates that the investment strategy should be set by the fiduciaries that have the best knowledge of the scheme, i.e. the trustees.

• Implementation was carried out by a single manager. Best practice dictates that decisions should be delegated to the appropriate level of expertise and a single manager is not necessarily superior in all asset classes. This departure from best practice is deemed acceptable by many when simple investment choices are being followed.

The most effective decision-making structure depends on each scheme’s investment complexity, access to in-house resource and experience.

Figure 4 – Possible decision-making structures (not mutually exclusive)

Asset Managers

Board of Trustees

Asset Managers

Board ofTrustees

PlanM

anage Im

plement

Asset Managers

Board of Trustees

Asset Managers

Board of Trustees

Asset Managers

Board of Trustees

InvestmentCommittee Internal CIO

Fiduciary Manager

Investment Committee

Investment Committee

?


INVESTMENT COMMITTEES (SECOND COLUMN) As the investment challenges increase—such as utilising specialist managers, increasing the range of alternatives included or implementing more dynamic strategies—it gets more difficult for trustees to manage the investments. In this case, trustees need access to a resource that has more time and expertise to manage the extra investment complexity.

A suitably-staffed Investment Committee (or other appropriate body, such as an internal Executive Office or CIO) provides one such resource. This was one of the recommendations of Myners and many large schemes have adopted this model. However, care needs to be taken that the role of the Investment Committee is clearly defined and decisions are delegated appropriately.

It is easy for trustees to delegate a decision to the Investment Committee but cover the subject in the same amount of detail when it comes to Board approval. This could suggest that they don’t trust the recommendation or feel that they need to cover the subject in the same detail to satisfy their responsibility. In this case, they have not delegated the decision, but duplicated it.

In other cases, trustees make a decision based on an Investment Committee recommendation without sufficient understanding or questioning. This can be acceptable in some circumstances, but in these cases it should be clear that these decisions have been delegated rather going through the charade that trustees are making that decision.

It is often misunderstood that Investment Committees are only suitable for large schemes. The need for an Investment Committee, or other bodies performing similar functions, should be driven by the scheme’s investment complexity rather than size alone. While size is often a proxy for complexity, the range of products available today means that size need not be a barrier to accessing sophisticated investment strategies.

INTERNAL CHIEF INVESTMENT OFFICER OR EXECUTIVE OFFICE (THIRD COLUMN) Another option open to trustees that have the scope to utilise or hire internal resource is to augment their capabilities with an experienced internal CIO (or Executive Office). This model is particularly popular with very large or multi-national schemes.

The CIO could be used in addition to or in preference to an Investment Committee. Where the two co-exist, the Investment Committee has the potential to take on responsibility higher up the governance ladder (in the planning stage) and could work with the Trustee Board more closely to help them make a decision.

Depending on internal resources and expertise available, day-to-day management could be conducted in-house, delegated to an external manager or executed by a hybrid of the two, with more intricate mandates outsourced externally. Also, in many cases pension schemes have put internal Executive Offices at arms length.

FIDUCIARY MANAGEMENT (FOURTH COLUMN) Another option open to trustees to manage complex investment arrangements is Fiduciary Management. Widely embraced in the Netherlands7, many providers in the UK are now offering similar sets of services—under the banner of Fiduciary Management or Implemented Consulting—to bring control and reduce costs through consolidation. A good fiduciary manager can help trustees find the right balance of decision delegation and, in general, is made responsible for:

Care needs to be taken that the role of the Investment Committee is clearly defined and decisions are delegated appropriately.

7 75% of externally managed Dutch pension fund assets are managed by fiduciary managers (if you include in-house managed assets) according to the Altis Investment Management Dutch Investment Management Survey—2009.


• Helping the trustees formulate comprehensive investment objectives and an investment strategy;

• Executing the investment strategy, including managing asset and overlay managers; and

• Controlling and monitoring the risk and performance profile of the overall portfolio and all its individual constituents, including transparent and comprehensive reporting on progress versus objectives and guidelines.

Many of the functions performed by fiduciary managers are similar to those executed by internal CIOs. In many cases, fiduciary managers can implement a much wider range of services, can offer a more dynamic service and may provide greater advice than an internal CIO.

Some commentators liken fiduciary managers to balanced managers where everything is outsourced to the manager. This is not a useful comparison. The three key differences are that a fiduciary manager will help the trustees (i) identify which decisions to delegate, (ii) incorporate liabilities into trustee discussions on strategy, and (iii) find the most appropriate implementation expertise, not just default to an in house or single-provider capability.

Acknowledgements

CONTRIBUTING STAFF The authors would like to acknowledge the valuable input of a number of Russell associates in the creation of this research piece, in particular D. Don Ezra, John Gillies, Lloyd Raynor, Joey Hong, David Millen, Sarah Higgins and Patricia Marron.

ABOUT IFF RESEARCH IFF Research is one of the largest independent research companies in the UK. Established in 1965, they conduct high quality strategic research for a wide range of private and public sector clients, with particular expertise in Business, Enterprise, the Environment; Branding, Marketing and Communications; Employment and Benefits; Financial Services; Learning and Skills; and Regulation.

Fiduciary managers play a similar role to internal CIOs, but in many cases can implement a much wider range of services, can offer a more dynamic service and may provide greater advice.

/ p 153

A model agenda for an investment committee

Investment committees play a central role in the management of most institutional investment programs. Committees have a bad name. There is a camp (and I am in it) that believes human beings are not well-adapted to the committee environment; they are not our natural way of getting things done. Nonetheless, they serve a purpose for institutional investors, so we are stuck with them.

And they can be made better. The 2004 piece that follows looks at the humble agenda and the role it can play in creating a better decision-making structure.

This article was first published as a Russell London Monograph in 2004.

Russell Investments // Risk management: A fiduciary’s guidebook // A model agenda for an investment committee

/ p 154

A model agenda for an investment committee Few practitioners in any field of business or professional occupation will deny that an effective meeting relies on a well defined agenda. The success of a meeting can stand or fall on this apparently innocuous list of business items, and as such the humble agenda deserves a modicum of thought and time in its preparation.

Two formal documents define the responsibilities of today’s pension fund Trustee: the newly strengthened statement of investment principles, setting out governance; and the forward-looking business plan, defining priorities. The investment committee agenda should harness the work of the committee to these two documents: A good agenda will turn plans into action.

This monograph sets out what we think characterises effective agendas for investment committees. There is no universal list of agenda items that define the perfect agenda: any such list would risk turning meetings into box-ticking marathons. However, three key characteristics make for effectiveness:

• Everything on the agenda should have a clear purpose;

• Time should be allotted to each item in accordance with the importance of that item;

• The agenda should follow a logical order.

We do not expect most fiduciaries to disagree with these simple rules. Regrettably, they are often ignored in practice. We discuss in this monograph how these principles can be incorporated into the stewardship of funds.

The governance backdrop Much has been written about pension fund governance and decision making: it is not the purpose of this monograph to revisit this in detail. However, it is useful to remind ourselves of the investment committee’s role in the fiduciary cycle to understand its remit.

Investment committees fill the managing fiduciary role, implementing decisions of the governing fiduciary (the Trustee) and monitoring the activities of the operating fiduciaries (the investment managers and custodian). They concern themselves primarily with the implementation and monitoring phase of the fiduciary cycle, with a duty to report back to the Trustee. The investment committee also prepares information to assist the Trustee in making higher-level decisions, in particular strategic asset allocation. The degree of delegation by the governing fiduciary can vary: it is rare for the investment strategy to be delegated, but common for the manager selection decisions to be fully delegated.

The diagram overleaf shows the broad areas of investment committee responsibility, and how meeting time might be allocated over time. It is our observation that many committees spend a disproportionate amount of time on the custody, administration and reporting aspects of their responsibilities (especially the recent investment performance).

By: Sorca Kelly-Scholte

April 2004

Russell Investments // Risk management: A fiduciary’s guidebook // A model agenda for an investment committee

/ p 155Russell Investments // Risk management: A fiduciary’s guidebook // A model agenda for an investment committee

Higher level issues are starved of time. Without sufficient time on strategy, trawling the performance figures in search of useful information is largely a waste of time.

Exhibit 1: A reasonable allocation of time

It is worth noting that discussions of objectives and strategy do not disappear from an agenda shortly after a triennial review (the much-maligned Asset Liability Modeling (ALM) study). There is a broad range of important decisions that can be categorised under ‘Objectives and Strategy’, from the broad equity/bond split to the allocations to alternative assets, or the policy on currency. All these decisions cannot be crammed onto one or two meetings and still enjoy the attention required to make sound decisions: far better to prioritise the decisions and spread them across meetings to allow time to consider them fully. Although the headline decisions such as the broad equity / bond split are normally made early on in the cycle, other decisions may be made later and may continue beyond a year. Indeed, a rolling programme of review of strategic allocation may be preferable to clustering all the decisions into a small number of meetings.

There also needs to be flexibility to identify and follow-up on new investment opportunities as they arise. For example, many UK pension funds had gilt-oriented bond strategies during the 1990s, and if they did not have sufficient flexibility to revisit strategic questions between formal triennial reviews, they may not have updated their bond strategies to reflect the market change that occurred with the explosion in non-government issuance in the late 1990s.

Time spent on custody, administration and reporting is fairly constant and these will be recurring items on agendas. However, at all times the emphasis should be on context, where results are reviewed against expectations.

Characteristics of an effective agenda With that background, let’s return to the three primary characteristics of an effective agenda.

1. CLEAR PURPOSE Each item on the agenda should have a clear purpose. It should be possible to identify that purpose by reference to the fund’s fundamental objectives and the investment committee’s terms of reference.

Given the range of activities generally covered by investment committees, a list of purposeful agenda items can probably be drawn up of which any investment committee agenda is a subset. To avoid the risk of box-ticking, the requirement for purpose should go further, in particular what is the desired outcome from the discussion of the agenda item?

Early in cycle

Late in cycle

Strategy

Implementation

Reporting


Notice that there are three basic types of agenda item:

1. Information-oriented ‘I am not required to make a decision, but I must review this information to ensure that my original decision remains appropriate.’

Items may be included purely to provide information to review previous decisions, for example to confirm that mandates delegated to portfolio managers are being complied with, or to help assess the ongoing suitability of a particular manager within the overall investment arrangements. The desired outcome from such an agenda item will generally be an affirmation of the original decision, or a set of action points if the information suggests that change is required.

The recurring parts of the fiduciary cycle will prompt the informational agenda items. For example, review of performance, and attendance of the fund’s managers at meetings can be characterised as informational items. Once the planning phase is completed, the majority of time will be spent on implementation and monitoring. The investment committee can therefore expect meetings to contain lots of information-oriented agenda items.

An informational agenda item will be successful if the right information is provided. It should go without saying that this should be considered and relayed to whomever is providing the information. This observation may seem to be obvious, yet it is so often the missing link in the chain. Information without context lacks meaning, and the time spent discussing strategy and manager structure will pay dividends in identifying what information is required. Most crucially, it allows the investment committee to review the updated information against original expectations and thereby focus the discussion on those areas where attention is most required. For example, investment managers who are reporting to the committee should be told what to include: the committee sets the agenda—not the investment manager. If the committee wants to hear why the portfolio appears to have changed in character, then that should be the focus of the manager’s report rather than simply reporting performance.

With this in mind, how often should you meet your investment managers? Should you starve broader reviews of portfolio structure, and information provided by your advisers on manager developments, just to ensure that your manager gets regular airtime? Probably not: if you can’t do both, cut down on time spent listening to the manager telling you what the performance has been for the last six months, and use the time to discuss whether performance patterns are in line with expectations, whether there are developments at the manager that undermine the rationale for your original decision, and whether the overall line-up of managers within the asset class still makes sense. If you have difficulty with this, ask yourself this question: how often has a routine manager presentation driven your decision making?

2. Decision-oriented ‘I am required to make a decision, and I have been provided with information to make this decision’

Decision-oriented agenda items should arise less frequently than information-oriented agenda items, and may have their origins in previous information-oriented items. For example, following a routine, information-oriented review of a manager it may have been decided that the manager needed to be replaced. This will give rise to a decision-oriented item at a subsequent meeting, i.e. which manager or managers should be appointed. Clearly a decision-oriented item is not devoid of information requirements, but the difference that characterises a decision-oriented agenda item is that the desired outcome is for a decision to be reached.


Decisions can be deferred on the basis that further information is required, but we should distinguish between a true need for further information and a reluctance to decide. A skillful chairman has to determine whether further information will help or simply delay.

3. Proposal-oriented ‘I have been asked to make a recommendation to the Trustee. I must review the information to decide on the most appropriate recommendation and identify which pieces of information the Trustee will need in order to approve my recommendation’.

Proposal-oriented items will be included on the agenda less frequently than either information or decision-oriented items, and pertain to decisions not fully delegated to the investment committee and needing approval by the Trustee. The desired outcome is a recommendation to the Trustee, and as such an element of decision making may be required—but here the discussion should also identify what information needs to be included in support of the proposal.

Getting the right information, and knowing what to do with it, is the thread that runs through all agenda items. The diagram below sets out how information flow can be sorted by an investment committee and used to identify issues, inform decision making, and facilitate the generation of proposals for the governing fiduciary to consider.

Exhibit 2: Illustration of proposal-oriented agenda item

2. APPROPRIATE TIME The amount of time allotted to each item will depend on the particular meeting, and the ascendancy of that item in the fiduciary ladder. For example, suppose that the investment committee has been asked to review the results of an asset-liability study in order to form a recommendation to the Trustee on strategy. The investment committee’s duties in monitoring existing investment arrangements are not suspended while it is reviewing the strategic asset allocation, but an agenda which allows only half an hour to discuss asset-liability results while allowing three hours to meet with existing managers is disproportionate.

3. LOGICAL ORDER An orderly agenda will avoid the need for previous discussions under earlier agenda items to be reopened at a later point in the meeting. This is achieved by ensuring that any item with potential consequences for other items occurs earlier in the agenda, and also by grouping together items which rely on common or similar information. For example, it is

Request information

Proposal-oriented

Decision-oriented

Information-oriented

Formproposal

Makedecision

Identifyissue

Reaffirm originaldecision

Investmentcommittee Trustee

Operatingfiduciaries

ManagersCustodianMeasurers

Info

rmat

ion

Prop

ose

Implement


usually more useful to include a general review of performance, portfolio structure and comments on developments at the managers before meeting with individual managers. This general discussion will help identify the appropriate focus for the discussion with the individual manager.

A sample agenda

A sample agenda is shown below. It includes recurring items such as review of the minutes of the previous meeting, and identifies the information required to support a full discussion, and an appropriate allotment of time. It is assumed in the example that the investment committee forms proposals or recommendations on investment strategy to the Trustee, but has itself responsibility for all subsidiary decisions.

Agenda item Broad area of responsibility

Item type Information requirements Time (hours)

Minutes Custody, administration and reporting


Minutes from previous meeting 0.25

Currency strategy Objective and strategy

Proposal-oriented

Analysis of impact of different currency strategies based on the current allocation

Analysis of the relative merits of active and passive currency management

1.5

Identify suitable fund-of-funds products to implement new private equity allocation

Portfolio structure and investment management

Decision-oriented

Analysis of available products

Proposals from products providers

1.5

Review current manager line-up

Portfolio structure and investment management


Analysis of style tilts at managers and at aggregate asset class level

Update on manager developments and advisor option

Reports from managers on portfolio positioning

1

Review fund performance

Custody, administration and reporting


Performance and attribution reports

Independent performance and attribution report

0.5

Administrative items such as trustee report and accounts, statement of investment principles, commission recapture, other reports

Custody, administration and reporting


Commission recapture report

Draft accounts from auditor

0.5

Review progress against business plan

Custody administration and reporting


Up-to-date business plan 0.25

Any other business


Summary

In this industry, it is very easy to become mired in information and statistics. The key to avoiding ‘information overload’ is to identify the purpose of the information provided, and the priority that should be afforded to its discussion. This is what the model agenda aims to achieve.

Each type of agenda item discussed above (information, decision and proposal-oriented) has information content: the division into the three categories identifies the different uses or purposes of that information.

Ordering the agenda and allotting the time recognises the relative priority and complexity of different items. The agenda is therefore a useful tool in categorising and ordering information to ensure it successfully contributes to your meeting rather than leaving participants feeling overwhelmed. Of course—where possible—the information should be sent in advance along with the agenda.

Effective agenda-setting harnesses the power of your decision-making structure. Armed with an effective agenda, the investment committee meetings can be efficient, and successful. They may even become enjoyable!

/ p 161

Mostly brick | Not yet a bonfire

We opened this guide with two papers —There’s nothing normal about risk and Risk management is the cornerstone of investing—which at the time they were written were intended to form part of a three-part series culminating with a case study. When it came time to write this case study, there was so much to include that it turned into two of them. And even then we were able to only skim the surface and begin to illustrate the thought process of how to apply in practice the principles we have been arguing for throughout this guidebook.

The case studies are written from the perspective of a newly-appointed Chief Investment Officer at a large corporate pension plan. The titles refer to the tale of the three little pigs (and the big bad wolf who huffed and puffed); in the first case study our protagonist examines the foundation of decision making at the plan and finds that it has areas of strength and areas of weakness (a house made of “mostly brick”). In the second case study, changes are made and more planned (the burning straw being “not yet a bonfire”). Each case study ends with questions for study and discussion.

Russell Investments // Risk management: A fiduciary’s guidebook // Mostly brick

/ p 162

By: Bruce Curwood and D. Don Ezra

July 2010

PART 1: A PENSION FUND CASE STUDY FOR DISCUSSION

Mostly brick

Call me Deep.

It’s a natural contraction of my first name, Deepak, so it’s what my colleagues call me. Even in my family, where we usually use full names, I’m called Deep, ever since an aunt, an immigrant struggling with her English, observed my silence as a child and proclaimed reassuringly, “The boy is ... deep.” So I’ve been known as Deep for as long as I can remember, a one-syllable contraction that makes me sound like a Brazilian soccer player.

Some months ago I took the latest step in my career. I joined the Colossal Pension Fund as its chief investment officer (CIO). My predecessor had resigned—“by mutual consent,” as the politically correct wording had it—following Colossal’s horrendous performance in that bizarre year, 2008. The plan’s funded ratio—the extent to which its assets provided backing for its benefit promises—dropped precipitously, from 106% to 79%, triggering all manner of retributive penalties. The investment performance itself, for the year, was worse than that of 97% of the funds in a well-respected comparative survey. And worst of all, the disaster came without warning.

I was hired as a change agent. But I didn’t know where to start.

Here’s the background.For many years everything seemed to be fine. Throughout most of the “noughties” decade the fund had sailed through in the top quartile of the same comparative survey. The Investment Committee (IC), designated by Colossal’s board of directors the fund’s “named fiduciary” under the Employee Retirement Income Security Act, had commissioned the traditional asset-liability projections and, confident of the sponsor’s (and its own) ability to stomach the downside of an aggressive investment approach, had adopted an atypical asset allocation policy. They used a high equity exposure (but a 50% hedged currency exposure in their foreign equities, perhaps due to a fear of taking an inappropriate amount of currency risk), and a high allocation to Treasury inflation-protected securities (TIPS).

The minutes of the IC meetings revealed little. They were couched, in minimalist terms, recording decisions that were taken “following discussion.” So I inquired further. Colossal’s chief financial officer, a prominent member of the IC (but not its chair—that was the

I was hired as a change agent. But I didn’t know where to start.

Russell Investments // Risk management: A fiduciary’s guidebook // Mostly brick

/ p 163Russell Investments // Risk management: A fiduciary’s guidebook // Mostly brick

CEO himself), was very helpful. He offered to review with me anything that I wanted to investigate as part of my research into what needed changing. And he referred me to the fund’s consultant, the felicitously named Ms. Goode, who represented a prominent firm that worked with many major corporations.

Goode led me quickly through the fund’s governance structure. The IC included the CEO, the CFO, the vice president of human resources (VP HR), a couple of business line managers and a retiree. It took major decisions only, such as the asset allocation policy, delegating most of the subordinate decisions—like the hiring and firing of investment managers—to its CIO, who in turn was supported by in-house staff and the external consultant. And of course the IC then monitored and reviewed the results, which as I say seemed better than satisfactory for a long period.

If you think of the IC, the staff and the consultant as The Three Little Pigs—I don’t mean to be insulting, I just think in simple terms, and that fable is a useful analogy—the house they had constructed certainly seemed to be built of bricks. Yet the big bad wolf of 2008 had huffed and puffed and not just blown it down, it had virtually blown the house away. The bricks must have concealed a lot of straw. Where was the straw? Unless I could locate it, I wouldn’t be able to rebuild the house and avoid similar wreckage the next time a storm occurred. I know it’s often a fatal conceit, but next time, I thought quietly, will be different.

Part of being Deep is being patient and systematic. I’m an investment professional. I know where the numbers come from: from asset allocation and the performance of the managers with their specialist assignments.

I started with asset allocation. The IC minutes told me little, so I went to the asset/liability study. It was massive. Pages and pages of data and charts with unhelpful headings. What can you learn from a page headed “Comparison of current policy with middle-of-the-road policy”? Nothing. You can’t tell if the stats are meant to imply that continuing the current policy is desirable. Or if changing to a more moderate policy is preferable. I wish consultants would place an interpretation at the top of each page. Either “Comparison shows no reason to change from current policy” or “Moderate policy dramatically controls downside risk” or something of that sort that would help a stranger interpret the mass of numbers. But if you weren’t there, at the meeting where the presentation took place, you missed the voice-over, and it must have been the voice-over that was all-important, because I certainly couldn’t see any obvious way to interpret the comparison.

So, to save time, I asked Ms. Goode to lead me through the presentation. Eventually I got what I wanted, all the voice-overs. I also got what I didn’t expect: an enormous defensiveness. Apparently Goode and her firm were among the potential victims of the debacle.

Colossal’s board had convened a special meeting. (It seems that the pension operation is the biggest part of Colossal’s business, in the sense that its gains and losses outstrip the normal profits from the business operations that Colossal’s shareholders believe they’re investing in.) The Chairman ranted and raved at factors and results beyond his control. Obviously he and his board must be above blame. He demanded a witch hunt. And you know the one certainty: when there’s a witch hunt, you’ll find a witch. Goode and her firm were cast in that role. Their status as appointed consultants was now up for grabs. And Goode knew, from experience, that they would have no chance of being reappointed. So she was keen to defend her advice—and in turn point out how some important elements of her report had been ignored.

I also learned an intriguing thing about the governance process, that didn’t appear in the formal governance chart. That massive report wasn’t the one that had been presented to

…the big bad wolf of 2008 had huffed and puffed and not just blown it down, it had virtually blown the house away.


the IC. What the committee saw was the one that Goode had presented to the former CIO and his staff. From that, the former CIO had made a recommendation to the IC with a brief document that supported staying with the current (aggressive) policy.

Really? Goode’s report never made it onto the IC’s agenda? Why?

Goode explained: IC agendas always seemed packed with historical performance data, administrative matters and manager meetings. Even when they allowed time for staff, which sometimes included the consultant, the few minutes so allocated were sparse. After all, these were high-powered business executives, who just needed summaries to make complex decisions. They just needed the essential facts to figure out the correct approach.

The IC, satisfied that things were going well, and not themselves fully conversant with the subtleties, simply rubber-stamped the recommendation.

Apparently the IC’s belief that the sponsor could withstand the potential downside stemmed from the CFO’s saying so. (And, it turned out, the CFO had at the time chaired the IC. The insertion of the CEO, who also took over as chair of the IC, was a development subsequent to the debacle.) And the belief of the IC members that they could withstand the psychological impact of the downside had never been tested. The members had not been together for the turn-of-the-millennium markets that constituted the previous test.

Was there, I asked Goode, any significant difference between her report and the CIO’s recommendation? Not huge, she confessed; but there were two aspects she wanted to draw to my attention.

One was that the downside, in her report, was measured in terms of the plan’s funded ratio: how low might it go, in adverse markets? Yes, the former CIO said he would present this aspect to the IC. But Goode also recommended that the funded ratio should therefore be made the primary focus of the monitoring reports that the IC looked at, every quarter. Nevertheless, these recommendations never made it to the IC. (Not, she added bitterly, that it would have made the slightest difference.) The IC was used to seeing a fat report of investment performance analytics, leading off with each manager’s performance, portfolio characteristics and trades, then aggregated gradually into asset class results, and finally the performance of the total fund compared with other large funds. Not a word about the benefit liabilities. So, when disaster struck, and the low funded ratio was revealed, that was the first time it hit the IC that there were real consequences to the poor performance.

And the fact that the plan’s benefits were out of sync with those of the funds in the comparative group also turned out to be relevant. Colossal’s plan provided, generously, a certain amount of inflation-indexing of retiree benefits: that’s why the fund held Treasury Inflation Protected Securities (TIPS), as a partial hedge. That also meant that the comparative group wasn’t really terribly relevant. But it’s what the IC was accustomed to using as its measure of success or failure. (Why? Goode said she knew exactly why: because the CIO went along with his staff’s sole focus on how their peers were doing. That’s how they got rewarded.)

She showed me a brief report card her firm had designed, that combined funded ratio measures with more traditional relative-to-market performance measures. Had she ever presented this to the IC? Well, actually, no. (Perhaps a failure to consult, or she just did not get the opportunity, I thought. After all, the consultant was rarely invited to the IC meetings, as staff generally made presentations. And when she was invited, there were usually strict time allocations and pre-scrutiny by the CIO. Worth looking into with Goode’s successor.)

And what was the second aspect that Goode regretted? That some other recommendations never saw the light of day, as far as the IC was concerned. The CIO and his staff had, in effect, overruled them. What were those recommendations? One was to restructure the

…the belief of the IC members that they could withstand the psychological impact of the downside had never been tested.


fixed income portfolio, to lengthen its duration so that it became a closer match to the length of the liabilities. That would increase the downside protection, because a greater proportion of the assets and liabilities would move in sync when interest rates moved, and the volatility of the funded ratio would be smaller. Not that this would have helped materially, given the overall aggressiveness of the asset allocation; but ignoring it was a process violation, as far as Goode was concerned. As was another ignored recommendation: to narrow the ranges in which the actual asset allocation could be allowed to fluctuate, as markets moved; otherwise, the central asset allocation policy had little meaning. Again, this wouldn’t have helped in the dramatic collapse of 2008; but it was another process violation.

And a third recommendation actually suggested reducing the degree of aggression somewhat, since markets had had a terrific run for several years and the prospective reward from aggression wasn’t likely to be as great as the fund had enjoyed in the past. But Goode’s heart wasn’t in that one, as she felt sure the IC’s current investment philosophy (to optimize returns) was at odds with her suggested approach, which was aimed at managing risk. That would necessitate a much broader discussion of risk management than time would permit, without which it would just sound like market timing.

Instead of all of these, apparently the staff had substituted their own recommendation: that diversification and potential return should both be improved through investment in alternatives. That had been shelved until another day. But the staff’s focus on peer-relative comparisons was again evident in their substitute recommendation.

I thought about all of this for a couple of days, seeing if I could get any further insights from the CFO and from my inherited staff. I got a lot of color commentary, but nothing deep.

It occurred to me that I had no vested interest in retaining Goode; in fact, given the ire of Colossal’s chairman, it might even be a career-threatening move for me to defend her. But I had learned some useful things from her, like the fact that the governance process worked in a more fragile way in practice than the robust expression of the formal governance matrix suggested. And I also knew that because virtually every fund had been severely affected by the freak events of 2008, perhaps it was unfair for Goode to have to wear the witch’s hat.

So I asked her if she could do some calculations for me that might also help her firm with its defense when the time came to award the next consulting contract. Could she do a

“What if?” calculation, showing, with the benefit of hindsight, the extent to which changes in policy might actually have protected the fund from its huge decline in funded ratio? She said she had already done so by herself, as she wanted to find out whether the fund’s policy was the cause, or whether it was the global financial crisis (GFC) itself that was the demon.

She showed me the results. The carnage all took place in the fourth quarter of 2008. So she asked hypothetically, what would have been the result of making changes as of October 1, 2008? The answer was: not much. Suppose 10% of the fund had been shifted from equities to fixed income. Suppose the 50% currency hedge ratio had been moved to either 0% or 100%, whichever would have worked better. Suppose the TIPS had been replaced by Treasuries. All three changes together would have protected the fall in the funded ratio by less than 10 percentage points. Or, to look at it in a different way, and using a different calculation, the impact of replacing 40 percentage points of equity exposure by fixed income would also have been less than 10 percentage points of funded ratio saved. And that sort of change in asset allocation would have taken far more courage to adopt than holding to the current degree of aggression.

So it really was the fault of the GFC, after all. Well, mostly.


What about the managers? Here I got a lot of color.

It turned out that at the special board meeting, Chairman Rant (that’s how I thought of him: I hadn’t actually met him yet) had gone almost apoplectic on the subject of one of the bond managers. Why had this manager been retained? Wasn’t it obvious that these guys were hopeless? What was Goode’s firm’s process for making such inane judgments?

As the designated witch, Goode had of course been summoned to defend herself and her firm at this meeting. (Her own CEO was there too, because Colossal was a “name” client, and retaining Colossal was important to the firm’s image, even more than to its profits.) It mattered little that the manager had contributed almost nothing to the total fund underperformance, or that Goode’s firm had in general a reasonable track record in selecting managers, or that even a sound process will fail from time to time. What really hurt her was her blunt defense, when she told Rant that he could relate anecdotes till the cows came home, but the plural of anecdote will always be anecdotes, not data. Not a tactful thing to say to the out-of-control chairman of an important client. A better consultant would have bitten her tongue.

That led to a rival firm of consultants being hired to look into Goode’s firm’s manager selection process, a humiliation that inevitably led to the verdict that there were elements in the process that could certainly be strengthened, which really satisfied nobody.

But it did muddy the waters, even more.

I asked the CFO why the board was so concerned with minutiae, which is what specific managers represent in a fund the size of Colossal’s. His response surprised me. He didn’t consider managers to be minutiae. The IC spent most of its time discussing the managers. What, I asked gently, even though their hiring and firing was delegated to me and my staff?

“Certainly,” was his response. “That’s part of our monitoring of how well you’re doing. It’s your main responsibility, isn’t it?” And so they trawled through the voluminous quarterly reports, questioning everything, second-guessing everything—in effect, never really delegating the decisions, because they insisted on being formally notified of every intent to hire or fire, and giving their assent in advance.

Hmm. Another process failure. But not one I could change without a huge amount of education. Perhaps there was more straw, and less brick, in the governance structure than I realized.

Let’s take a look at those quarterly reports.

Data. That’s all, just data. I remember reading somewhere that the essential difference between data and information is human intervention. There had evidently been very little human intervention in the production of these reports. And my staff and the consultant (and my predecessor) were to blame, I thought. This was the nitty-gritty that they themselves loved. It enabled them to keep tabs on the managers and understand what the managers were doing and ensure that the managers were behaving consistently with the styles and decision processes that they had been hired and given a mandate to operate. And so the staff gathered up all their data and proudly, with the best intent in the world, presented it to the IC, in the manner of a child proudly exhibiting homework to a parent: “Look at what I’ve done!”

Not necessarily a good job, I thought, but they didn’t know enough to tell the difference.

But the IC didn’t have a clue what it all meant. How could they tell (without the voice-over) that there was a telltale number on page 17 that was horrifying? It looked no different from the thousand other numbers in the report. And the IC, also with the best intent in the world,

It mattered little that the manager had contributed almost nothing to the total fund underperformance…


but not knowing exactly what it ought to look at, diligently looked at everything, and satisfied itself that it had done a thorough job. Not necessarily a good job, I thought, but they didn’t know enough to tell the difference.

Clearly a big task for me: redesigning the quarterly report with a lot of human intervention, to make information out of data. Tell them what they need to know, and briefly. Make it self-evident what’s going right and what’s going wrong. And (a real challenge) educate them on why a few pages help them do their job far better than a report several inches thick.

How to do that? There’s always a solution. It’s not necessarily optimal, but a solution always exists. I decided to ask Goode and my staff to take the last voluminous report and recast it, to show what a brief monitoring report could look like. (And to incorporate Goode’s funded ratio notion, too.) Then we could show it to the IC and it should be self-evident that it was far more useful. And we could take the old report along too, so that when the IC asked questions, there’d be “deep dive” material available to answer them.

Self-evident? I smiled to myself. Nothing in human nature is that predictable, or that easy. It would be a slog, more likely. But a necessary first step.

So where does all of this leave me? Still worried, but at least with a few ideas. And a few problems I don’t know how to solve. I’ll have to ask the CFO—and the CEO, since he interviewed me before I got the job—for advice.

1. Chairman Rant. We can’t go on this way. He has everyone terrified. His intervention has been wild. And yet it isn’t difficult to understand why he was so upset. It had just become clear to him that Colossal was really two businesses, one with the operations he was familiar with, and another (bigger) one that was a sort of hedge fund, long assets and short liabilities. And he knew nothing about his hedge fund, until it blew up.

One possible solution is to design a quarterly report for him and his board, telling them briefly how their hedge fund business is performing, converting the results to business measures that they are all used to seeing. “No surprises” will be the watchword. And if the board has input for us, well, that’s their right.

A second possibility is to ask Rant to join the IC itself. Or if he doesn’t have the time, then to send a trusted board member. My idea here is that being on the IC will get him closer to research and education. What’s more, he’ll see issues before the fact, before he can second-guess us. He will better appreciate that every event has many possible outcomes, and so risk management becomes vital.

In fact, investing is ultimately all about risk management. We need to reverse the whole thought process on return and risk. Even experts can’t forecast returns with any degree of accuracy or consistency,* so how can they be optimized? So let’s focus on the things we can control, establishing a better governance process and determining an acceptable level of risk. In time, that will get us to a board-level discussion on how much risk the pension fund/hedge fund should be exposed to, as a maximum; and then the board will be in control of Colossal’s hedge fund line of business, too.

Which route is preferable? I don’t know. As I said, I need the advice of the CFO and CEO.

Not necessarily a good job, I thought, but they didn’t know enough to tell the difference.

* “Both statistical models and people have been unable to capture the full extent of future uncertainty and have been surprised by large forecasting errors they did not consider. Expert judgment is typically inferior to simple statistical models.” Dance with Chance: Making Luck Work for You, Makridakis, Hogarth & Gaba. OneWorld Publications, 2009; p. 256.


2. The governance process. Looks great on paper. Looks like a design for a house of bricks. But screwed up in its implementation. And it only needs a little bit of straw to make it capable of being blown away. The governance chain, it turned out in Colossal’s case, was as strong as its weakest link—and there were many of those.

Again, in a way, this is understandable. It turns out that the IC members who established the governance process many years earlier are no longer around. The structure fell into disrepair. Years of healthy capital markets made everything look good: as they say, a rising tide floats all boats. With no background training, no simple vision to guide them, the IC members got mesmerized by all the moving parts and saw the system as extremely complex; the shiny bits (let’s talk about managers!) captured their attention; and for all their efforts, they had no way to establish control—really, their only mode was reactive. Nobody knew how to steer the bus; but they were all along for the ride.

Solutions? Many parts. One is education: they need to know what they need to know! Once they understand how much (or how little) control they have—hey, they can’t control the capital markets, they can only control their asset allocation and their governance process—let’s get them to focus on the big-picture stuff and genuinely delegate the smaller stuff (even if it’s shiny!) to me and my staff. But that means giving them confidence that they won’t then lose control. Lots of communication, plus carefully designed reports.

3. Relationship with CFO and CEO: Must get them onside. Must understand what they feel their roles are, and what they (really, deep down) expect of me. We must be aligned. I can learn from them. They can learn from me. Let’s be honest: if the three of us become a team, the other members of the IC are simply cats to be herded.

4. The new consultant: I want to be involved in the selection process. Here’s what I expect the role to be. First, a resource: a consultant has time and expertise that the IC and I don’t have. Second, help me in my role as change agent: bring new ideas, teach me, align with me. Third: bring some expertise on the governance process: my Deep ideas are no more than common sense; I want something stronger to rely on. Fourth: sounding board, educator, conveyor of market intelligence and attitudes—in short, my ambassador both to and from the world beyond our fund.

5. My staff: Bless them, they’re dedicated and terrific. But it’s my role to give them direction that’s aligned with the broader purpose of the fund. Paying them on the basis of peer-relative performance is fine for those who are hired to outperform their peers at other similar funds; for others, it creates perverse incentives. They need to understand that their investment role exists in a world that starts with the definition of the liabilities. Get my team pointed in the right direction, then loosen the reins: they’re a great bunch.

In fact, investing is ultimately all about risk management.


Possible questions for study and discussions1. What do you think of Deep’s analyses? What did he think through well, badly or

incompletely? What would you do differently? How would you solve his problems?

2. This case study was presented from Deep’s perspective. How would it look from Rant’s perspective? From Goode’s perspective? From the staff’s perspective? How would you interact with Deep if you were the CFO or CEO?

3. Do you have any views on the pension fund as seen as a hedge fund operated by Colossal? What are the consequences for action?

4. How would you design a quarterly report for the IC? Or for the board? What risk measures would it contain?

5. Deep said nothing about periodically reviewing the governance process in action. It’s often said that what gets measured gets managed; and governance doesn’t get measured. How would you conduct such a review?

6. Do you recognize any similarities (or notable differences) when you compare your situation with those of any of the protagonists in this little drama? What lessons do you learn? What are you grateful for? What would you now do differently?

The case study presented here is hypothetical in nature and is not representative of any actual organization. It is presented for discussion purposes only.

/ p 170

PART 2: A PENSION FUND CASE STUDY FOR DISCUSSION

Not yet a bonfire Well, I’m surprised. Pleasantly surprised. After two years we’re finally making progress.

I mentioned earlier1 that the governance house of Colossal’s pension fund seemed to be mostly brick, but evidently contained enough straw for it to be blown away by the big bad wolf of 2008. We’ve located some of the straw, and lit it. It’s not yet a bonfire; there’s more to be located and burnt, but we’ve replaced some of it with solid brick. That’s satisfying.

And as I guessed, the process has not been straightforward.

One big change proved to be less troublesome than all of us anticipated.

We quickly replaced Ms. Goode and her firm as consultants. Now we have Rob Crewe, the assigned consultant from another big-name consulting firm, supported by a team of SMEs (subject matter experts: they speak in acronyms). It remains to be seen how much we’ll see of the SMEs, but together they made a pretty picture and spoke well and reassuringly at a presentation to our investment committee (IC). That, along with their good fortune that a market recovery had relieved some of the IC’s anxieties, got them hired.

No, I’m being cynical. Crewe did his homework and showed that he really understood our situation: our plan, guessing why we were angry, identifying the issues that were probably worrying and scaring us. Yes, the IC felt, we can already feel the comfort that he brings. But I’m glad to say that he’s bringing more than just good chemistry; he’s bringing some substance too.

And one person has stepped up to the plate in a way he never did before: our CEO. He hadn’t taken his role on the IC very seriously before. Yes, he was its chair, but he pretty much let the CFO lead the way. Evidently the chairman’s rant at the special board meeting got to him. And he then took command in an understated but no-nonsense way.

He talked with the chairman first, displaying a flair for psychology that nobody had known existed. He never even mentioned the rant. He simply thanked the chairman for opening his eyes to the fact that the pension fund was, in effect, like it or not, a dominant line of business for Colossal and should be run as such. He promised a plan of action that would treat it as a business, would reduce the risk exposure and would be reflected in monthly reports for the board in a familiar business style.

By: Bruce Curwood and D. Don Ezra

July 2010

1 See “Mostly Brick” by Bruce Curwood and D. Don Ezra, Russell Research Viewpoint, July 2010.

Russell Investments // Risk management: A fiduciary’s guidebook // Not yet a bonfire

/ p 171Russell Investments // Risk management: A fiduciary’s guidebook // Not yet a bonfire

Having dealt with that, he then summoned the CFO, and me and Crewe and gave us instructions. Find a way to reflect the position of the pension fund in familiar corporate accounting terms: balance sheet, profit and loss statement, cash flow and so on. “Don’t argue about measurement, just get it done. It won’t be perfect the first time, but having something reasonably defensible is fine. We’ll improve it as we go along.”

This reporting framework actually turned out not to be too difficult for the CFO to prepare. But as our small group saw the pension fund in a corporate financial perspective, all sorts of things became apparent. The size of the risk. The nature of the risk. How little we had control over. The inadequacy of the resources devoted to what was Colossal’s dominant line of business. And that was just the start.

Hence another bold step by the CEO: an annual retreat for the IC, a day of education. The whole IC needed to understand that perspective, to own it. “Crewe, that’s your assignment. And I don’t want it to be couched in traditional investment and actuarial lingo. You have the good luck that we’ve had a shock and our eyes have been opened. We’re willing to confront the hard questions. Put good stuff in front of those open eyes. Benjamin Disraeli said that ‘there’s no education like adversity’. We’ve had the adversity. But it’s only half of what we need. Now complete the other half, and educate us about how we can run this pension business better. And don’t go off in some esoteric direction. Check in with me a couple of times so I agree that the direction of the day is appropriate.”

Yes, sir!

Cut to the chase. The initial retreat was a success, really. Certainly none of us could have guessed how we would emerge from it. Because nothing went smoothly. I suppose that’s what happens when groups tackle big issues together for the first time: people discover they’re all different and they don’t share the same views. But by the end of the day the verdict was: we wish we had done this before the meltdown.

And so I rate it a success.

I’ll try to capture some of the flavor.

Crewe started with: “Are you ready for change?” And then reduced our discomfort by adding: “Wherever I can, with each change, I’ll show you the difference in terms of what it would have done to the outcome.” One member of the IC, a business line manager, protested: “What’s the point? It’ll just be fighting the last battle.” Response:

“The approach will be forward-looking, designed for the next battle. It may add to your confidence by showing what would have happened with the last one.” OK.

Crewe took the opportunity to continue: “Well then, let’s start with the last battle. Here’s an insight from a couple of guys named Connors and Smith, from their book “How Did That Happen? Holding People Accountable for Results the Positive, Principled Way.” They say: ‘When something goes wrong, there is usually something wrong with what I am doing.’ ”

Dubious looks all round. Crewe was going to have to kick-start this. Confession and introverted analysis weren’t going to work with this bunch. “OK, tell me this: what were you trying to achieve?” And, little by little, tentatively at first and then faster as they started to disagree, some answers emerged. And that in itself was revealing, because it suggested that we had too many goals. Yes, we wanted to keep the benefits fully funded. But we also wanted to minimize costs. And outperform our peers. And there were goals associated with accounting measures. And the inevitable “no surprises.”

Forced to pick one, we ended up just where the IC said it had started: with a focus on the funded ratio. The urge to add “but at a reasonable cost” was irresistible. The benefit

We’re willing to confront the hard questions. Put good stuff in front of those open eyes.


level that we aspire to could be beyond our means if we simply matched assets to the liabilities: everyone knew that. That’s why we had to take some risk. Reducing benefits to a level that was deliverable safely couldn’t be the solution: we’d be uncompetitive, unless the whole system changed so that our competitors did the same. Crewe couldn’t resist a sidebar: “Actually, that’s the Dutch system. But let’s move on, because you’re right, that’s not a feasible path on your own. By the way, we’ve just learned a lesson. Your problem was misdiagnosed, last time. You’ve just said again exactly what you said in your investment policy statement, that your main target is the funded ratio. And yet you ignored it, and consistently focused on asset performance only. And then you blamed poor asset performance. Perhaps if you had focused explicitly on the funded ratio, you might have done things a lot differently. As to what you might have done differently, we’ll explore that shortly.”

So we discussed risk. With some prodding, we identified more types of risk than we had considered before, and because it was now clear to us that a key risk arose from mismatching assets and liabilities, we first identified the things in the asset structure that didn’t match the liabilities. Inflation risk; interest rate risk; longevity risk; and that was before we got to the holding of equities, which matched nothing. Then risk in manager selection and in active management. Wearily, we started down the path of implementation going wrong, and operational risks, until there came an explosion: “For heaven’s sake, what’s the point of all this? The more kinds of risk we identify, the less chance we have of leaving here at the end of the day with a solution!”

“Fair enough,” allowed Crewe. “Let’s just say there’s another lesson here. Risk is multidimensional, and it’s not just one number.” And with a transparent attempt at psychology, which didn’t go down well: “You’ve identified that yourselves. Time for a break. Let’s come back and talk about a risk budget.”

It was a good idea to narrow the focus. But in no time we were arguing again among ourselves. And eventually two more lessons emerged. There’s a difference between risk capacity (how much financial risk we could afford to take) and risk attitude (how much we wanted to take). And there are two completely different consequences of risk-taking going wrong: financial loss and reputational loss. What academics called our “risk tolerance” was essentially how we weighed these four aspects in the balance, and since we had never considered them explicitly, our “revealed preferences” came through in the decisions we made. But explicit consideration is always better.

Rightly or wrongly, we resolved to focus on the financial side. That’s when the CFO pulled out the reporting framework he had designed for the board, and explained it to the IC. And Crewe superimposed on it the fund’s position on 1/1/2008 and its position on 1/1/2009, and the P&L for the year 2008 that had gone so spectacularly wrong. This was more like it for the business line managers: this was what they were used to, and could get their teeth into. (“Why didn’t we get this sort of thing before?” Crewe: “That’s one of the lessons we’ve learned. You can see right away how it might have helped in the last battle, to have this month by month. And it’s obviously a tool you’ll want for the next battle.”)

The CFO then showed the integrated financial statements, consisting of business lines plus DB plan, and that was a revelation. He and Crewe had together prepared some “what ifs,” using stress testing and scenario analysis, and the IC devoured these and wanted more. One big conclusion emerged, even without all the detailed analyses they demanded: that the amount of risk required to generate the reward that in turn kept the cost within bounds, was too large for the main operations to bear. And that eventually led to two further conclusions. One was that the benefit levels and the desired cost were incompatible (and we fled from any involvement in its resolution). The other was that, even at the cost of higher contributions, we didn’t want to take nearly as much risk ever again.

…it was now clear to us that a key risk arose from mismatching assets and liabilities…


Was that just hindsight? Didn’t the asset/liability simulations show exactly what that trade-off was likely to be? Yes, with two caveats. The risks turned out to be bigger than anybody anticipated. (This was a general observation by now; we knew it wasn’t just our own forecasting model that got everything wrong.) And the asset/liability simulations had been presented to the IC with extreme brevity and with an actuarial and investment slant, whereas these new numbers were in a language the business line managers understood completely. Lesson: communication counts. Without this new translation of the impact of risk, the IC hadn’t really owned the asset strategy before; but now it would. And the direction of the new asset strategy would be to reduce risk.

Crewe suggested that we not make the precise decision on strategy immediately. At a regular session of the IC he would show us what other sponsors were doing: resolving to reduce risk over the long term, but not yet (why lock in the losses?), and reducing it gradually, taking risk off the table as the funded ratio rose to specified levels. (They called this liability-responsive asset allocation, or LRAA—a step up in the acronym stakes from three letters to four.)

That was a good point at which to break for lunch. Unlike the previous break, this one saw the group satisfied that they had achieved something worthwhile: a deeper understanding, a direction they were sure they wanted to follow, and a form of monthly report that enabled them to really monitor what was happening. (Thank goodness! I think the day would have unraveled if we hadn’t made a big breakthrough before lunch.)

After lunch Crewe planned to get us into the nitty-gritty, the modeling, the numbers. “I promise you insights here, too.”

We already knew, from the earlier exercise, that models are inadequate. But that’s all we’ve got, in the absence of a crystal ball. Of course, with a crystal ball there’s no such thing as risk, anyway: you just foresee the future and act to get an advantage from your foresight. In practice, you hire people who give you a better than 50/50 chance of having moderate foresight. That’s active management. But without a crystal ball, without any benefit from active management, the strategic imperative is first to manage risk; with average luck the reward ought to come, for the type of risk the academics call systematic. We resolved to write down explicitly which forms of risk we believe are likely to be rewarded, over the long term, simply by taking them. That sort of belief statement would be an integral part of our future statement of investment policy.

As would a focus on risk management. Risk management is the cornerstone of investing. Other than the relatively small amount we hoped to get from active management, the rewards for systematic risk were completely beyond our ability to control. Risk first; reward follows—so we hope and expect. All we can do is decide how much risk we can take, and what exposure to different forms of risk we choose to tolerate.

By now we were eager to look at risk measures. But this led to disappointment. Crewe and I expected this; it was familiar territory to the two of us, to him because of his work with other large sponsors, and to me from my previous career as a money manager.

Risk measurement systems are expensive. Money managers pay up for them. They have to, both because the systems are so helpful and because the managers are fiduciaries—and they can afford the systems. Sponsors are fiduciaries too, but most simply don’t have a financial budget even close to being large enough to afford the risk systems. Can’t they just get the risk measures on their own portfolios from their money managers, and add them up? Not really: managers use different systems, and to add up the measures they need to have a customized interface written, which is also very expensive. And so even large sponsors talk about the issue much more than they take action.

…without a crystal ball, without any benefit from active management, the strategic imperative is first to manage risk…


Crewe showed us what “big picture” risk measures look like, the sort of strategic risks we had identified earlier: inflation risk, interest rate risk, equity risk, longevity risk. And those were fascinating in themselves. So too were the stress tests and scenario projections he showed us. (“Let’s not get too involved. We’ll look at them in detail at a regular IC session, when you can set the new asset allocation policy. Today this is just to show you the possibilities.”) The idea of not going further, not building a set of bottom-up, holdings-based risk measures to complement the top-down strategic measures, because we didn’t have the budget for it, was a difficult one for the IC to take.

And then the CEO intervened. “You’re right, we really do need those measures. Not just out of curiosity, but because we need tools to run this very important line of business. The new reality, that this is a line of business, needs to permeate our thinking. The size of the business justifies creating a budget for tools of this sort. So my new thinking is: let’s take a look at what’s available, and at what cost.” And so the CFO and I had an assignment to investigate risk systems, with Crewe to do some legwork for us.

After another break, Crewe led us into the final session. Here he planned to focus on process. He introduced this via a Harvard Business Review article that discussed ways in which companies mismanage risk. The article listed six ways: relying on historical data; focusing on narrow measures; overlooking knowable risks; overlooking concealed risks; failing to communicate; and not managing in real time.

Crewe suggested that with our new insights we ought to be able to identify whether our previous processes had succumbed to each of these sorts of failure—and whether (and how) our new determination would protect us.

Focusing on historical measures? Yes, we had done so. The problem was that we would continue to do so, perhaps with better history after the Global Financial Crisis, but what else is there but historical numbers? True, conceded Crewe. But there’s one aspect we hadn’t identified. Even the measures we had used in the past were all focused on looking backward: our quarterly reviews measured backward-looking performance, backward-looking tracking error, and so on. There was not the slightest attempt to ask: with our current portfolio, what are forward-looking risk measures? And if forward-looking measures are too high, what portfolio-related action should we take to bring them back within range? So we knew we had taken risk, and we had measures of how much we had taken; but we never asked how much risk was still embedded in the current portfolio, even if those numbers would themselves be based on backward-looking estimates of risk.

And even with the CFO’s new integrated financial statements, those were still backward-looking statements. Hence the significance of the CEO’s imaginative rationale for investigating forward-looking risk systems.

Focusing on narrow measures? Embarrassment: not just narrow, but not even the one measure (funded ratio) that the IC had proclaimed as most important. Resolution: create a dashboard that has the appropriate focus. And a surprise: by now, for the first time in any group I have worked with, the IC took it as natural that the old quarterly reports, with their focus on peer-relative performance, were an anachronism. Interesting, certainly, and probably worth continuing to report on an exception basis, but hardly fundamental.

Overlooking knowable risks? Yes, we had done that, in spades. The new dashboard should report on the ongoing extent of all the identified strategic risks, as well as (on an exception basis) the hoped-for, bottom-up, holdings-based risk measures.

Overlooking concealed risks? We’d never really know. Being aware of the state of the art was the best we could do. This category is like Donald Rumsfeld’s “unknown unknowns.”

…so even large sponsors talk about the issue much more than they take action.


Failing to communicate? Heavens, yes. This session was the first real communication among the group, and we were discovering things about our attitudes that we never realized before. An annual retreat should become part of the agenda. And the CEO’s line of communication with the chairman about the pension fund, as part of their regular discussions, was another new action step that would have been awfully valuable before.

Finally, not managing in real time. Yes to this too, historically. Going forward we’d get monthly reports, and we resolved that if my staff and I ever felt that events were moving rapidly, I should get together with the CFO and Crewe, make a quick assessment of the situation and, if necessary, hold an emergency IC meeting by conference call. Other actions that might be necessary in real time, like monitoring and replacing managers, were already provided for in our governance structure.

I thought at this stage that Crewe would have us declare the day successful and wind up with a quote he had found that provided an interesting angle on prudence: “A prudent man sees danger and takes refuge, but the simple keep going and suffer for it” (Proverbs 22:3)—a nice contrast between ourselves now and our pre-retreat selves.

But the strangest thing happened. A side conversation between a business line manager and the vice president of human resources got heated and attracted everybody’s attention. Apparently the HR chief was complaining about too much importance being placed on obviously imperfect tools, and the business line manager, an impatient and numerate man, was pooh-poohing the HR chief’s faith in process; and the tools-versus-process debate suddenly got personal and heated. And others joined in and the IC split down the middle.

“Tools are better than process.” “Process is more reliable than tools.” “Good process doesn’t guarantee good results.” It was like being in a school playground. Our new Chief Psychologist, the CEO, called a halt to the debate with a conciliatory “You know, it doesn’t have to be one or the other, we’re fortunate enough to have both process and tools.”

And so ended the retreat, on a discordant note, but still representing much progress.

Where did all of this leave me? With a few things to do and a few new concerns (preferable to the old ones, though!). Here’s what I listed, at the time:

1. Set up an asset allocation session as soon as possible, for the IC. This group is now ready to own the decision, not just rubber-stamp a recommendation.

2. Start investigating risk measurement systems.

3. Understand the CFO’s integrated financial statements much better than I do.

4. Get Crewe to do a mock-up of a dashboard that I can discuss with him and that we can together then take to the CFO and then the CEO.

5. Get Crewe to take a stab at a first draft of a set of belief statements for the investment policy statement, about the risks that ought to lead fairly automatically (at least over the long term) to rewards, and the risks that aren’t expected to bring rewards and therefore ought to be closely managed and if possible eliminated. Also active management: why do we think we in particular can add value, when the average fund doesn’t? Can we test whether our beliefs are defensible? (I’m a bit nervous about this last bit, I admit.)

6. With the new mantra that risk management is paramount, should my staff and I anticipate possible new ways of measuring our contribution? And what does that mean for our remuneration?

Failing to communicate? Heavens, yes.


7. Despite all of us saluting the funded ratio as the new dominant measure, I’m uneasy. What’s our GPS for success? Is there a measure, given that we’re really on a journey to an unknown destination in an indefinite time frame? And if that’s an accurate analogy, how do we know if we’re on track?

Time to stop. I’m becoming too philosophical. Yet even that great philosopher Yogi Berra hinted at this same issue when he said, in his inimitable way, “You’ve got to be very careful if you don’t know where you’re going, because you might not get there.” Indeed!

Time has passed since then. As I said at the start, we’re finally making progress. But that’s a story for another time.

Possible questions for study and discussion1. For what proportion of companies, do you think, does the DB fund constitute a major

operating division? How big is it for your company? What measures would you use to assess its significance? If it isn’t a major operating division for you, how would that affect your attitude and your decisions?

2. Colossal’s IC didn’t take any further the fact that their benefits were insupportable, except at more risk than they could afford. How would you deal with that situation? What’s the appropriate forum for that sort of discussion? Which parties would want to be involved in finding a solution?

3. What kinds of things going wrong would lead to reputational risk? What sorts of actions might guard against them? How would they interact with actions to manage financial risk?

4. How would you decide how much of each type of risk is acceptable? What part does active management risk play in this?

5. How do you think active managers should be compensated? Should measures be risk-adjusted?

Related readingConnors, Roger, and Tom Smith. “How did that happen? Holding people accountable for results the positive, principled way.” Portfolio Hardcover, Knoxville, TN, 2009.

Curwood, Bruce. “Risk management is the cornerstone of investing.” Russell Research Viewpoint, February 2010.

Stulz, René M. “Six ways companies mismanage risk.” Harvard Business Review, March 2009.

The case study presented here is hypothetical in nature and is not representative of any actual organization. It is presented for discussion purposes only.

/ p 177Russell Investments // Risk management: A fiduciary’s guidebook // About the authors

Sorca Kelly-Scholte, FIAManaging Director, Consulting and Advisory Services, EMEA

Shashank Kothare, FIADirector, Investment Communications, EMEA

Janine Baldridge, CFA, CAIAGlobal Head of Consulting and Advisory Services

Bruce Curwood, CFA, CIMADirector, Investment Strategy, Canada

D. Don Ezra, FIA Co-Chairman, Global Consulting, Director, Investment Strategy, Americas Institutional

Bob Collie, FIAManaging Director, Investment Strategy and Consulting, Americas Institutional

About the authorsRUSSELL ASSOCIATES

Nathan Dudley, CFADirector Investment Services

Chris HenselSenior Practice Consultant, Americas Institutional

John H. Ilkiw, CFADirector Global Consulting Practice

George OberhoferSenior Practice Consultant, Americas Institutional

Jennie TyndallClient Portfolio Strategist

FORMER RUSSELL ASSOCIATES

/ p 178

For more information:Call Russell at 800-426-8506or visit www.russell.com/risk

Important information

This document is intended to provide general information about financial topics but not specific advice or market predictions. This material is intended for the use of investment professionals and other institutional investors only, it is not directed at private individuals.There is no guarantee that any stated objectives or expectations in this material will be met.This piece is based on information that we consider reliable, but it is not guaranteed as accurate and complete by Russell Investment Group or its affiliates.Not all papers were written in the U.S. Please note that although we believe the overall concepts and information provided can be useful to institutions globally, results could vary across different regions.Please remember that all investments carry some level of risk, including the potential loss of principal invested. They do not typically grow at an even rate of return and may experience negative growth. As with any type of portfolio structuring, attempting to reduce risk and increase return could, at certain times, unintentionally reduce returns.Diversification does not assure a profit and does not protect against loss in declining markets.Please remember that all investments carry some level of risk. Although steps can be taken to help reduce risk it cannot be completely removed. Copyright © Russell Investments 2010. All rights reserved. This material is proprietary and may not be reproduced, transferred, or distributed in any form without prior written permission from Russell Investments. It is delivered on an “as is” basis without warranty.Russell Investment Group is a Washington, USA corporation, which operates through subsidiaries worldwide, including Russell Investments, and is a subsidiary of The Northwestern Mutual Life Insurance Company.The Russell logo is a trademark and service mark of Russell Investments.First used: August 2010USI-7657-08-12

About Russell Investments

Founded in 1936, Russell Investments is a global financial services firm that serves institutional investors, financial advisers and individuals in more than 40 countries. Over the course of its history, Russell’s innovations have come to define many of the practices that are standard in the investment world today, and have earned the company a reputation for excellence and leadership.

Through a unique combination of wide-ranging and interlinked businesses, Russell delivers financial products, services and advice. A pioneer, Russell began its strategic pension fund consulting business in 1969 and today is trusted by many well-known worldwide institutions for investment advice. The firm has $140 billion in assets under management (as of 6/30/10) in its mutual funds, retirement products, and institutional funds, and is well recognized for its depth of research and quality of manager selection. Russell offers a comprehensive range of implementation services that helps institutional clients maximize their assets. The Russell Indexes calculate over 50,000 benchmarks daily covering 65 countries and more than 10,000 securities.

For more information about Russell Investments and our

products and services, please go to www.russell.com/risk

risk management: a fiduciary's guidebook - national association of

Documents