abcd december 2007

Introduction

According to a report in the Financial Times,

‘European nations are to draw up radical proposalsto improve transparency in financial markets and tochange the way credit rating agencies operate in anattempt to prevent any recurrence of the financialturmoil arising from the credit squeeze.’1

Are transparency in financial markets and betterdesigned rating agencies indeed key to preventing arecurrence of the kind of mess we have been experienc-ing in the world's most developed financial economiesfor these past four months? I intend to take a rompthrough the crisis to see what lessons it holds for poli-cymakers and market participants.

The problems we have recently witnessed across theindustrialised world (but not, as yet, in the emergingmarkets), were created by a 'perfect storm' bringingtogether a number of microeconomic and macroeco-nomic pathologies. Among the microeconomic sys-temic failures were: wanton securitisation, fundamentalflaws in the rating agencies' business model, the pro-cyclical behaviour of marked-to-market leverage (seeAdrian and Shin (2007a,b) and also of the Basel capitaladequacy requirements, privately rational but sociallyinefficient disintermediation, and competitive interna-tional de-regulation. Proximate local drivers of the spe-cific way in which these problems manifested them-selves were regulatory and supervisory failure in the UShome loan market.

In the UK, the problems were aggravated by: 1. a flawed Tripartite arrangement between the

Treasury, the Bank of England and the Financial

Services Authority (FSA) for dealing with finan-cial crises;

2. supervisory failure by the FSA;3. flaws in the Bank of England's liquidity-oriented

open market policies (too restrictive a definitionof eligible collateral and an unwillingness to tryto influence market rates at maturities longerthan overnight, even during periods of seriouslack of market liquidity);

4. flaws in the Bank of England's discount windowoperations (too restrictive a definition of eligiblecollateral; only overnight lending; too restrictivea definition of eligible discount window counter-parties).

Both shortcomings in the Bank of England's operatingarrangements and procedures were due to a flawedunderstanding in that institution of (1) the nature anddeterminants of market (ill)liquidity, of (2) the Bank ofEngland's unique role in the provision of market liquid-ity because of its ability to create unquestioned liquid-ity instantaneously and costlessly, and of (3) the condi-tions under which there is a trade-off between moralhazard (bad incentives for future bank behaviour) andthe ex-post provision of liquidity to (a) markets and (b)specific individual institutions with the aim of prevent-ing unnecessary collateral damage to the financial sys-tem and the real economy.

Among the macroeconomic pathologies that con-tributed to the crisis were the following:

(1) An ex-ante global saving glut, brought about bythe entry of a number of high-saving countries (notablyChina) into the global economy and a global redistrib-ution of wealth and income towards commodityexporters that also had, at least in the short run, high-er propensities to save than the losers from the globalincrease in commodity prices.

(2) Excessive liquidity creation by the world's twoleading central banks, the Fed and (to a lesser extent theECB) reinforced by the desire of many new industrialis-ing and oil and gas exporting countries to limit theappreciation of their currencies vis-à-vis the US dollar.

DECEMBERDECEMBER 20072007

CE

PR

PO

LIC

YIN

SIG

HT

No.

18

POLICY INSIGHT No. 18 abcd

To d o w n l o a d t h i s a n d o t h e r P o l i c y I n s i g h t s v i s i t w w w. c e p r. o r g

Lessons from the 2007 FinancialCrisisWILLEM H. BUITEREuropean Institute, LSE, Universiteit van Amsterdam and CEPR

a

Author’s note: This paper was submitted to the UK Treasury SelectCommittee, as background for my appearance before the Committeeon Tuesday, November 13, 2007.

1 Financial Times, October 8, 2007, ‘EU plans market reforms toavert crisis’

http://www.ft.com/cms/s/0/0c1cae8c-75d5-11dc-b7cb-0000779fd2ac.html

The behaviour of these central banks may be in partrationalised as a response to the Keynesian effectivedemand weaknesses that many feared would resultedfrom (1).

The microeconomic pathologies ofmodern finance

Securitisation

OriginsTraditionally, banks borrowed short and liquid and lentlong and illiquid. On the liability side of the banks' bal-ance sheets, deposits withdrawable on demand andsubject to a sequential service (first come, first served)constraint figured prominently. On the asset side, loans,secured or unsecured, to businesses and householdswere the major entry. These loans were typically held tomaturity by the banks (the 'originate and hold' model).Banks therefore transformed and extended maturity andcreated liquidity. Such a combination of assets and lia-bilities is inherently vulnerable to bank runs by depositholders.

Banks were deemed to be systemically important,because their deposits were a key part of the paymentmechanism for households and non-financial corpora-tions, because they played a central role in the clearingand settlement of large-scale transactions and of secu-rities. To avoid systemically costly failures by banks thatwere solvent but had become illiquid, the authoritiesimplemented a number of measures to protect andassist banks. Deposit insurance was commonly intro-duced, paid for either by the banking industry collec-tively or by the state. In addition, central banks provid-ed lender of last resort (LoLR) facilities to individualdeposit-taking institutions that had trouble financingthemselves.

In return for this assistance and protection, banksaccepted regulation and supervision. This took the formof minimum capital requirements, minimum liquidityrequirements, other prudential restrictions on what thebanks could hold on both sides of their balance sheets,as well as reporting and transparency obligations.

In the 1970s, Fannie Mae (Federal National MortgageAssociation), Ginnie Mae (Government NationalMortgage Association) and Freddie Mac (Federal HomeLoan Mortgage Corporation) began the process of secu-ritisation of residential mortgages. Asset securitisationinvolves the sale of income generating financial assets(such as mortgages, car loans, trade receivables (includ-ing credit card receivables) and leases) by a company(the originator of the financial assets) to a special pur-pose vehicle (SPV). The SPV, which might be a trust ora company, finances the purchase of these assets by theissue of bonds, which are secured by those assets. TheSPV is supposed to be bankruptcy-remote from theoriginator, that is, it has to be an off-balance sheet enti-ty vis-à-vis the originator. Cash-flow securitisationworks in a similar way, as when the UK governmentagreed to create the International Finance Facility whichis supposed to securitise future development aid com-mitments.

Private institutions, especially banks, immediatelytook advantage of these securitisation techniques toliquefy their illiquid loans. The resulting 'originate anddistribute' model had major attractions for the banksand also permitted a potential improvement in the effi-ciency of the economy-wide mechanisms for intermedi-ation and risk sharing. It made marketable the non-marketable; it made liquid the illiquid. There wasgreater scope for trading risk, for diversification and forhedging risk.

Securitisation generally involves the 'tranching' of thesecurities issued against a given pool of underlyingassets or cash flows. The higher tranche has priority(seniority) over the lower tranches. This permits thehighest tranche secured against a pool of high-riskmortgages, say, to achieve a much better credit ratingthan the average of the assets backing all the tranchestogether (the lower tranches, of course, have a corre-spondingly lower credit rating). In addition, various'enhancements' are frequently packaged with the secu-rities. A common example is insurance against defaultrisk, which was obtained from specialised financial insti-tutions, called 'monolines ' that had sprung into beingto enhance the creditworthiness (and credit ratings) ofsecurities issued by US municipalities.

ProblemsThere are three problems associated with securitisation(and the generally associated creation of off-balancesheet vehicles).

1. The greater opportunities for risk trading createdby securitisation not only made it possible tohedge risk better (that is, to cover open posi-tions); it also permitted investors to seek out andtake on additional risk, to further 'unhedge' riskand to create open positions not achievablebefore. When risk-trading opportunities areenhanced through the creation of new instru-ments or new institutions, and when new popu-lations of potential investors enter the risk-trad-ing markets, we can only be sure that the risk willend up with those most willing to bear it. Therecan be no guarantee that risk will end up beingborne by those most able to bear it.

2. The 'originate and distribute' model destroysinformation compared to the 'originate andhold' model. The information destruction occursat the level of the originator of the assets thatare to be securitized. Under the 'originate andhold' model the loan officer collecting the infor-mation on the creditworthiness of the would-beborrower is working for the Principal in theinvesting relationship (the originating bank ornon-bank lending institution). Under the 'origi-nate and distribute' model, the loan officer ofthe originating banks works for an institution(the originating bank) that is an Agent for thenew Principal in the investing relationship (theSPV that purchases the loans from the bank andissues securities against them). With asymmetricinformation and costly monitoring, the agencyrelationship dilutes the incentive for information


CE

PR

PO

LIC

YIN

SIG

HT

No.

18DECEMBER 2007 2

gathering at the origination stage. Reputationconsiderations will mitigate this problem, butwill not eliminate it.

3. Securitisation also puts information in the wrongplace. Whatever information is collected by theloan originator about the collateral value of theunderlying assets and the credit worthiness ofthe ultimate borrower, remains with the origina-tor and is not effectively transmitted to the SPV,let alone to the subsequent buyers of the securi-ties issued by the SPV that are backed by theseassets. By the time a hedge fund owned by aFrench commercial bank sells ABSs (asset backedsecurities) backed by US sub-prime residentialmortgages to a conduit owned by a smallGerman Bank specialising in lending to small andmedium-sized German firms, neither the buyernor the seller of the ABS has any idea as to whatis really backing the securities that are beingtraded.

Partial solutionsThe problems created by securitisation can be mitigatedin a number of ways.

1. Simpler structures. The financial engineering thatwent into some of the complex securitised structuresthat were issued in the last few years before the ABSmarkets blew up on August 9, 2007, at times becameludicrously complex. Simple securitisation involved thepooling of reasonably homogeneous assets, say, resi-dential mortgages issued during a given period with agiven risk profile (e.g. sub-prime, alt-A or prime). Thesewere pooled and securities issued against them weretranched. However, second-tier and higher-tier-securi-tisation then took place, with tranches of securitisedmortgages being pooled with securitised credit-cardreceivables, car loan receivables etc. and tranched secu-rities being issued against this new, heterogeneous poolof securitised assets. Myriad credit enhancements wereadded. In the end, it is doubtful that even the design-ers and sellers of these compounded, multi-tiered secu-ritised assets knew what they were selling, knew its riskproperties or knew how to price it. Certainly the sellersdid not.

There is a simple solution: simpler structures. This willin part be market-driven, but regulators too may putbounds on the complexity of instruments that can beissued or held by various regulated entities. Centralbanks could accept as collateral in repos or at the dis-count window only reasonably transparent classes ofABS.

2. Unpicking' securitisation. This 'solution' is theultimate admission of defeat in the securitisationprocess. A number of American banks with residentialmortgage-backed securities (RMBS) on their balancesheets have been scouring the entrails of the asset poolsbacking these securities and have sent staff to specificaddresses to assess and value the individual residentialproperties. This inversion of the securitisation matrix is,of course, very costly and means that the benefits fromrisk pooling will tend to be ignored. It is an ignomin-ious end for the securitisations involved.

3. Retention of equity tranche by originator. Whenthe originator of the loans is far removed from the ulti-mate investor in the securities backed by these loans,the incentive for careful origination is weakened. Oneway to mitigate this problem is for the originator toretain the 'equity tranche' of securitised and tranchedissues. The equity tranche or 'first-loss tranche' is thehighest-risk tranche - the first port of call when theservicing of the loans is impaired. It could be made aregulatory requirement for the originator of residentialmortgages, car loans etc. to retain the equity tranche ofthe securitised loans. Alternatively, the ownership of theequity tranche could be required to be made publicinformation, permitting the market to draw its ownconclusions.

4. External ratings. The information gap could beclosed or at least reduced by using external rating agen-cies to provide an assessment of the creditworthiness ofthe securitised assets. This has been used widely in thearea of RMBS and of ABS. This 'solution' to the infor-mation problem, however, brought with it a whole slewof new problems.

Rating agencies

A small number of internationally recognised ratingagencies (really no more than three: Standard & Poor's,Moody's and Fitch) account for most of the rating ofcomplex financial instruments, including ABS. They gotinto this business after for many years focusing mainlyon the rating of sovereign debt instruments and of largeprivate corporates. They have been given a formal reg-ulatory role, (which will be greatly enhanced under theabout-to-be-introduced Basel 2 Capital Adequacyregime) because their ratings determine the risk weight-ing of a whole range of assets bank hold on their bal-ance sheets.

Their role raises a number of important issues becauseit creates a number of problems.

Problems1. What do they know? This is a basic but importantquestion. One can imagine that, after many years, per-haps decades, of experience, a rating agency wouldbecome expert at rating a limited number of sovereigndebtors and large private corporates. How would therating agency familiarise itself with information avail-able only to the originators of the underlying loans orother assets and to the ultimate borrowers? How wouldthe rating agency, even if it knew as much about theunderlying assets as the originators/ultimate borrowers,rate the complex structures created by pooling hetero-geneous underlying asset classes, slicing and dicing thepool, tranching and enhancing the payment streamsand making the ultimate pay-offs complex, non-linearfunctions of the underlying income streams? These rat-ings were overwhelmingly model-based. The modelsused tended to be the models of the designers and sell-ers of the complex structures, who work for the issuersof the instruments. The potential for conflict of interestin the design and use of these models is obvious. Inaddition, even honest models tend to be useless duringperiods of disorderly markets, because we have too fewC

EP

RP

OL

ICY

INS

IGH

TN

o.18


DECEMBER 2007 3

observations on disorderly markets to construct reason-able empirical estimates of the risks involved.

2. They only rate default risk. Rating agencies pro-vide estimates of default risk (the probability of defaultand the expected loss conditional on a default occur-ring). Even if default risk is absent, market risk or pricerisk can be abundant. Liquidity risk is one source ofprice risk. As long as the liquidity risk does not mutateinto insolvency risk, the liquidity risk is not reflected inthe ratings provided by the rating agencies. The factthat many 'consumers' of credit ratings misunderstoodthe narrow scope of theses ratings is not the fault of therating agencies, but it does point to a problem thatneeds to be addressed. First, there has to be an educa-tion campaign to make investors aware of what the rat-ings mean and don't mean. Second, the merits of offer-ing (and requiring) a separate rating for, say, liquidityrisk should be evaluated.

3. They are conflicted. Rating agencies are subject tomultiple potential conflicts of interest.

a. They are the only example of an industry wherethe appraiser is paid by the seller rather than thebuyer, even though the buyers is likely to havethe greatest information deficiency.

b. They are multi-product firms that sell advisoryand consulting services to the same clients towhom they sell ratings. This can include sellingadvice to a client on how to structure a securityso as to obtain the best rating and subsequentlyrating the security designed according to thesespecifications.

c. The complexity of some of the structuredfinance products they are asked to evaluatemakes it inevitable that the rating agencies willhave to work closely with the designers of thestructured products. The models used to evaluatedefault risk will tend to be the models designedby the clients. This is not just the problem that'marking-to-model' can become 'marking-to-myth' or 'garbage in, garbage out'. There is thefurther problem that the myth will tend to beslanted towards the interest of the seller of thesecurities to be rated.

Partial solutionsThere is no obvious solution other than 'try harder anddon't pretend to know more than you know' for the firstproblem. The second problem requires better educationof the investing public. The third problem can be miti-gated in a number of ways.

1. Reputational concerns. Reputation is a key assetof rating agencies. That, plus the fear of law suits willmitigate the conflict of interest problem. The funda-mental agency problem cannot be eliminated this way,however. Even if the rating agencies expect to bearound for a long time (a necessary condition for repu-tation to act as a constraint on opportunistic and inap-propriate behaviour), individual employees of ratingagencies can be here today, gone tomorrow. A person'sreputation follows him/her but imperfectly.Reputational considerations are therefore not a fullyeffective shield against conflict of interest materialising.

2. Remove the quasi-regulatory role of the ratingagencies in Basel II and elsewhere. Just as the publicprovision of private goods tends to be bad news, so theprivate provision of public goods leaves much to bedesired ('the best judges money can buy etc.'). The offi-cial regulatory function of private credit risk ratings inBasel I and II should be de-emphasized and preferablyended altogether.

I may get my wish here, because Basel II appearsfatally holed below the waterline. It was long recognisedto have unfavourable macroeconomic stabilisation fea-tures, because the capital adequacy requirements arelikely to be pro-cyclical (see Borio, Furfine and Lowe(2001), Gordy and Howells (2004) and Kashyap andStein (2004)). On top of this, the recent financial tur-moil showed that the two key inputs into Pillar 1, theratings provided by the rating agencies and the internalrisk models of the banks, are deeply flawed.

As regards internal risk models, there are two prob-lems. The first is the unavoidable 'garbage in - garbageout' problem referred to earlier, that makes any quanti-tative model using parameters estimated or calibratedusing past observations useless during times of crisis,when every crisis is different. We have really only hadone instance of a global freeze-up of ABS markets,impairment of wholesale markets and seizure of leadinginterbank markets simultaneously in the US, theEurozone and the UK. Estimates based on a size 1 sam-ple are unlikely to be useful. Second, the use of internalmodels is inherently conflicted. The builders, maintain-ers and users of these models are perceived by the oper-ational departments of the bank as a constraint ondoing profitable business. They will be under relentlesspressure to massage their models to produce the resultsdesired by the bank's profit centres. They cannot beshielded effectively from such pressures. Chinese wallsinside financial corporations are about as effective inpreventing the movement of purposeful messagesacross them, as the original Great Wall of China was inkeeping the barbarians out and the Han Chinese in -that is, utterly ineffective.

3. Make rating agencies one-product firms. Thepotential for conflict of interest when a rating agencysells consultancy and advisory services is inescapableand unacceptable. Even the sale of other products andservices that are not inherently conflicted with the rat-ing process is undesirable, because there is an incentiveto bias ratings in exchange for more business in func-tionally unrelated areas. The obvious solution is torequire any firm offering rating services to provide justthat. Having single-product rating agencies should alsolower the barriers to entry.

4. End payment by the issuer. Payment by the buyer(the investors) is desirable but subject to a 'collectiveaction' or 'free rider' problem. One solution would be tohave the ratings paid for by a representative body forthe (corporate) investor side of the market. This couldbe financed through a levy on the firms in the industry.Paying the levy could be made mandatory for all firmsin a regulated industry. Conceivably, the security issuerscould also be asked to contribute. Conflict of interest isavoided as long as no individual issuer pays for his own


CE

PR

PO

LIC

YIN

SIG

HT

No.

18DECEMBER 2007 4

ratings. This would leave some free rider problems, butshould permit a less perverse incentivised rating processto get off the ground. I don't think it would be neces-sary (or even make sense) to socialise the rating process,say by creating a state-financed (or even industry-financed) body with official and exclusive powers toprovide the ratings.

5. Increase competition in the rating industry.Competition in the rating process is desirable. The cur-rent triopoly is unlikely to be optimal. Entry should beeasier when rating agencies become single-productfirms, although establishing a reputation will inevitablytake time.

The procyclical behaviour of leverage and of theBasel capital adequacy criteria

As documented extensively in a number of contribu-tions by Adrian and Shin (2007a,b), leverage is stronglyprocyclical for financial intermediaries that operatemainly through the capital markets. This includes secu-rities brokers and dealers, hedge funds and investmentbanks but not commercial banks. When assets aremarked-to-market, as regulators increasingly requirethem to be, increases in asset prices therefore tend to beassociated with rising leverage and falling asset priceswith declining leverage. If financial intermediaries werepassive and did not adjust their balance sheets inresponse to changes in net worth caused by changes inthe prices of the assets they hold, leverage would becountercyclical. Higher leverage will put upward pres-sure on asset prices, creating a positive feedback loop.The response of the intermediaries to asset price-changes is therefore systemically destabilising.

Adrian and Shin also document the procyclical behav-iour of the value at risk to equity ratio. A possibleexplanation of the procyclical nature of leverage, givenby Adrian and Shin, is that financial intermediaries tar-get some value at risk to equity ratio, which inducesthem to increase leverage when value at risk fallsbecause of rising asset prices. This of course onlychanges the statement of the puzzle; it does not solveit.

This pattern of procyclical leverage is reinforcedthrough the Basel capital adequacy requirements.Banks have to hold a certain minimal fraction of theirrisk-weighted assets as capital. Credit ratings are pro-cyclical. Consequently, a given amount of capital cansupport a larger stock of assets when the economy isbooming then when it is slumping. This further rein-forces the procyclical behaviour of leverage.

Partial solutionsThere is no convincing explanation as to why financialintermediaries might target their value at risk to equityratio (the 1996 Market Risk Amendment of the Baselcapital accord only prescribes a lower floor for the reg-ulatory capital of banks relative to value at risk2). Nordo we have much insight about the drivers of leveragefor banks and non-bank financial intermediaries. It is,

however, interesting that there is no procyclical (orcountercyclical) behaviour of commercial bank leverage.If the procyclical behaviour of leverage is deemed aproblem, bringing commercial bank regulatory practicesto bear on other banks and non-bank financial institu-tions may deserve consideration.

The procyclical effect of the Basel capital require-ments has been well-documented (see Kashyap andStein (2003)). This undesirable feature (and the morerecent doubts about the quality of the rating processitself) should lead to an immediate re-opening andrethinking of Basel II. It is rather disappointing havingto go back to the drawing board of capital adequacyeven before Basel II has been formally launched, but inview of its manifest flaws, there is no other choice.

Excessive disintermediation

There are no doubt solid economic efficiency reasonsfor taking certain financial activities out of commercialbanks and out of investment banks, and putting themin special purpose vehicles (SPVs), StructuredInvestment Vehicles (SIVs, that is, SPVs investing inlong-term, often illiquid complex securitised financialinstruments and funding themselves in the short-termwholesale markets, including the asset-backed commer-cial paper (ABCP) markets), Conduits (SIVs closely tiedto a particular bank) and a host of other off-balance-sheet and off-budget vehicles. Incentives for efficientperformance of certain tasks, including appropriate riskmanagement, can, in principle, be aligned better in asuitably designed SPV than in a general-purpose com-mercial bank. The problem is that it is very difficult tocome up with any real-world examples of off-balancesheet vehicles that actually appear to make sense onefficiency grounds.

Most of the off-balance sheet vehicles (OBSVs) I amfamiliar with are motivated primarily by regulatory arbi-trage, that is, by the desire to avoid the regulatoryrequirements imposed on banks and other deposit-tak-ing institutions. These include minimal capital require-ments, liquidity requirements, other prudential con-straints on permissible liabilities and assets, reportingrequirements and governance requirements. Others arecreated for tax efficiency (i.e. tax avoidance) reasons orto address the needs of governments and other publicauthorities for off-budget and off-balance sheetfinance, generally to get around public deficit or debtlimits.

OBSVs tend to have little or no capital, little or notransparency and opaque governance. When opaqueinstitutions then invest in opaque financial instrumentslike the ABS discussed earlier, systemic risk is increased.This is reinforced by the fact that much de-jure or de-facto exposure remains for the banks that have spun offthe off-balance-sheet vehicles (the 'sponsoring' banks)to these vehicles . There exists de jure exposure whenthe bank is a shareholder or creditor of the OBSV, whenthe OBSV has an undrawn credit line with the bank orwhen the bank guarantees some of the OBSV's liabili-ties. De-facto exposure exists when, for reputationalreasons, it is problematic for the bank to let an OBSVthat is closely identified with the bank go under.C

EP

RP

OL

ICY

INS

IGH

TN

o.18


DECEMBER 2007 5

2 Regulatory capital should not be less that three times the 10-day,99 percent value at risk.

Banks in many cases appear not to have been fullyaware of the nature and extent of their continued expo-sure to the OBSVs and the ABS they carried on their bal-ance sheets. Indeed the explosion of new instrumentsand new financial institutions so expanded the popula-tions of issuers, investors and securities, that many mar-ket participants believed that risk could not only betraded and shared more widely and in new ways, butthat risk had actually been eliminated from the systemaltogether. Unfortunately, the world of risk is not adoughnut: it does not have a hole in it. All risk sold bysomeone is bought by someone. If the system workswell, the risk ends up being born by those both willingand most able to bear it. Regrettably, it often ends upwith those most willing but not most able to bear it.

Partial solutions Mitigation of the problems created by excessive disin-termediation will be partly market-driven and partlyregulatory.

1. Re-intermediation. Either Conduits, SIVS andother OBSVs are taken back on balance sheet by theirsponsoring banks, or the ABS and other illiquid securi-ties on their balance sheets are sold to the banks. TheOBSVs then either wither away or vegetate at a low levelof activity.

2. Regulation. We can anticipate a regulatoryresponse to the problem of opaque instruments held byopaque OBSVs in the form of reporting requirements,and consolidation of accounts requirements that aredriven by broad principles ('duck tests'), with constantadaptation of specific rules addressing particular insti-tutions and instruments. For instance, if the SingleMaster Liquidity Enhancement Conduit (M-LEC) orSuperfund proposed by JPMorgan Chase, Bank ofAmerica and Citigroup, with the active verbal support ofthe US Treasury, ever gets off the ground, it is question-able whether the US regulators will permit the partici-pating banks to keep it off-balance-sheet for reportingpurposes, including earnings reports. This would not ofcourse, solve the problem that M-LEC, were it to get offthe ground, could be too successful in preventing salesof distressed illiquid assets held by various OBSVs atrock-bottom prices. There is a material risk that theparticipating banks would use M-LEC to buy eachother's bad assets at sweetheart prices. They wouldthus be able to postpone further the marking to marketof these assets at realistic values. This would mean sys-temically costly further delays in the resolution of theparalyzing uncertainty about who has lost how muchthrough what exposure.

Competitive Global Deregulation

Regulators of financial markets and institutions areorganised on a national basis and are, in part, cheer-leaders and representatives of the interests of theirnational financial sectors. While regulation is national,finance is global. The location of financial enterprisesand markets is endogenous; many are very footloose. Athriving financial sector creates jobs and wealth, and isgenerally environmentally friendly. So regulators try toretain and attract financial businesses to their jurisdic-

tions in part by offering more liberal, less onerous reg-ulations. This competition through regulatory standardshas led to less stringent regulation almost everywhere.

There have been occasional reversals in this process.The Sarbanes-Oxley Act of 2002 was a response to thecorporate governance, accounting and reporting scan-dals associated with Enron, Tyco International,Peregrine Systems and WorldCom. It undoubtedly con-tributed to a loss of business for New York City as aglobal financial centre. Because Sarbanes-Oxley compli-ance is mainly a matter of box-ticking (like most real-world compliance, especially compliance originating inthe USA), it has not materially improved the informa-tional value of accounting or the protection offered toinvestors.

Is this global competitive deregulation process a wel-come antidote to a tendency to excessive and heavy-handed regulation, or a race to the bottom in whicheveryone loses in the end? I believe the jury is still outon this one, although I am inclined, if pushed, to sug-gest that the following are likely to be true

• Principles-based regulation (allegedly what wehave in the UK) vs. rules-based regulation is anunhelpful distinction. You need both. You needprinciples that spell out the fundamental 'ducktest': (a) Does the institution lend long and bor-row short? (b) Does it lend in illiquid form andborrow in markets that are liquid in normal timesalthough they may turn illiquid during period ofmarket turbulence? Do banks have substantialexposure to the institution? If so, it should beeither consolidated for reporting purposes withthe bank or treated as a bank it its own right.Then you also need rules that are constantlyadapted to keep up with developments in instru-ments and institutions.

• Self-regulation is no regulation unless backed upcredibly with the threat that, unless effectiveself-regulation is implemented, external regula-tion will be imposed.

• Voluntary codes of conduct are without signifi-cance unless they can be and are used by theregulator (through 'comply or explain' rules, forinstance) to impose and enforce standards. Thatmeans that if the explanation is not to the reg-ulator's satisfaction, consequences follow andultimately compulsion can be used.

• The UK's 'light-touch' regulation has become'soft-touch' regulation and needs to be tight-ened up in a large number of areas.

Partial solutions1. Greater international cooperation between regula-tors. This is a no-brainer, but very hard to achieve.

2. A single EU-wide regulatory regime for banks,other financial institutions and financial markets.National financial regulators in the EU should go theway of the dodo. An EU-level FSA separate from theESCB would be a good idea, although the central banks(the ECB and, from January 1, 2007, 16 national cen-tral banks) should collect more information about indi-vidual banks than the Bank of England has done since


CE

PR

PO

LIC

YIN

SIG

HT

No.

18DECEMBER 2007 6

it lost banking supervision and regulation in 1997 whenthe Bank became operationally independent for mone-tary policy. Adequate information on the liquidity posi-tions of systemically important banks and other finan-cial institutions should be collected routinely by all cen-tral banks.

3. A crackdown on "regulators of convenience".This requires tough measures towards 'regulationhavens', some found in the Caribbean, others closer tothe UK. One effective approach would be the non-recognition and non-enforceability of contracts, courtjudgements and other legal and administrative rulingsmade by non-compliant jurisdictions.

The global macroeconomic setting

The macroeconomic background to the crisis is the'Great Moderation' - the low and stable global inflationand the high and stable global real GDP growth of thepast decade. Actually, this moderation is more apparentfrom the inflation figures than from the GDP figures.Figure 1 shows the spectacular decline and recent sta-bility of global inflation.

Figure 2 demonstrates two points. First, the stabilityof global GDP growth does not appear to haveincreased since the early 1980s. Second, the commonbelief that global growth over the past 4 years has aver-aged over 5 percent is based on the wrong statistics -that is, on data that weigh national GDPs at purchasingpower parity (PPP) exchange rates rather than marketexchange rates. PPP exchange rates are the best con-version factors if comparisons between national stan-dards of living are to be made. To get the best estimateof developments in global economic activity, marketexchange rates should be used. GDP growth at marketexchange rates has averaged around 3.5 percent perannum over the past few years. The difference betweenthe two measures is due to the fact that PPP exchangerates give a much greater weight to developing coun-tries and emerging markets than do market exchangerates. Since emerging markets (China, India, Vietnam,South Africa) have been the fastest growers by far overthe past decade, global growth measured at marketexchange rates has been well below global growthmeasured at PPP exchange rates. The view that globalgrowth has been good but not spectacular is confirmedC

EP

RP

OL

ICY

INS

IGH

TN

o.18


DECEMBER 2007 7

0

5

10

15

20

25

30

35

40

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

2006

%

Global Investment (% of GDP) Inflation, average consumer prices (Annual % change)

Figure 1 Global inflation and investment

Source: IMF World Economic Outlook November 2007

0

1

2

3

4

5

6

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

2006

Ann

ual %

cha

nge

Real GDP growth (PPP) Annual % change Real GDP Growth (Market) Annual % change

Figure 2 World growth at PP and market exchangerates, 1980-2008

Source: IMF World Economic Outlook November 2007

Figure 3 Global saving, investment and real interest rate

Source: Desroches and Francis (2007)

by the observation that by 2006, the global share ofinvestment in GDP was only slightly above its previouspeak value achieved in 1994 (see Figure 1).

Another striking feature of the global macroeconom-ic environment has been the declining level of realinterest rates since 2001, and specifically the markeddecline since the bursting of the tech bubble at the endof 2000. This is shown clearly by Figure 3, which istaken from Desroches and Francis (2007).3

The proximate determinant of the trend decline inglobal real interest rates is an ex-ante saving glut,caused by the rapid growth of new emerging marketslike China, which have extraordinarily high propensitiesto save, and, in more recent years, the global redistrib-ution of wealth and income towards a limited numberof producers of primary energy sources (especially oiland natural gas) and raw materials. For a number ofyears, the absorptive capacity of the beneficiaries couldnot keep up with their new-found wealth, and vastamounts of savings had to be recycled. The extremefinancial conservatism of many of the big savers (inChina, Japan, India, Russia, most South East Asian andLatin American countries and in the Gulf states, theseoften were the central banks) meant that much of theincreased demand for financial assets was directedtowards default-risk-free financial instruments, espe-cially US Treasury bonds. With no response of supply,risk-free real rates fell very low indeed (see Caballero,2006).

In addition, the response of the US monetary author-ities to the bursting of the tech bubble, the continuedliquidity trap in Japan and, for a while also the ratherrelaxed monetary policy in the Euro area resulted inmassive and excessive global liquidity growth, especial-ly from 2003 till the end of 2006. Many rapidly grow-ing and high-saving emerging markets and a number ofkey oil producers (including the 6 members of the GulfCooperation Council) pursued policies of undervaluednominal exchange rates and sterilized intervention,which although only partially effective, resulted in anunprecedented accumulation of foreign exchangereserves and, until recently, growing demand for high-grade sovereign debt instruments.

As a result of this, not only were long-term risk-freenominal and real interest rates extraordinarily low since2003, but unprecedentedly low credit risk spreads (thatis, default risk spreads) prevailed across the board.There was also an explosion of leverage, although inter-estingly enough not in the non-financial corporate sec-tor. Households leveraged up and so did the financialsector. Prima facie, commercial banks did not increasetheir leverage very much. The increased leverage in thefinancial sector took place outside the commercialbanks - in investment banks, hedge funds, private equi-ty funds and a whole range of new financial institutions(SIVs, conduits etc), often using the new securitisation-based financial instruments discussed earlier. It was

insufficiently appreciated, by regulators, by the banksand by the new financial institutions themselves, thatbeing off-balance-sheet for certain regulatory, auditingand reporting purposes, does not mean that there is nosubstantive (and potentially substantial) financial, com-mercial, economic and reputational exposure.

Partial Solutions Low global risk-free real interest rates have been risingsince the end of 2006, as the absorptive capacity of theoil and gas exporters has risen and as central banks atlast lost control of the management of the externalassets acquired in the high-saving emerging markets.The transfer of these resources to sovereign wealthfunds with a much greater willingness to take risk anda thirst for returns, means that at first the incrementalflows, but increasingly also the existing stocks of exter-nal assets are being shifted out of high-grade sovereignobligations and into such things as equity, infrastruc-ture and other riskier but higher-yielding investments,including commodities.

As regards excessive liquidity creation, it looks asthough both Japan and the US may be repeating (or beabout to repeat) the policies of the beginning of thedecade. Japan appears to be sliding back into recession,with renewed deflationary pressures and no prospectsfor an early normalisation of nominal interest rates. TheBernanke Fed has turned out to be more like theGreenspan Fed than I would have expected or hoped,and has, since the crisis started in August 2007, cut theFederal Funds target rate by 100 bps and the primarydiscount rate by 125bps, despite the presence of seriousinflationary pressures. While the exchange rates ofmany oil and gas producers have appreciated somewhatagainst the dollar, there has been considerable interven-tion to keep down the rate of appreciation. The samehas been true in China and India. It looks as though thefoundations for the next global liquidity glut are beinglaid while the world is still struggling with the (market)liquidity crunch that started this summer.

The onset of the financial crisis

Facts can be ignored for a long time, but not forever.The realisation that risk may have been underpriceddawned first in the USA to holders of securities backedby sub-prime mortgages. During the second half of2005, the delinquency rate on these mortgages beganto creep up from a low of 10 percent at an annual rate(see Figure 4).

During 2006, the delinquency rate rose further and byearly 2007 it had reached 15 percent. It became clearthat, because many of the mortgages granted in 2005and 2006 had up-front 'teaser rates', which during2007 and 2008 would reset at much higher levels, therewas only one direction delinquencies were going to go:up.

The prices of sub-prime mortgage credit defaultswaps began to fall late in 2006 (see Figure 5) anddropped like a stone by the middle of the year, indicat-ing higher perceived default risk for the underlyingassets.


CE

PR

PO

LIC

YIN

SIG

HT

No.

18DECEMBER 2007 8

3 Nothing much can be concluded from eyeballing the ex-post sav-ing and investment rates in Figure 4. They are supposed to beidentically equal, and any difference represents just measurementerror.

CE

PR

PO

LIC

YIN

SIG

HT

No.

18


DECEMBER 2007 9

Figure 4 US sub-prime mortgage delinquency ratea

Note: (a) 30+ days delinquent.

Source: Mortgage Bankers Association and Thomson Datastream.

Figure 6 Sterling coporate bond spreadsa

Note: (a) Option-adjusted spreads over goverment bond yields..

Source: Merrill Lynch

Figure 7 Global CDO issuancea

Note: (a) Funded CDOs refer to instruments backed by coporatebonds; unfunded CDOs refer to instruments backed by creditdefault swaps. (b) Unfunded data for September not available

Source: JPMorgan Chase & Co.

Figure 5 Prices of US sub-prime mortgage creditdefault swapsa

Note: (a) 2006 H2 vintage.

Source: JPMorgan Chase & Co.

Figure 8 US$-denominated commercial paper outstanding

Source: Board of Governors of the Federal Reserve.

The widening of credit risk spreads that followed wasnot confined to sub-prime related instruments andinstitutions. As is clear from Figure 6, which shows thebehaviour of Sterling corporate bond spreads by rating,the global underpricing of risk had affected virtuallyevery private financial instrument, and the sovereigninstruments issued by all but a small number of highlycreditworthy sovereigns.

The US sub-prime mortgage crisis was just the triggerof the global crisis. To illustrate, early in 2007, a largeamount of unsecured household debt (consumer credit)had to be written down/off by UK banks.

In August 2007, we say something we had never seenbefore. The simultaneous global freezing up of virtual-ly all wholesale capital markets, including the interbankmarkets, CDO markets, markets for asset-backed-com-mercial paper (ABCP) (where the crisis hit Canada first)and markets for all but the very best asset-backed secu-rities. Global new CDO issuance dropped precipitously(see Figure 7) and it became impossible to roll over out-standing stocks of commercial paper, especially asset-backed commercial paper, which as a result declinedsharply (see Figure 8).

The financial turmoil did not just touch securities andinstitutions associated with sub-prime lending in theUS. The underpricing of credit risk had been a globalphenomenon, and the re-pricing of credit risk, which isby no means over at the time of writing (December 11,2007), has affected other financial markets, as I discussin the Conclusion.

The 'monolines' or credit risk insurers, from the largestones like MBIA and Ambac to smaller ones like ACA,FGIC, Security Capital Assurance found themselves inthe spotlight and under pressure. The value of the cred-it risk enhancement they can provide depends entirelyon their own credit rating. A 'monoline' without a tripleA rating no longer has a viable business model. It istherefore key that they are well-capitalised or arebacked by well-capitalised parents or sponsors who canreplenish their capital should the need arise. That thisis a real issue became clear when at the end ofNovember 2007, two French banks pledged $1.5 bn torecapitalise a small French mononline, CIFG. This isunlikely to be the last rescue of a monoline in this cred-it cycle.

How did the world's leading centralbanks respond to the crisis?

None of the world's leading central banks exactly cov-ered themselves with glory, although some did betterthan others, and the Bank of England probably did theworst job.

The Federal ReserveAfter the crisis erupted on August 9, the Federal Reservedecided to reduce its (primary) discount rate by 50 basispoints from 6.25 percent to 5.75 percent on August 16.This was at best a meaningless gesture. There were noUS financial institutions for whom the differencebetween able to borrow at the discount rate at 5.75 per-cent rather than at 6.25 percent represented the differ-

ence between survival and insolvency; neither would itmake a material difference to banks consideringretrenchment in their lending activity to the real econ-omy or to other financial institutions. It was a reduc-tion in the discount window penalty margin (previously100 basis points) of interest only to institutions alreadywilling and able to borrow there (because they had thekind of collateral normally expected at the discountwindow). It was small subsidy to such banks - a smalltreat for their shareholders.

A possible rationalisation of this action - that it wasa way for the Fed to say 'we feel your pain; we knowand we care', without doing anything substantive, likea cut in the Federal Funds target rate - really makes lit-tle sense, as from a substantive viewpoint, the Fed'saction on August 16 was cheap talk.

Subsequently, on September 18, the Fed cut theFederal Funds Rate by 50 basis points, with a furtherreduction of 25 basis points following on October 31.In both cases, the Discount rate was reduced by thesame amount as the Federal Funds target rate.

The Fed also extended the maturity of loans at thediscount window from overnight to up to one month.It also injected liquidity into the markets at maturitiesfrom overnight to 3-month. The amounts injected weresomewhere between those of the Bank of England(allowing for differences in the size of the US and UKeconomies) and those of the ECB.

Throughout the four months of the crisis, it is diffi-cult to avoid the impression that the Fed is too close tothe financial markets and leading financial institutions,and too responsive to their special pleadings, to makethe right decisions for the economy as a whole.Historically, the same behaviour had characterised theGreenspan Fed. It came as something of a surprise tome that the Bernanke Fed, if not quite a clone of theGreenspan Fed, displays some of the same excess sensi-tivity to Wall Street concerns. There is an always-pres-ent danger of a regulator getting too close to the indus-try it is supposed to be regulating in the public interest.Even if conscious regulatory capture is avoided, the reg-ulator is at risk of internalising the objectives, fears andworldview of the regulated industry to such an extent,that it interferes with the regulator's ability to make animpartial judgement about what actions are most likelyto serve its official mandate.

There can be no doubt in my view that the Fed underGreenspan treated the stability, well-being and prof-itability of the financial sector as an objective in its ownright, regardless of whether this contributed to theirlegal triple mandate of maximum employment, stableprices and moderate long-term interest rates. While theBernanke Fed has but a short track record, its ratherpanicky reaction and actions since August suggest thatit too may have a distorted and exaggerated view of theimportance of the financial sector for macroeconomicstability. Time will tell.

The ECB The European Central Bank injected liquidity bothovernight and at longer maturities on an very large scaleindeed, but with limited success (see Figure 9 below). It


CE

PR

PO

LIC

YIN

SIG

HT

No.

18DECEMBER 2007 10

did not cut the policy rate or its discount rate, but itrefrained from raising rates as it had planned to do, andhad effectively pre-announced following its last pre-cri-sis Governing Council rate-setting meeting on August 2.Since then there have been four more meetings whererates have been kept on hold, but where the rhetoricstrongly hints at a bias towards further rate increases.The longer talk without action persists along these lines,the lower the credibility of the forward-looking state-ments of President Trichet.

The Bank of JapanThe Bank of Japan did nothing in particular, but did itvery well. This is justified if the absence of evidence (ofsignificant exposure of Japanese banks to sub-prime-backed securities or to other devalued financial instru-ments) is indeed evidence of absence (of such expo-sure). There is, unfortunately, a long history ofJapanese banks not owning up to asset impairments,and refusing to write down underperforming assets.Japanese banks continue to be opaque, even by themodest standards of the rest of the banking sectors ofthe advanced industrial countries.

The Bank of EnglandThe Bank of England cut neither its discount rate norits policy rate until December 7, when it cut both by 25basis points. It injected liquidity on a modest scale, atfirst only in the overnight interbank market. Rather latein the day, it reversed this policy and offered to repo at3-month maturity, but subject to an interest rate floor100 basis points above Bank Rate, that is, effectively ata penalty rate. No one came forward to take advantageof this facility.

The Bank now manages the Liquidity Support Facilityfor Northern Rock, although the Treasury is on the hookfor any losses the Bank may suffer through its exposureto the mortgages that it is taking from Northern Rockas collateral for its use of the Facility.

Just before the Northern Rock crisis blew up, on 12September 2007 (in a Paper submitted to the TreasuryCommittee by Mervyn King, Governor of the Bank ofEngland) the Bank told the world the following:

‘…the moral hazard inherent in the provision of expost insurance to institutions that have engaged inrisky or reckless lending is no abstract concept’.

On 13 September 2007, the announcement came thatthe Bank of England, as part of a joint action by HMTreasury, the Bank of England and the FinancialServices Authority (according to the Memorandum ofUnderstanding between these three parties), had bailedout Northern Rock, a specialist mortgage lender, by pro-viding it with a credit line (the purpose-designedLiquidity Support Facility). Without this, Northern Rock,which funds itself mainly in the wholesale markets,would not have been able to meet its financial obliga-tions.

Even today we don't know any of the details of howthis reported credit line is secured, or how any draw-downs of this credit line are collateralised. If Northern

Rock had sufficient collateral eligible for rediscountingat the Bank of England's Standing (collateralised)Lending Facility, it presumably would have done so,rather than invoking this emergency procedure involv-ing the Bank, the FSA and the Treasury. Collateral eligi-ble for rediscounting at the Standing Lending Facilityconsists of sterling and euro-denominated instrumentsissued by UK and other European Economic Area cen-tral governments, central banks and major internation-al institutions rated at least Aa3 (and, exceptionally, USTreasury bonds). Such assets are said to be scarce on thebalance sheet of Northern Rock. The severity of thepenalty rate charged Northern Rock will also be impor-tant in determining the long-term moral hazard dam-age caused by this operation.

The Bank's September 12 Paper recognises conditionswhen this kind of bail out is justified:

"…, central banks, in their traditional lender of lastresort (LOLR) role, can lend "Against good collater-al at a penalty rate" to any individual bank facingtemporary liquidity problems, but that is otherwiseregarded as solvent. The rationale would be thatthe failure of such a bank would lead to seriouseconomic damage, including to the customers ofthe bank. The moral hazard of an increase in risk-taking resulting from the provision of LOLR lendingis reduced by making liquidity available only at apenalty rate. Such operations in this country arecovered by the tripartite arrangements set out inthe MOU between the Treasury, Financial ServicesAuthority and the Bank of England. Because theyare made to individual institutions, they are flexiblewith respect to type of collateral and term of thefacility".

The MOU states in paragraph 14:

14. In exceptional circumstances, there may be aneed for an operation which goes beyond theBank's published framework for operations in themoney market. Such a support operation is expect-ed to happen very rarely and would normally onlybe undertaken in the case of a genuine threat tothe stability of the financial system to avoid a seri-ous disturbance to the UK economy."

It is clear that the conditions for a justifiable LoLR oper-ation, as specified in the MOU and reiterated in theBank's September 12 Paper, were not satisfied.

First, no evidence has been offered to support the fre-quently-heard assertion (from Northern Rock, theTreasury, the Bank of England and the FSA) thatNorthern Rock (total assets £113 bn as of 30 June 2007)suffered just from illiquidity rather than from the threatof insolvency. Delinquencies on its mortgages are saidto be below the average of the UK mortgage lendingindustry, and that indeed is good news. However, theorganisation had followed an extremely aggressive andhigh-risk strategy of expansion and increasing marketshare, funding itself in the expensive wholesale marketsfor 75% of its total funding needs, and making mort-C

EP

RP

OL

ICY

INS

IGH

TN

o.18


DECEMBER 2007 11

gage loans at low and ultra-competitive effective ratesof interest. In the first half of 2007, Northern Rockaccounted for over 40 percent of the gross mortgagelending in the UK, and for 20 percent of the net. It ishard to see how with such a breakneck rate of expan-sion, it is possible to maintain adequate quality controlover the lending process. Creditworthiness vetting musthave slipped - there are limits to the speed of organicgrowth. In addition, the bank reputedly offered mort-gages up to six times annual income, and packages ofmortgage and personal loans adding up to 125 percentof the value of the collateral for the mortgage. Thatseems reckless and an strategy designed to end up withnon-performing loans. There is some information sure-ly in the fact that Northern Rock's share price had beenin steep decline since February of this year, well beforethe financial market turmoil hit.

In my view, the solvency of Northern Rock is a mat-ter still to be determined. As usual, there is no hardinformation to go by.

Second, it is hard to argue that the survival ofNorthern Rock is necessary to avoid a genuine threat tothe stability of the UK financial system, or to avoid aserious disturbance to the economy. The bank is not'too large to fail'. As the fifth largest mortgage lenderin the UK, it is not systemically significant. When all elsefails, the 'threat of contagion' argument can be invokedto justify bailing out even intrinsically rather small fish,but irrational contagion, that is, contagion not justifiedby objective balance sheet and off-balance sheet inter-dependencies, is extremely rare in practice, and couldhave been addressed directly had it, against the odds,occurred, following the insolvency Northern Rock. Witha reasonable deposit insurance arrangement (say oneinsuring personal retail deposits up to £50,000 andcapable of making full payment on the insured deposits

in no more than a couple of working days), NorthernRock could and should have been left to sink or swimon its own, or with any private sector assistance it mightbe able to drum up without the support of the UK tax-payer.

In a well-designed financial system, Northern Rockcould have been taken into public ownership, with thedeposits ring-fenced and distributed swiftly to thedepositors, and with the bank remaining open to man-age existing exposures and commitments. This wouldgive everyone involved time to discover the best longer-term destination for Northern Rock, its assets andstakeholders. The US legal and regulatory frameworkfor dealing with the insolvency of banks has this prop-erty.

Talking tough at first but then providing liquiditysupport to Northern Rock, and describing liquidity sup-port to the markets at longer maturities as creatingmoral hazard (an erroneous view, in my opinion) butsubsequently offering to provide such support after all,have undermined the credibility of the Bank. I believethe Bank recognises this and is taking steps to avoid arecurrence of such mishaps.

One of the ironies (and surprises) of this set of eventsis that despite the contrast between the low-key andsmall-scale interventions of the Bank of England, themassive liquidity injections at all maturities, including 3months, of the ECB,. and the rate cuts and continuedmoderate liquidity injections of the Fed, the effect ofthese policies on one key measure of money market dis-tress, the spread between -3 month Libor (the interbankrate) and the 3-month OIS rate or Overnight IndexedSwaps, is now about the same for sterling, the euro andthe US dollar. Figure 9 makes that clear. The spread ofLibor over the Overnight Indexed Swap rate is a betterindicator of the market's view of default risk plus liq-


CE

PR

PO

LIC

YIN

SIG

HT

No.

18DECEMBER 2007 12

0.00

0.20

0.40

0.60

0.80

1.00

1.20

02/01

/2006

02/02

/2006

02/03

/2006

02/04

/2006

02/05

/2006

02/06

/2006

02/07

/2006

02/08

/2006

02/09

/2006

02/10

/2006

02/11

/2006

02/12

/2006

02/01

/2007

02/02

/2007

02/03

/2007

02/04

/2007

02/05

/2007

02/06

/2007

02/07

/2007

02/08

/2007

02/09

/2007

02/10

/2007

02/11

/2007

02/12

/2007

US Dollar 3-Month Libor-OIS Spread (%) Euro 3-Month Euribor-OIS Spread (%) Sterling 3-Month Libor-OIS Spread (%)

Figure 9 3-month interbank rate – OIS spread, 3 Jan. 2006 – 6 Dec. 2007 (%)

Source: Haver.com & Goldman Sachs International.

uidity risk than the spread of Libor over the policy rate,because over a 3-month horizon, the policy rate can beexpected to change.4 This has obviously been the casefor the Federal Funds target rate since the beginning ofthe crisis.

Lessons to be learnt by the UK authorities

The way the crisis unfolded damaged the prestige andinternational standing of the City of London - thefinancial capital of the world - more than the otherleading financial centres.5 The damage is manageableand remediable, but only if effective steps are taken tocorrect the many manifest weaknesses of the UK finan-cial system that were brought to light by the crisis.

I believe there are thirteen lessons for the UK author-ities.

(1) The Tripartite arrangement between the Treasury,the Financial Services Authority and the Bank ofEngland, for dealing with financial instability is flawed.Responsibility for this design flaw must be laid at thedoor of the man who created the arrangement - the for-mer Chancellor and current Prime Minister, GordonBrown. The Treasury, as the dominant partner in thearrangement, also bears primary responsibility for theway in which the Tripartite arrangement has performedoperationally and continues to perform during this cri-sis.

The main problem with the arrangement is that itputs the information about individual banks in a differ-ent agency (the FSA) from the agency with the liquidfinancial resources to provide short-term assistance to atroubled bank (the Bank of England). This happenedwhen the Bank lost banking sector supervision and reg-ulatory responsibility on being made operationally inde-pendent for monetary policy by Gordon Brown in 1997.It's clear this separation of information and resourcesdoes not work.

There are two solutions. Either the relevant elementsof banking supervision and regulation (those having todo with liquidity management) are returned to the Bankof England, or the FSA is given an uncapped and open-ended credit line with the Bank of England, guaranteedby the Treasury. With discretionary access to liquidresources, the FSA can perform the Lender of LastResort function vis-à-vis individual troubled institu-tions. The Bank of England would of course retain the

Market Maker of Last Resort Function of providing liq-uidity to markets and supporting systemically importantfinancial instruments.

If the Bank were to regain all of banking supervisionand regulation, two deeply political activities, its inde-pendence would be jeopardised, especially its opera-tional independence for monetary policy. One solutionto this problem could be to take the Monetary PolicyCommittee out of the Bank of England. The Governorof the Bank of England would no longer be theChairman of the MPC, although I suppose he (or she)could still be an external member. The MPC would justset the target rate for the overnight interbank market.The Bank would act as agent for the MPC in keepingthe overnight rate as closely to the official target aspossible. Anything else (including liquidity-orientedinterventions at maturities longer than overnight, andforeign exchange market intervention) would be theprovince of the Bank of England, not of the MPC.

(2) Three months after the creation of the LiquiditySupport Facility and the granting of deposit insurancecover to Northern Rock (and to any other bank thatmight fight itself in similar circumstances), NorthernRock is still on life support, having drawn over £25 bnfrom the LSF - just under 25 percent of its assets. Thisis a shambles. First, it never should have been necessaryto provide both liquidity support and a deposit guaran-tee for Northern Rock. By effectively guaranteeingaccess to funds for Northern Rock and insuring virtual-ly all unsecured creditors to Northern Rock (and allother UK banks who might find themselves in similarstraights), the UK has socialised all risk to the liabilityside of the banking sector balance sheet.

Several courses of action would have been preferable.They include the following:

(a) Let Northern Rock sink or swim on its ownstrength (i.e. no Liquidity Support Facility), butguarantee all deposits, and ensure a speedy pay-ment of all insured deposits. This would proba-bly have resulted in the insolvency of NorthernRock.

(b) Let Northern Rock sink or swim, but guaranteeall personal retail deposits up to £50,000, andensure a speedy payment of all insured deposits.This would probably have resulted in the insol-vency of Northern Rock and a much smaller runon the bank by depositors than actually tookplace.

(c) Take Northern Rock into public ownership. Thiswould probably result in lawsuits by existingshareholders who feel they should have got abetter deal from the tax payer.

(3) The UK deposit insurance arrangements (which havebeen in place since 1982) are flawed. The amount cov-ered (£2000 outright and 90 percent of the next£33,000) was too low; the deductible for deposits over£2000 was an invitation to run, and the time (alleged-ly up to 6 months) it could take for depositors to gettheir money back was far too long. Responsibility lieswith the Chancellor, although the Bank and FSA couldC

EP

RP

OL

ICY

INS

IGH

TN

o.18


DECEMBER 2007 13

4 An Overnight Indexed Swap is a fixed/floating interest rate swapwith the floating leg tied to a published index of a daily overnightrate reference. The overnight rate is close to the policy rate, so thefixed leg of an OIS swap can be interpreted as the market's expec-tation of the policy rate over a three-month horizon.

5 The damage done by weaknesses in the design of the frameworkfor financial stability and the implementation of policy by thethree key players, the Treasury, the FSA and the Bank of Englandshould not be exaggerated. The position of London as the world'sprimary financial centre is threatened more by its grossly inade-quate transportation infrastructure, its excessive cost of living(especially housing) and sub-standard and/or wildly expensive pri-mary and secondary education facilities than by anything con-nected with the recent financial crisis.

have been better advisors and counsellors to the gov-ernment in these matters. The necessary reforms areobvious.

(4) The FSA did not properly supervise Northern Rock.It failed to recognise the risk attached to NorthernRock's funding model. Stress testing was inadequate.The 'war-games' organised by the three parties to theTripartite arrangement also seem to have suffered froma lack of imagination.

(5) The much-vaunted 'light touch' UK model of regu-lation (based on principles) turned out to be instead ofmodel of 'soft touch' regulation. It is clear that theprinciples vs. rules debate is vacuous. You need both.The principles should state a clear 'duck test'. E.g. if afinancial institution borrows short and lends long, if itborrows liquid (during normal times, but with the risk ofoccasional illiquidity in its usual funding channels) andlends illiquid, and if banks are substantially exposed toit, then it should be regulated like a bank, even if it says'Hedge Fund' on the letterhead. The rules shouldaggressively chase the unceasing attempts, throughinstitutional and instrument innovation, to avoid regu-lation.

(6) Bank insolvency law in the UK is flawed. A bankthat goes into administration has its deposits frozen.The UK needs a US-style arrangement, where the regu-lator can take a threatened bank promptly into publicownership, ring-fence its deposits so they can be trans-ferred within at most two days to the depositors, andreopen the bank immediately to manage its existingactivities and commitments while a longer-term planfor is worked out. Disgruntled existing shareholdersshould have to meet a much tougher test for such banknationalisations to be actionable through lawsuits andother legal remedies.

Provided a troubled and potentially failing bank canbe taken into public ownership, I don't believe there isany need to give banks a dispensation from the lawsgoverning its take-over by, sale to or merger withanother institution. Despite the assertions to the con-trary by the Governor of the Bank of England, the EUMarket Abuse Directive was never an obstacle to anundercover rescue or support operation for NorthernRock.

(7) Following the announcement of the LiquiditySupport Facility, there should have been a joint appear-ance by the Prime Minister, the Chancellor of theExchequer, the Governor of the Bank of England, theChairman of the FSA and the CEO of the FSA, lookingsolemn and reliable, and intoning jointly: 'your moneyis safe'. It might not have prevented the banana-repub-lic-style bank run that started on the 14th, but it wouldhave been worth a try.

(8) In case even the joint appearance of the TalkingHeads would not do the job, the Treasury should haveguaranteed the personal retail deposits of NorthernRock at the same time the LSF was announced.

(9) The Bank of England has a flawed liquidity policy,both in the money markets and at the discount window.It accepts as collateral, both at the Standing LendingFacility (discount window), and in liquidity-orientedopen market operations (sale and repurchase agree-ments or repos) only instruments that are already liquid(UK and European Economic Area government bonds,bonds issued by a few highly-rated international organ-isations and, under exceptional circumstances, USTreasury securities). It should emulate the ECB and theFed and accept as collateral also private instruments,including illiquid and non-traded instruments such asmortgages and asset-backed securities. Provided thiscollateral is priced severely or even punitively, and has afurther 'haircut' or discount applied to it, there will beno moral hazard and the Bank can expect not to losemoney.

The Bank does not need to have information superi-or to that available to the private sector, to ensure thatthe prices it pays for illiquid and nontraded securitiesare not excessive. Many auctions, including the reverseDutch auction, are (reservation) price discovery mecha-nisms. With the Bank acting as a monopolistic buyer atthese auctions, it could (provided there is no collusionamong the sellers) cream off most of the surplus overand above the reservation prices of the sellers.

The Bank would not have to form a view on the trueor fundamental value of these securities following theauction either. It could simply hold them on its booksuntil maturity. That's the advantage of being the oneinstitution that is never illiquid.

(10) The Bank should recognise that the spreadbetween, say, three month Libor and the expected pol-icy rate over the three month period (as measured, forinstance, by the spread of 3-month Libor over the fixedleg of the 3-month Overnight Indexed Rate Swap) canreflect liquidity risk premia as well as default risk pre-mia. In its memo to the Treasury Committee ofSeptember 12, it got close to arguing that this spreadreflected just anticipated default risk. That makes nosense.

Liquidity can vanish today, because market partici-pants with surplus liquidity fear that both they them-selves and their potential counterparties will be illiquidin the future (say, three months from now), when theloans would have to be repaid. A credible commitmentby the Central Bank to provide liquidity in the future(three months from now) would solve the problem, butit is apparent that the required credibility simply doesnot exist. Therefore, the only time-consistent solution,in the absence of a credible commitment mechanism, isto intervene today at a three-month maturity.

The Bank of England should aim, through repos atthese longer maturities, to eliminate as much of the'term structure of liquidity risk premia' as possible. Thiscorrects a market failure. It does not create moral haz-ard if the collateral in the repos is priced properly.

Points (9) and (10) assign to the Bank the responsibil-ity to be the Market Maker of Last Resort, to provide thepublic good of liquidity when disorderly markets disrupt


CE

PR

PO

LIC

YIN

SIG

HT

No.

18DECEMBER 2007 14

financial intermediation and threaten fundamentallyviable institutions.

(11) The Bank should lend at the discount window atlonger maturities than overnight. Loans of up to onemonth should be available (properly priced, with a 'shortback and sides' haircut, and at a punitive rate). Givenpoints (9) and (10), the discount window wouldbecome, for all banks and on demand, what theLiquidity Support Facility purpose-built for NorthernRock is now.

(12) Northern Rock should have known about the Bankof England's repo and discount window policy. Giventhese policies, its funding policies were reckless.

No party involved in this debacle comes out of itsmelling of roses. At least the Bank of England appearsto be willing to learn, and even to admit that it madesome errors. We are still waiting for the Treasury toadmit to anything less than perfection.

(13) My last observation concerns the failure of effec-tive Parliamentary scrutiny of and oversight over thelaws, rules, regulations and institutions that brought usthis debacle. Parliament has done little more than snip-ing ex-post at the other principals in this drama.Finger-pointing and blame allocation are not, however,substitutes for effective ex-ante Parliamentary scrutinyof the laws, rules and regulations and institutions, at thepoint that they can still be moulded and shaped. Wherewas Parliament when it could have done some good?

Conclusion: is the sky falling in bothWall Street and Main Street?

When all the relevant lessons have been learnt and allappropriate recommendations implemented, we still willnot have a system in which banks cannot fail or inwhich systemic instability cannot take hold.

Capitalism, based on greed, private property rightsand decentralised decision making, is both cyclical andsubject to bouts of financial manic-depressive illness.There is no economy-wide auctioneer, no enforcer ofsystemic 'transversality conditions' to rule out periodicexplosive bubble behaviour of asset prices in speculativemarkets. It's unfortunate, but we have to live with it.The last time humanity tried to do away with theseexcesses of capitalism, we got central planning, and weall know now how well that worked. Hayek and Keyneswere both right.

Regulation should try to curb some of the more egre-gious excesses of a decentralised capitalist market econ-omy, but without killing the goose that lays the goldeneggs. In the UK, the pendulum towards de-regulationand self-regulation has probably swung too far. It will,however, be difficult to tighten up unilaterally, as busi-ness would no doubt be lost to other jurisdictions withmore relaxed standards. Regulation of financial marketsand institutions at the EU level would be a major stepforward. After that, intergovernmentalism, that is,cooperation between national (or supranational) regula-tors and tax authorities, will have to take over, to stop

the regulatory race to the bottom from discreditingfinancial globalisation altogether.

The present financial crisis has not yet run its course.This is clear from Figure 9, which shows the spread ofthe 3-Month Interbank Rates over the OIS rate in theUS, the Euro Area and the UK rising sharply again fromthe middle of November until the time of writing (11December 2007). Both the sterling spread and the USdollar spread exceed 100 basis points and the eurospread is not much below 90 basis points. These mas-sive 3-Month spreads cannot be attributed to technicalyear-end liquidity effects. They reflect a complete lackof trust and confidence among the leading banks, lacedwith irrational fear and panic. The failure of the threemonetary authorities involved to regularise the opera-tion of the interbank markets through continuous inter-vention in the repo markets at all maturities where irra-tional spreads manifest themselves, is no tribute to theirunderstanding of the issues or to their decisiveness inaddressing them. The central banks dither while themarkets freeze.

The correction of the global underpricing of risk from2003 till the beginning of 2007 will manifest itselfbeyond the US sub-prime residential mortgage markets,the instruments backed by these mortgages and theinstitutions exposed to them. Higher-rated residentialmortgages in the US and in Europe will suffer similarcorrections. So will commercial real estate-backedmortgages and securities backed by them, securitiesbacked by car loans and credit card receivables, andunsecured consumer credit of all kinds. Unsustainableconstruction, housing market and residential lendingbooms occurred not only in the US, but also in the UK,Spain, Ireland, the Baltic states and other CEE countrieslike Bulgaria. Until quite recently, industrial countryequity markets have continued to perform well, unaf-fected by the re-appraisal and repricing of risk that hasshaken many of the other markets for financial instru-ments. Further equity market corrections, in theadvanced industrial countries and certainly in some ofthe more bubbly emerging markets, are due.

There remains pervasive uncertainty about the valueof the credit ratings granted to complex structuredproducts during the period 2003-2006, and about thevalue of the various enhancements to these products,including the credit risk insurance provided by the'monolines'

Sovereign risk is being re-priced. Even within theEurozone, the spread of 10-year Treasury bond yieldsover Bunds has increased from the 10 bps to 20 bpsrange to the 30 bps to 40 bps range for highly indebt-ed, fiscally fragile countries like Greece and Italy.Belgium's spread over 10-year Bunds is now in the 20s.These spreads are likely to widen further when thebudgetary positions of these countries worsen as theEurozone goes into a cyclical downturn.

Emerging market risk continues to be underpriced,especially non-sovereign emerging market risk, a situa-tion that will no doubt be corrected before long.

There are, however, also signs that the outline of asystemic stabilisation and recovery sometime in the sec-ond half of 2008 is beginning to take shape. LeadingC

EP

RP

OL

ICY

INS

IGH

TN

o.18


DECEMBER 2007 15

commercial banks are beginning to put their off-bal-ance-sheet offspring back onto their balance sheets.HSBC's announcement on November 26, 2007, that itwas taking onto its balance sheet $45bn of debt, muchof it mortgage-linked, owned by SIVs it manages is, Ibelieve, a harbinger or things to come.

The apparent failure of the Single Master LiquidityEnhancement Conduit, aka 'Superfund' proposed byCitigroup, JPMorgan Chase and Bank of America, withthe active verbal encouragement of the US Treasury, toget off the ground is another positive sign. It supportsthe view that it is no longer acceptable or possible forprivate financial institutions to avoid the recognition ofcapital losses on assets held in SIVs, conduits and otheroff-balance-sheet vehicles, by selling them to eachother at sweetheart prices. The enforced revelation ofwhere the losses are located will reduce the uncertaintyand fear about counterparty risk that have been killingliquidity in so many markets.

Money from the 'New Global Moneybags' - sovereignwealth funds from the Gulf and from other emergingmarkets like China, Singapore and Russia - is beginningto find its way into some of the depressed financialmarkets. Citigroup announced on November 26, 2007,that it had raised $7.5bn in new capital from the AbuDhabi Investment Authority, albeit at 'junk' rates of 11percent. On December 10, UBS raised $11.5bn worth ofcapital in Swiss francs from Singaporean and GulfSovereign Wealth Funds, again at junk rates. More dealslike this will follow. When the dust settles on this cri-sis, a significant share of the North-American and West-European financial sectors will be owned and controlledby residents of emerging markets, including the emerg-ing sovereigns. This will be accompanied by a shift indiplomatic and political power to the new creditornations.

The monetary authorities of the leading industrialcountries may have learnt their lessons about the pub-lic good nature of market liquidity. While liquidity canbe managed privately, by private financial institutionshoarding liquid assets, this is socially inefficient if itextends beyond the private provision of liquidity fororderly market conditions. It is more likely today that,even in the UK, the monetary authorities are willing andready to do what simple applied welfare economics tellsthem to do: to provide liquidity on a large scale shouldthe need arise, say, because of disorderly conditions insystemically important financial markets. Theannouncements in late November 2007 by the Fed andthe Bank of England about their plans for year-end liq-uidity are an example of this greater official prepared-ness.

Most importantly, the credit boom of 2003-2006 hasnot led to a massive bout of over-investment in physi-cal capital, except in a few emerging markets like China.The only sectoral exceptions in the industrial countriesare residential construction in the US, Spain, Ireland, theBaltics and a few other emerging markets in CEE, andoverexpansion of the financial sector almost everywherein the industrial world. In these countries the contrac-tionary effects of lower residential investment is nowbeing felt (the US) or will be soon (Spain, Ireland, the

Baltics). But in the most systemically important of thesecountries, the US, residential construction accounts forbarely 4.5 percent of GDP. The damage even a com-plete collapse of house prices can do through the resi-dential construction channel is therefore quite limited.

There is therefore little threat of widespread excesscapacity from the 'supply side' of the economy. Thefinancial position (balance sheets and financial deficits)of the non-financial corporate sectors throughout theindustrial world is strong. The bulk of the financialexcess has stayed inside the financial sector or hasinvolved the household sector.

The key question then becomes whether and to whatdegree the decline in housing wealth (in the US) and thegeneral tightening of the cost and availability of creditwill adversely affect household spending in theadvanced industrial countries. While the sign of theeffect is clear - consumption will weaken - its magni-tude is not. The increasing cost and decreasing avail-ability of household credit is likely to affect and con-strain mainly those households wishing to engage innew or additional borrowing. The increased burden ofservicing outstanding household debt, especially unse-cured debt, is as likely to lead to higher defaults as toreduced consumer spending. Personal bankruptcy is,especially in the US, such an easy and relatively painlessoption, that it is the shareholders of the financial insti-tutions that have made the unsecured loans, as much asthe households that took out these loans, that will suf-fer the financial impact of the increased cost anddecreased availability of credit. If these shareholders aretypically not liquidity-constrained, unlike the defaultingborrowers, the net effect on consumption should bemild. There can be further effects on spending throughthe credit channel if, as a result of the write-offs andwrite-downs, the financial institutions whose debt hasbeen defaulted on become capital-constrained and cur-tail further lending. As always, those most affected willbe new would-be borrowers, households and corpo-rates.

It is still likely, in my view, that the economic fall-outfrom the financial crisis will be contained mainly with-in the financial sector. It is clear that, following theoverexpansion of the residential construction sector inthe US and in a few European countries, and followingthe massive overexpansion of the financial sector justabout everywhere in the industrial world during the pastdecade, there is now likely to be a retrenchment in bothsectors, through lower employment, lower profits andlower valuations. From the point of view of the effi-cient allocation of resources in the medium and longterm, the relative (probably even absolute in the shortrun) contractions of the residential construction sectors(in a few countries) and of the financial sectors almosteverywhere in the industrial world, is a desirable devel-opment. For a number of years now, the private returnsin the financial sector have exceeded the social returnsby an ever-growing margin. Too much scarce analyti-cal and entrepreneurial talent has been attracted intoactivities that, while privately profitable and lucrative,were socially zero-sum at best. In the short run, thiscutting down to size of 'Wall Street' and 'the City' will


CE

PR

PO

LIC

YIN

SIG

HT

No.

18DECEMBER 2007 16

inevitably have some negative side effects for MainStreet also. In the medium and long term, however, amore balanced sectoral allocation of the best and thebrightest will be beneficial.

The short-run pain, concentrated in the financial sec-tor, and especially in the banking and investment sectorand its off-balance-sheet offspring, is not suffered insilence. There is an army of reporters and newscastersstanding by to report each groan and moan from everyCEO whose bank has just written down another chunkof careless CDO exposure. But as long as the monetaryauthorities take their mandates seriously - includingtheir duty to act, at a price, as lenders of last resort andmarket makers of last resort - and as long as the grow-ing financial market hysteria does not spread to the realeconomy, the financial market kerfuffle should result inno more than a mild cyclical downturn around a robustupward trend.

References

Adrian, T. and H. S. Shin (2007a), ‘Liquidity, MonetaryPolicy and Financial Cycles’, forthcoming in CurrentIssues in Economics and Finance.

Adrian, T. and H. S. Shin (2007b), ‘Liquidity andLeverage’, Mimeo, Princeton University, September.

Borio, C., C. Furfine and P. Lowe (2001), ‘Procyclicalityof the financial system and financial stability: issuesand policy options’, in ‘Marrying the macro- and micro-prudential dimensions of financial stability’, BIS Papers,no 1, March, pp 1-57.

Caballero, R. J. (2006), ‘On the Macroeconomics ofAsset Shortages’, November 2006. Forthcoming: TheFourth European Central Banking Conference 2006Volume.

Desroches, B. and M. Francis (2007), ‘World Real InterestRates: A Global Savings and Investment Perspective’,Bank of Canada Working Paper 2007-16, March 2007.

Gordy, Michael B. and B. Howells (2004), ‘Procyclicalityin Basel II: Can We Treat the Disease Without Killing thePatient?’, Board of Governors of the Federal ReserveSystem; First draft: April 25, 2004. This draft: May 12,2004.

Kashyap, A. K. and J. C. Stein (2004), ‘Cyclical implica-tions of the Basel II capital standards’, EconomicPerspectives, Federal Reserve Bank of Chicago, FirstQuarter, pp 18-31.

CE

PR

PO

LIC

YIN

SIG

HT

No.

18


DECEMBER 2007 17

The Centre for Economic Policy Research (www.cepr.org), founded in 1983, is a network of over 700researchers based mainly in universities throughout Europe, who collaborate through the Centre in research and itsdissemination. The Centre’s goal is to promote research excellence and policy relevance in European economics.CEPR Research Fellows and Affiliates are based in over 237 different institutions in 28 countries. Because it drawson such a large network of researchers, CEPR is able to produce a wide range of research which not only address-es key policy issues, but also reflects a broad spectrum of individual viewpoints and perspectives. CEPR has made keycontributions to a wide range of European and global policy issues for over two decades. CEPR research may includeviews on policy, but the Executive Committee of the Centre does not give prior review to its publications, and theCentre takes no institutional policy positions.The opinions expressed in this paper are those of the author and notnecessarily those of the Centre for Economic Policy Research.

Willem Buiter is Professor of European Political Economy at the European Institute of the London School ofEconomics and Political Science. He was a member of the Monetary Policy Committee of the Bank of England (1997-2000) and Chief Economist and Special Adviser to the President at the European Bank for Reconstruction andDevelopment (EBRD) (2000-2005). He has held academic appointments at Princeton University, the University ofBristol,Yale University and the University of Cambridge and has been a consultant and advisor to the InternationalMonetary Fund,The World Bank,The Inter-American Development Bank, the EBRD, the European Communities anda number of national governments and government agencies. Since 2005, he is an Advisor to Goldman SachsInternational. He has published widely on subjects such as open economy macroeconomics, monetary and exchangerate theory, fiscal policy, social security, economic development and transition economies. He obtained his PhD inEconomics from Yale in 1975.

http://www.bis.org/publ/bispap01a.pdf

http://econ-www.mit.edu/files/171

http://www.bis.org/bcbs/events/rtf04gordy_howells.pdf

abcd december 2007

Documents