international investment 2005-2006
DESCRIPTION
International Investment 2005-2006. Professor André Farber Solvay Business School Université Libre de Bruxelles. Notions of Market Efficiency. An Efficient market is one in which: Arbitrage is disallowed: rules out free lunches - PowerPoint PPT PresentationTRANSCRIPT
International Investment 2005-2006
Professor André FarberSolvay Business SchoolUniversité Libre de Bruxelles
PhD 01-1 |2April 22, 2023
Notions of Market Efficiency
• An Efficient market is one in which:– Arbitrage is disallowed: rules out free lunches– Purchase or sale of a security at the prevailing market price is never a
positive NPV transaction.– Prices reveal information
• Three forms of Market Efficiency• (a) Weak Form Efficiency
• Prices reflect all information in the past record of stock prices• (b) Semi-strong Form Efficiency
• Prices reflect all publicly available information• (c) Strong-form Efficiency
• Price reflect all information
PhD 01-1 |3April 22, 2023
Efficient markets: intuition
Expectation
Time
Price
Realization
Price change is unexpected
PhD 01-1 |4April 22, 2023
Weak Form Efficiency
• Random-walk model:– Pt -Pt-1 = Pt-1 * (Expected return) + Random error– Expected value (Random error) = 0– Random error of period t unrelated to random component of any past
period
• Implication:– Expected value (Pt) = Pt-1 * (1 + Expected return)– Technical analysis: useless
• Empirical evidence: serial correlation– Correlation coefficient between current return and some past return– Serial correlation = Cor (Rt, Rt-s)
PhD 01-1 |5April 22, 2023
S&P500 Daily returns
-0.08
-0.06
-0.04
-0.02
0
0.02
0.04
0.06
0.08
-0.08 -0.06 -0.04 -0.02 0 0.02 0.04 0.06 0.08
Return day t
Ret
urn
day
t+1
PhD 01-1 |6April 22, 2023
Semi-strong Form Efficiency
• Prices reflect all publicly available information
• Empirical evidence: Event studies• Test whether the release of information influences returns and when this
influence takes place.
• Abnormal return AR : ARt = Rt - Rmt
• Cumulative abnormal return:• CARt = ARt0 + ARt0+1 + ARt0+2 +... + ARt0+1
PhD 01-1 |7April 22, 2023
Efficient Market Theory
-16-11-6-149
141924293439
Days Relative to annoncement date
Cum
ulat
ive
Abn
orm
al R
etur
n (%
)Announcement Date
PhD 01-1April 22, 2023
Example: How stock splits affect value
05
10152025303540
Month relative to split
Cumulative abnormal return %
-29 0 30
Source: Fama, Fisher, Jensen & Roll
PhD 01-1 |9April 22, 2023
Event Studies: Dividend Omissions
Cumulative Abnormal Returns for Companies Announcing Dividend Omissions
0.146 0.108
-0.72
0.032-0.244-0.483
-3.619
-5.015-5.411-5.183
-4.898-4.563-4.747-4.685-4.49
-6
-5
-4
-3
-2
-1
0
1
-8 -6 -4 -2 0 2 4 6 8
Days relative to announcement of dividend omission
Cum
ulat
ive
abno
rmal
retu
rns
(%)
Efficient market response to “bad news”
S.H. Szewczyk, G.P. Tsetsekos, and Z. Santout “Do Dividend Omissions Signal Future Earnings or Past Earnings?” Journal of Investing (Spring 1997)
PhD 01-1 |10April 22, 2023
Strong-form Efficiency
• How do professional portfolio managers perform?
• Jensen 1969: Mutual funds do not generate abnormal returns
• Rfund - Rf = + (RM - Rf)
• Insider trading
• Insiders do seem to generate abnormal returns
• (should cover their information acquisition activities)
PhD 01-1 |11April 22, 2023
What moves the market
• Who knows?• Lot of noise:
– 1985-1990: 120 days with | DJ| > 5%• 28 cases (1/4) identified with specific event(Siegel Stocks for the Long Run Irwin 1994 p 184)
– Orange juice futures (Roll 1984)• 90% of the day-to-day variability cannot explained by
fundamentals
• Pity financial journalists!
PhD 01-1 |12April 22, 2023
Trading Is Hazardous to Your Wealth(Barber and Odean Journal of Finance April 2000)
• Sample: trading activity of 78,000 households 1991-1997• Main conclusions:
1. Average household underperforms benchmark by 1.1% annually2. Trading reduces net annualized mean returns
Infrequent traders: 18.5% Frequent traders: 11.4%3. Households trade frequently (75% annual turnover)4. Trading costs are high: for average round-trip trade 4% (Commissions 3%, bid-ask spread 1%)
PhD 01-1 |13April 22, 2023
US Equity Mutual Funds 1982-1991(Malkiel, Journal of Finance June 1995)
• Average Annual Return• Capital appreciation funds 16.32%• Growth funds 15.81%• Small company growth funds 13.46%• Growth and income funds 15.97%• Equity income funds 15.66%
• S&P 500 Index 17.52%
• Average deviation from benchmark -3.20% (risk adjusted)
PhD 01-1 |14April 22, 2023
: Excess Return
• Excess return = Average return - Risk adjusted expected return
Risk
Return Expected return
Average return
Risk
PhD 01-1 |15April 22, 2023
Jensen 1968 - Distribution of “t” values for “”115 mutual funds 1955-1964
0
5
10
15
20
25
30
35
-5 -4 -3 -2 -1 0 1 2 3 4
Not significantly different from 0
PhD 01-1 |16April 22, 2023
US Mutual FundsConsistency of Investment Result
Successive Period PerformanceInitial Period Performance Top Half Bottom HalfGoetzmann and Ibbotson (1976-1985)Top Half 62.0% 38.0%Bottom Half 36.6% 63.4%Malkiel, (1970s)Top Half 65.1% 34.9%Bottom Half 35.5% 64.5%Malkiel, (1980s)Top Half 51.7% 48.3%Bottom Half 47.5% 52.5%
Source: Bodie, Kane, Marcus Investments 4th ed. McGraw Hill 1999 (p.118)
PhD 01-1 |17April 22, 2023
Decomposition of Mutual Fund Returns(Wermers Journal of Finance August 2000)
• Sample: 1,758 funds 1976-1994
• Benchmark 14.8%+1%
• Gross return 15.8%
• Expense ratio 0.8%• Transaction costs 0.8%• Non stock holdings 0.4%
• Net Return 13.8%
Stock picking +0.75%No timing abilityDeviation from benchmark +0.55%
Funds outperform benchmark
Not enough to cover costs
PhD 01-1 |19April 22, 2023
• Complicated subject• Theoretically correct measures are difficult to construct• Different statistics or measures are appropriate for different types of
investment decisions or portfolios• Many industry and academic measures are different• The nature of active management leads to measurement problems
Introduction
PhD 01-1 |20April 22, 2023
Dollar-weighted returns• Internal rate of return considering the cash flow from or to investment• Returns are weighted by the amount invested in each stock
Time-weighted returns• Not weighted by investment amount• Equal weighting
Dollar- and Time-Weighted Returns
PhD 01-1 |21April 22, 2023
Text Example of Multiperiod Returns
Period Action
0 Purchase 1 share at $50
1 Purchase 1 share at $53
Stock pays a dividend of $2 per share
2 Stock pays a dividend of $2 per share
Stock is sold at $108 per share
PhD 01-1 |22April 22, 2023
Period Cash Flow
0 -50 share purchase
1 +2 dividend -53 share purchase
2 +4 dividend + 108 shares sold
%117.7)1(
112)1(
5150 21
rrr
Internal Rate of Return:
Dollar-Weighted Return
PhD 01-1 |23April 22, 2023
Time-Weighted Return
%66.553
25354
%1050
25053
2
1
r
r
Simple Average Return:
(10% + 5.66%) / 2 = 7.83%
PhD 01-1 |24April 22, 2023
Averaging Returns
Arithmetic Mean:
n
t
t
nrr
1
Geometric Mean:
1)1(/1
1
nn
ttrr
Text Example Average:
(.10 + .0566) / 2 = 7.81%
[ (1.1) (1.0566) ]1/2 - 1
= 7.83%
Text Example Average:
PhD 01-1 |25April 22, 2023
• Past Performance - generally the geometric mean is preferable to arithmetic
• Predicting Future Returns- generally the arithmetic average is preferable to geometric– Geometric has downward bias
Comparison of Geometric and Arithmetic Means
PhD 01-1 |26April 22, 2023
What is abnormal?Abnormal performance is measured:• Benchmark portfolio• Market adjusted• Market model / index model adjusted• Reward to risk measures such as the Sharpe Measure:
E (rp-rf) / p
Abnormal Performance
PhD 01-1 |27April 22, 2023
• Market timing• Superior selection
– Sectors or industries– Individual companies
Factors That Lead to Abnormal Performance
PhD 01-1 |28April 22, 2023
1) Sharpe Index
rp = Average return on the portfolio
rf = Average risk free rate
= Standard deviation of portfolio returnP
Risk Adjusted Performance: Sharpe
P
fP rr
PhD 01-1 |29April 22, 2023
M2 Measure
• Developed by Modigliani and Modigliani
• Equates the volatility of the managed portfolio with the market by creating a hypothetical portfolio made up of T-bills and the managed portfolio
• If the risk is lower than the market, leverage is used and the hypothetical portfolio is compared to the market
PhD 01-1 |30April 22, 2023
M2 Measure: Example
Managed Portfolio: return = 35% standard deviation = 42%
Market Portfolio: return = 28% standard deviation = 30%
T-bill return = 6%
Hypothetical Portfolio:
30/42 = .714 in P (1-.714) or .286 in T-bills
(.714) (.35) + (.286) (.06) = 26.7%
Since this return is less than the market, the managed portfolio underperformed
PhD 01-1 |31April 22, 2023
2) Treynor Measure
rp = Average return on the portfolio
rf = Average risk free rate
ßp = Weighted average ß for portfolio p
Risk Adjusted Performance: Treynor
p
fp rr
PhD 01-1 |32April 22, 2023
Risk Adjusted Performance: Jensen
3) Jensen’s Measure
p= Alpha for the portfolio
rp = Average return on the portfolio
ßp = Weighted average Beta
rf = Average risk free rate
rm = Avg. return on market index port.
])([ pfmfpp rrrr
PhD 01-1 |33April 22, 2023
Appraisal Ratio
Appraisal Ratio = p / (ep)
Appraisal Ratio divides the alpha of the portfolio by the nonsystematic risk
Nonsystematic risk could, in theory, be eliminated by diversification
PhD 01-1 |34April 22, 2023
It depends on investment assumptions
1) If the portfolio represents the entire investment for an individual, Sharpe Index compared to the Sharpe Index for the market.
2) If many alternatives are possible, use the Jensen or the Treynor measureThe Treynor measure is more complete because it adjusts for risk
Which Measure is Appropriate?
PhD 01-1 |35April 22, 2023
• Assumptions underlying measures limit their usefulness
• When the portfolio is being actively managed, basic stability requirements are not met
• Practitioners often use benchmark portfolio comparisons to measure performance
Limitations
PhD 01-1 |36April 22, 2023
Adjusting portfolio for up and down movements in the market• Low Market Return - low ßeta• High Market Return - high ßeta
Market Timing
PhD 01-1 |37April 22, 2023
Example of Market Timing
***
*
**
*
*
*
**
**
******
* ***
rp - rf
rm - rf
Steadily Increasing the Beta
PhD 01-1 |38April 22, 2023
• Decomposing overall performance into components
• Components are related to specific elements of performance
• Example components– Broad Allocation– Industry– Security Choice– Up and Down Markets
Performance Attribution
PhD 01-1 |39April 22, 2023
Set up a ‘Benchmark’ or ‘Bogey’ portfolio• Use indexes for each component• Use target weight structure
Process of Attributing Performance to Components
PhD 01-1 |40April 22, 2023
• Calculate the return on the ‘Bogey’ and on the managed portfolio
• Explain the difference in return based on component weights or selection
• Summarize the performance differences into appropriate categories
Process of Attributing Performance to Components
PhD 01-1 |41April 22, 2023
)(
&
1
11
11
BiBi
n
ipipi
n
iBiBi
n
ipipiBp
n
ipipip
n
iBiBiB
rwrw
rwrwrr
rwrrwr
Where B is the bogey portfolio and p is the managed portfolio
Formula for Attribution
PhD 01-1 |42April 22, 2023
Contributions for Performance
Contribution for asset allocation (wpi - wBi) rBi
+ Contribution for security selection wpi (rpi - rBi)
= Total Contribution from asset class wpirpi -wBirBi
PhD 01-1 |43April 22, 2023
• Two major problems– Need many observations even when portfolio mean and variance are
constant– Active management leads to shifts in parameters making measurement
more difficult• To measure well
– You need a lot of short intervals– For each period you need to specify the makeup of the portfolio
Complications to Measuring Performance
Passive Portfolio Management
Professor André FarberSolvay Business SchoolUniversité Libre de Bruxelles
PhD 01-1 |46April 22, 2023
Academic Foundations of Passive Investment
• Portfolio Theory (Markowitz 1952)– Benefits of diversification
• Capital Asset Pricing Model (Sharpe, Lintner)– Relationship between expected return and risk
• Market Efficiency (Fama 1970)– Stock prices reflect all available information.
• Mutual Fund Performance (Jensen 1968)– Professionally managed portfolio seem unable to make consistent
abnormal returns
PhD 01-1 |47April 22, 2023
Portfolio Theory
• Portfolio characteristics– expected return– risk (standard deviation)
• Risk determined by covariances• Efficient frontier• If riskless asset: one optimal
portfolio
• Expected return
Risk
PhD 01-1 |48April 22, 2023
Capital Asset Pricing Model
• Equilibrium model, optimal portfolio = market portfolio• Risk of individual security = beta (systematic risk)• Risk - expected return relationship
E(r) = Risk-free rate + Market risk premium x Beta
Beta
Expected return
1
E(rmarket)
PhD 01-1 |49April 22, 2023
Efficient Market Hypothesis (EMH)
• Strong version:“Security prices fully reflect all available information”
• Weaker version:“Prices reflect information to the point where the marginal benefit of
acting on information (the profit to be made) do not exceed the marginal costs”
(Fama 1991)
PhD 01-1 |50April 22, 2023
EMH (continued)
• A theoretical result: – Bachelier (1900) Théorie de la spéculation – Samuelson (1965) Proof that properly anticipated prices fluctuate
randomly.• A vast empirical litterature
– “weak-form tests”: do past returns provide information?– “semistrong-form”: is public information reflected in stock prices?– “strong-form tests”: do stock prices reflect private information?
PhD 01-1 |51April 22, 2023
Implications of the EMH for Investment Policy
• Technical Analysis• Fundamental Analysis• Active Portfolio Management
– market timing– stock selection
PhD 01-1 |52April 22, 2023
Mutual Fund Performances
• Malkiel (Journal of Finance June 1995)• 239 equity funds 1982-1991• Average excess return
Benchmark Portfolio: Wilshire 5000 S&P500 Net returns -0.93% -3.20% Gross returns 0.18% -2.03%
• Persistence of Fund Performance:– Winner: rate of return > median– Percent Repeat Winners: 51.7%
PhD 01-1 |53April 22, 2023
Mutual Fund Performances (cont.)
• Otten and Barms, WP 2000• 506 European equity funds 1991-1998
No Mean Market Funds Return ReturnFrance 99 10.9 12.7Germany 57 13.9 15.3Italy 37 15.2 14.9Netherland 9 22.0 21.0United Kingdom 304 12.3 14.2
PhD 01-1 |54April 22, 2023
EMH: faith or fact?
• All empirical tests based on asset pricing model:Excess return = Realized return - Expected return
• Any test of the EMH is a joined test• Still looking for the Capital Asset pricing model
– anomalies (calendar, size)– missing factors (book-to-market, value vs growth)– time variation of market risk premium– international diversification
PhD 01-1 |55April 22, 2023
From Theory to Practice
• First index fund: 1971 launched by Wells Fargo (Samsonite pension fund) and American National Bank
• Fidelity vs Vanguard• Benchmarking• Asset classes• Expense ratio• Outliers: talent or luck?
PhD 01-1 |56April 22, 2023
Optimal portofolio with borrowing and lending
Optimal portfolio: maximize Sharpe ratio
M
PhD 01-1 |57April 22, 2023
Capital asset pricing model (CAPM)
• Sharpe (1964) Lintner (1965)• Assumptions
• Perfect capital markets• Homogeneous expectations
• Main conclusions: Everyone picks the same optimal portfolio• Main implications:
– 1. M is the market portfolio : a market value weighted portfolio of all stocks
– 2. The risk of a security is the beta of the security:• Beta measures the sensitivity of the return of an individual security to the
return of the market portfolio• The average beta across all securities, weighted by the proportion of each
security's market value to that of the market is 1
PhD 01-1 |58April 22, 2023
Market equilibrium: illustration
Wealth Risk free asset
MarketPortfolio
Firm 1 Firm 2 Firm 3
Optimal portfolio 100% 20% 50% 30%
Alan 10 -10 20 4 10 6
Ben 20 -5 25 5 12.5 7.5
Clara 30 15 15 3 7.5 4.5
Market 60 0 60 12 30 18
PhD 01-1 |59April 22, 2023
Capital Asset Pricing Model
Expected return
Beta
Risk free interest rate
Rj
RM
1βj
jFMFj RRRR )(
PhD 01-1 |61April 22, 2023
Arbitrage Pricing Theory
• Starts from statistical characterization of returns• Consider one factor model for stock returns:
• Rj = realized return on stock j
• E(Rj) = expected return on stock j• F = factor – a random variable E(F) = 0• εj = unexpected return on stock j – a random variable
• E(εj) = 0 Mean 0
• cov(εj ,F) = 0 Uncorrelated with common factor
• cov(εj ,εk) = 0 Not correlated with other stocks (=key assumption)
jjjj FRER )(
PhD 01-1 |62April 22, 2023
Diversification
• Suppose there exist many stocks with the same βj.• Build a diversified portfolio of such stocks.
• The only remaining source of risk is the common factor.
FRER jjj )(
PhD 01-1 |63April 22, 2023
Created riskless portfolio
• Combine two diversified portfolio i and j.• Weights: xi and xj with xi+xj =1• Return:
• Eliminate the impact of common factor riskless portfolio
• Solution:
FxxRExREx
RxRxR
jjiijjii
jjiiP
)())()((
0 jiii xx
ji
jix
ji
ijx
PhD 01-1 |64April 22, 2023
Equilibrium
• No arbitrage condition:• The expected return on a riskless portfolio is equal to the risk-free rate.
Fjji
ii
ji
j RRR
j
Fj
i
Fi RRERRE )()(
At equilibrium:
PhD 01-1 |65April 22, 2023
Risk – expected return relation
jFj RRE )(
FM RRE )(
Linear relation between expected return and beta
For market portfolio, β = 1
Back to CAPM formula:
jFMFj RRERRE )()(
PhD 01-1 |66April 22, 2023
Generalization
• The approach can easily be generalized to several factors
N
jkjkjj FRER )(
PhD 01-1 |67April 22, 2023
Empirical challenges
• Explaining the cross section of returns
• Explaining changes in expected returns
PhD 01-1 |68April 22, 2023
Beta
NoDu
Durb
Oil
Chem
Manu
Telc Util
ShopMone
Other
MktPort
RF
0.00
2.00
4.00
6.00
8.00
10.00
12.00
14.00
16.00
18.00
0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60
Beta
Ave
rage
retu
rn
NoDuDurbOilChemManuTelcUtilShopMoneOtherMktPortRF
PhD 01-1 |69April 22, 2023
Average return vs market beta for 25 FF stock portfolios 1926-2004
Mkt
RF
S1S2
S3S4
S5
GovB
CorpB
BM1
BM2
BM3
BM4
BM5
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
1.80
2.00
-0.20 0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60 1.80
Beta
Mea
n re
turn
Size S: S1 smallest - S5 biggest B/M: BM1 lowest - BM5 highest
Average monthly returns Small BigLowB/M 0.91 1.01 1.08 1.01 0.92 0.99 1.29 1.33 1.26 1.10 0.92 1.18 1.50 1.46 1.30 1.30 1.03 1.32 1.69 1.51 1.40 1.35 1.11 1.41HighB/M 1.83 1.64 1.53 1.46 1.34 1.56 1.45 1.39 1.32 1.24 1.07
PhD 01-1 |70April 22, 2023
Size and B/M
12
34
5
Low B/M
S2
S3
S4
High B/M
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
1.80
2.00
mean
monthly return
%
Size
Value
Low B/MSeries2Series3Series4High B/M
Small
Big
PhD 01-1 |71April 22, 2023
Small BigAv.Ret. 1.45 1.39 1.32 1.24 1.07St.Dev. 9.47 7.59 7.03 6.70 6.45Beta 1.35 1.17 1.15 1.13 1.05
SIZE
Low HighAv.Ret. 0.99 1.18 1.32 1.41 1.56St.Dev. 7.38 6.91 6.75 7.10 8.85Beta 1.17 1.12 1.10 1.13 1.32
B/M
Based on monthly data 192607 200411
File: 25_Portfolios_5x5_monthly.xls
PhD 01-1 |72April 22, 2023
Fama French
Fama French Factors - Annual
-60
-40
-20
0
20
40
60
80
1927
1930
1933
1936
1939
1942
1945
1948
1951
1954
1957
1960
1963
1966
1969
1972
1975
1978
1981
1984
1987
1990
1993
1996
1999
2002
RM SMB HML
PhD 01-1 |73April 22, 2023
Predictability: Interest Rates and Expected Inflation
Sample period (Sample Size) γ
1831-2002 (2,053) -2.073(-3.50)
1831-1925 (1,136) -3.958(-4.58)
1926-1952 (324) 0.114(0.03)
1953-1971 (228) -5.559(-2.57)
1972-2002 (357) -1.140(-1.08)
Schwert, W., Anomalies and Market Efficiency,WP October 2002 http://ssrn.com/abstract_id=338080
tftmt RR
PhD 01-1 |74April 22, 2023
Predictability: D/P
Price/dividend ratio
0.00
10.00
20.00
30.00
40.00
50.00
60.00
70.00
80.00
90.00
100.00
1926
1928
1930
1932
1934
1936
1938
1940
1942
1944
1946
1948
1950
1952
1954
1956
1958
1960
1962
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
PhD 01-1 |75April 22, 2023
Predictability
Nobs 77
R(t+1)=a+b*R(t)+e(t+1)Mean StDev Slope Standerror t R²
Stock 0.1190 0.2050 0.03 0.1154 0.27 0.001Tbill 0.0421 0.0350 0.92 0.0465 19.79 0.838Excess 0.0769 0.2083 0.04 0.1155 0.31 0.001
Excess(t+x) = a + b (D/P)(t) + eHorizon1 year 4.17 1.60 2.61 0.0822 year 8.13 2.26 3.60 0.1473 years 11.27 2.62 4.30 0.2004 years 13.69 2.95 4.64 0.2285 years 15.02 3.21 4.67 0.233
PhD 01-1 |76April 22, 2023
ER(+5)=a+b*(D/P)(t)+e
-1
-0.5
0
0.5
1
1.5
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
D/P
Exce
ss R
etur
n +5