stylized facts on security returns - princeton universitymarkus/teaching/fin501/01lecture.pdf ·...

STYLIZED FACTS ON SECURITY RETURNSadopted from Heinz Zimmermann, Elmar Mertens

2

AGENDA

We will look at how security returns behave . . .

. . . across asset classes

. . . compared with their "risk"

. . . once they are grouped into baskets

. . . in relation to the macroeconomy

. . . depending on firm characteristics

. . . with regard to prior performance

. . . when there is new information

. . . and what investment managers get out of them

3

AGENDA










4

AVERAGE RETURNS ON FINANCIAL AND PHYSICAL ASSETSPercent p.a. in U$, average over the 1980s

Source: Malkiel (1996), p. 383

Zero returnin real terms

Historical returns on various asset classes differ considerably

5

TODAY‘S VALUE OF 1$ INVESTED IN 1972Including reinvestment of interests and dividends

$0

$5

$10

$15

$20

$25

1972 1976 1980 1984 1988 1992 1996

S&P 500 (w/Dividends) US Government Bonds (long term) US Government Bonds (1 year)Source: Mertens, Data from Ibbotson Associates

Exponentialgrowth of stock

prices

1y-Bonds earnedless, but grewvery smoothly

Stockaccountdipping

below $1

The long-term gains from the stock market have been astounding

6

TYPICAL RETURNS ON U.S. STOCKS AND BONDSPercent p.a.

Source: Elton and Gruber (1995), p. 20

Do you see a linkbetween average

return and variability?

The variability in returns differs, too

7

AGENDA










8

RISK AND RETURN OF INTERNATIONAL ASSET CLASSESPercent per month

0.00%

0.20%

0.40%

0.60%

0.80%

1.00%

1.20%

1.40%

1.60%

0.00% 2.00% 4.00% 6.00% 8.00% 10.00% 12.00%

Stocks Bonds Money

Source: Mertens, Data from Investment Consulting Group

Average

Standard Deviation

The relation between average return and dispersion is not straightforward

9

RISK AND RETURN OF SOME U.S. STOCK PORTFOLIOSPercent per month

Source: Mertens, Data from Fama and French (1992)

0.6%

0.8%

1.0%

1.2%

1.4%

1.6%

1.8%

4.0% 4.5% 5.0% 5.5% 6.0% 6.5% 7.0% 7.5% 8.0%

Average

Standard Deviation

BACKUP

Could a negativerelationship make

any sense?

For a single asset class, like stocks, there is almost no relationship

10

AGENDA










11

ADDING STOCKS IN ALPHABETIC ORDER TO A PORTFOLIOVolatility of portfolio returns (dispersion around mean) in percent p.a.

15%

17%

19%

21%

23%

25%

27%

ABB

+ Alus

uisse

+ Balo

ise+ C

iba

+ Clar

iant

CS Hold

ing+ E

ms

+ Hold

erban

k

+ Nes

tlé

+ Nov

artis I

+ Nov

artis N

+ Swiss

Life

+ Roc

he

+ SAir

Group

+ Swiss

Re+ S

GS

+ SMH N

+ SMH I

+ Sulz

er+ U

BS

+ Zuri

ch Al

lied

Relationshipnot monotone

Source: Zimmermann

Even a naïve mix of just a few stocks reduces risk considerably

12

COMOVEMENT OF STOCKS WITH MARKETReturns in percent per month

-40%

-30%

-20%

-10%

0%

10%

20%

30%

40%

-20% -10% 0% 10% 20%

SGS vs Market (x-axis)

-50%

-40%

-30%

-20%

-10%

0%

10%

20%

30%

40%

50%

-20% -10% 0% 10% 20%

Credit Suisse vs Market (x-axis)

Correlation:0.3

Correlation:0.8

Source: Mertens

Some stocks move more, other less closely with the market

13

AGENDA










14

RECESSIONS AND THE U.S. STOCK MARKET SINCE 1926Indexed S&P 500 (with dividends), logarithmic scale

Most recessionsaccompanied by falls

in stock market.Which causation?

Source: Mertens, Data from NBER and Ibbotson Associates

There is a well established link between business cycles and stockmarket returns

15

U.S. STOCK MARKET AND THE MACROECONOMYPatterns of yearly returns / changes (different scales)

Annual returnon S&P 500

Yearly averageof 1-monthT-Bill rates

BACKUP

Change inannual GDP

Source: Mertens, Data from Investment Consulting Group

The relation between stock market, GDP and short rates is notstraightforward

16

AGENDA










17

TYPICAL FIRM CHARACTERISTICS

• Size

• Industry affiliation

• Accounting Ratios:

– Price-Earnings

– Book-to-Market

– Price-to-Cash-Flow

– Leverage ratio

– . . .

• Location of Headquarters and the place of major share listing

• Type of securities issued (stock, preferred, bonds, derivatives)

• Type of activities: conglomerate, start-up etc.

• . . .

Accounting Ratios aresupposed to conveygrowth expectations.Note: Most ratios are

scaled prices

Source: Mertens

18

RETURNS AND FIRM SIZEMillion U$, value of $10,000 invested from 1952-1994

Source: O‘Shaugnessy, „What works on Wall Street“, 1996, Figure 2-4

All Stocks

Sorted by Market Capitalization

Small firms have higher returns, but the most extraordinary results applyonly to „micro-caps“

20

AVERAGE RETURNS ON U.S. STOCKS DEPENDING ON SIZE AND B/MPercent per month

Source: Mertens, Data from Fama and French (1992)

tinysmall

midlarge

x-large

Low

Mid

High

0.0%

0.2%

0.4%

0.6%

0.8%

1.0%

1.2%

1.4%

1.6%

1.8%

Firm Size

Book-to-Market

High B/M issimilar to low P/E,it means „value“.The opposite is

„growth“

BACKUP

Small „value“ companies have higher returns

21

PORTFOLIO OF BUYING „VALUE“ AND SELLING „GROWTH“ ´68-´90Percent p.a., U.S. Stocks BACKUP

Source: Lakonishok, Shleifer and Vishny (1994)

R: RecessionD: Down Market

A crash- andrecession-proof

strategy?

For a long time, the performance of buying „value“ companies seemedvery persistent

22

PORTFOLIO OF BUYING „VALUE“ AND SELLING „GROWTH“ ´90sPercent p.a., STOXX Euro Style Indices BACKUP

Source: Mertens, Data from STOXX

-50%

-40%

-30%

-20%

-10%

0%

10%

20%

30%

40%

1998 1999 2000 2001

Many „value“investors got

burned. Do youknow some oftheir names?

Recently, the tide has turned against „value“ investors

23

AGENDA










24

RETURNS TO PREVIOUS 5-YEAR‘S WINNER/LOSER STOCKS (U.S.)Market adjusted returns

Source: DeBondt and Thaler (1985) reproduced in Thaler (1993)

In the long-term, returns of extreme winner/loser tend to reverse

25

RETURN TO BUYING SHORT-RUN WINNER AND SELLING LOSERMarket adjusted return, international sample of stocks

Confidenceinterval (95%)

Source: Rouwenhorst (1998)

Short-run continuations seem to be persistent, too

26

AGENDA










27

REACTION TO ANNOUNCEMENT OF CORPORATE ACTIONS„Event Returns“ when stock splits are announced

Source: Mertens, Chart from Fama, Fischer, Jensen and Roll (1969)

• The paper of Fama,Fischer, Jensenand Roll (1969) isa.k.a. „mother of allevent-studies“

• Hypothesis: Splitsare sign of „goodprofits“

• Fama: „We werelucky. We expectedto have some drift “

Clearly, stocks react to news about a company. And they react swiftly

28

RETURNS AROUND EARNINGS ANNOUNCEMENTS

Source: MacKinlay (1997)

Good News

Bad News

No News

Stocks rise when earnings are good, fall when they are bad and remainflat when the announcement has „no news“

29

LONG-TERM REACTION AFTER NEWSBACKUP

Source: Bernard (1993) following Ball and Brown (1968)

Ball and Brown(1968) preceded thework of FFJR (1969)!

Surprisingly, stocks returns drift even once the news are well known

30

NEWS, RETURNS AND VOLUME: ANALYST RECOMMENDATIONS (A)BACKUP

Source: Womack (1996)

Prices and volume react to analyst recommendations . . .

31

NEWS, RETURNS AND VOLUME: ANALYST RECOMMENDATIONS (B)BACKUP

Source: Womack (1996)

. . . and the reaction is markedly stronger for sell-recommendations

32

WARREN BUFFET: ORACLE OF OMAHA1$ invested in 1964

0

500

1'000

1'500

2'000

2'500

1964 1967 1970 1973 1976 1979 1982 1985 1988 1991 1994 1997 2000

Berkshire Hathaway (Book Value)

S&P 500 (w/Dividends)

That is the„exponential growth“curve we saw earlier!

Source: Mertens, Data from Berkshire Hathaway

Some people leave the market way behind

33

FUNDS UNDERPERFORMING WILSHIRE 5‘000 INDEXPercent of total U.S. general equity funds

Source: Malkiel (1996), p. 215

50%

Only in 8 out of 22years lower than 50%

For the average fund, the odds of beating the market are less than even

34

PERSISTENCE OF INVESTMENT MANAGERS‘ SKILLSProbability of next year‘s ranking in performance

At best, chances evenworse than 1:3!

Initial year‘s fund ranking

Follwing year‘s ranking

„good“

„bad“

„good“

„bad“

Source: Carhart (1997)

Spotting a good manager in the past is not much help

35

PERFORMANCE AND EXPENSES: A CLASSICFrequency distribution of outperformance

Before Expenses

Source: Jensen (1968)

After Expenses

Management fees raise the threshold for a fund‘s performance and dilutethe record even further

36

SUMMING UP: KEY QUESTIONS

• What determines asset prices?

• What is risk and where does it come from?

• Which factors influence the stock market and how?

• What information do we need for pricing?

• How shall we invest?

STYLIZED FACTS ON SECURITY RETURNSby Heinz Zimmermann, Elmar Mertens

37

LIST OF REFERENCES 1/2

Ball/ Brown (1968). „An Empirical Evaluation of Accounting Income Numbers“, Journal ofAccounting Research, pp. 159 - 178Bernard (1993). „Stock Price Reaction to Earnings Announcements“ in Thaler (1985),Advances in Behavioral Finance, Russel Sage Foundation, New York, chapter 11Carhart (1997). „On Persistence in Mutual Fund Performance“, Journal of Finance 52,pp. 57-82DeBondt/ Thaler (1985). „Does the Stock Market Overreact?“, Journal of Finance 40:3,pp. 793-807Elton/ Gruber (1995). Modern portfolio theory and investment analysis, Wiley, New YorkFama/ Fischer/ Jensen/ Roll (1969). „The Adjustment of Stock Prices to New Information“,International Economic Review 10, pp. 1-21Fama/ French (1992). "The cross-section of expected stock returns", Journal of Finance 47,pp. 427-465Jensen (1968). „The Performanceof Mutual Funds in the Period 1945-1964“, Journal ofFinance 23, pp. 389-416Lakonishok/ Shleifer/ Vishny (1994). "Contrarian Investment, Extrapolation, and Risk,"Journal of Finance 49:5, pp. 1541-1578

38

LIST OF REFERENCES 2/2

bLowenstein (1995). Buffet - The Making of an American Capitalist, Random HouseMacKinlay (1997). „Event Studies in Economics and Finance“, Journal of Economic Literature35, pp. 13-39bMalkiel (1996). A Random Walk Down Wall Street, Norton, New YorkbO‘Shaugnessy (1996). What works on Wall Street, McGraw-Hill, New YorkRouwenhorst (1998). „International Momentum Strategies“, Journal of Finance 53:1,pp. 267-284Thaler (1993). Advances in Behavioral Finance, Russel Sage Foundation, New YorkWomack (1996). „Do Brokerage Analyst Recommendations Have Investment Value?“,Journal of Finance 51, pp. 137-167

b : „bed-time“ reading (and still useful in daylight)

stylized facts on security returns - princeton universitymarkus/teaching/fin501/01lecture.pdf ·...

Documents