Recap

• Last Week – Market predominantly not random – But pattern of market movements very hard to work out – Fractal markets hypothesis Behavioural • Market dynamics follow highly volatile patterns • “Fractal” dimensions to system with underlying “deterministic” pattern plus noise • Measured by Box Dimension and Hurst Exponent Lecture 06 – Latter covered in addendum to lecture 5 Inefficient Markets Hypothesis • This week – If there are patterns to prices, what are they? • The “Inefficient Markets Hypothesis”

The impossibility of efficiency A Keynesian view

• Key assumption of EMH • Key issue is uncertainty, not risk – can accurately predict the future – Cannot possibly estimate expected returns far into • “The first assumption is complete agreement…” future: – (“investors are assumed to agree on the • “our basis of knowledge for estimating the ten prospects of various investments”; Sharpe 1964) years hence of [an investment] amounts to little... • And this distribution is the true one—that is, it is • those who seriously attempt to make any such the distribution from which the returns we use to estimate are often so much in the minority that test the model are drawn.” (Fama & French 2004) their behaviour does not govern the market.” • Reality—future uncertain – Therefore investors can’t know “fundamental value” – How do investors cope with • Versus essential aspect of CAPM: investors can • Not knowing the future work out “real value” of shares • And yet having to invest? • Share values therefore always speculative • Keynes: they form “conventions” about the future…

A Keynesian view Keynes’s view

– Without knowledge of future, investors develop • Investors profit by picking shifts in confidence: “conventions” to cope with uncertain future. They – “the professional and speculator are ... • “assume that the present is a ... serviceable guide to concerned, not with making superior -term the future… forecasts of the probable yield of an investment over • that the existing state of ... prices ... is based on a its whole life, but with foreseeing changes in the correct summing up of future prospects…; [and] conventional basis of valuation a time ahead of • we endeavor to fall back on the judgment of the the general public… this behaviour... is an inevitable rest of the world which is perhaps better result of an investment market... For it is not sensible informed.” to pay 25 for an investment of which you believe the prospective yield to justify a value of 30, if you also believe that the market will value it at 20 three months hence.” [OREF II] • Markets thus conducted by on immediate behaviour of other speculators, rather than rational calculation:

1 Keynes’s view The “Price system” and Asset Markets

• The Stockmarket as a beauty contest and “the third • Normal micro theory: degree”: – Supply a positive function of price – “… pick out the six prettiest faces … the prize being – Demand a negative function of price awarded to the competitor whose choice most nearly – Supply and demand independent corresponds to the average preferences of the competitors as a whole... We have reached the third • If price rises degree where we devote our intelligences to – Supply rises anticipating what average opinion expects the average – Demand falls opinion to be.” – Tendency towards equilibrium • The practicality of rational calculation?: • But finance markets – “Investment based on genuine long-term expectation is – Supply (of assets, shares) possibly a positive function … scarcely practicable. He who attempts it must surely of price … run greater risks than he who tries to guess better – Demand also a positive function of price: than the crowd how the crowd will behave…”

The “Price system” and Asset Markets The “Inefficient Markets Hypothesis”

• If price of assets (shares, real estate, etc.) rising, • Argument that investors demand also rises – React slowly to news – Buyers hope to buy and sell on a rising market – Under-react and Over-react • The faster the rate of price increase (generally speaking) – Ignore “reversion to the mean” the faster the growth of demand • Series of good reports leads to expectation of more good • Tendency to move away from “equilibrium” (“fundamental news value”, historic price to earnings ratios, etc.) • Firm valuation rises, seen as “” • Price thus destabilises an asset market – rise becomes self-fulfilling; bandwaggon buying • Far-from-equilibrium process means • Firm cannot sustain above sector/economy performance – Overvaluation of popular “growth” indefinitely – Undervaluation of unpopular “value” stocks… • Initial “bad news” reports ignored as firm “reverts to mean” • Finally, “bear” valuations set in; bandwaggon selling – “growth stock” underperforms in medium term

The “Inefficient Markets Hypothesis” The “Inefficient Markets Hypothesis”

• 90% of price variability due to internal dynamics of • Key outcomes of “Inefficient Markets Hypothesis” (IEH) speculators watching other speculators: – Shares with low outperform the market – EMH idea of investors focusing solely upon expected • Opposite of EMH risk/return wrong: – Markets characterised by Instead, • Slow reaction by investors to news speculators watch other • Under and over reaction at different times speculators – Institutional investors behave differently to individuals • Forced by short time horizon to match Index • Advantage for individuals over institutions – Best stocks to buy are ones doing poorly now • Likely to have better growth and lower downside volatility in future

2 The “Inefficient Markets Hypothesis” The “Inefficient Markets Hypothesis”

• Companies with good results now • “Contrarian strategy” of buying poor performers now – Tend to become complacent – Won’t work in short-medium term – Attract competitors • Market over-valuation of “good” companies will give them good short-medium term results – Get high valuations – Will work in medium-long term • Companies with poor results now • Persistent failure of “good” companies to maintain – Face “improve or die” pressure results slows share price rise – If in “dull” industries, don’t face many competitors • “Unexpected” good performance of “poor” – Get low stock market valuations companies • Inversion of future performance results – Yields good – Leads to eventual market revaluation of shares – “Good” results now often followed by poor ones • So non-institutional investors can “outperform the Index” – “Poor” results now often followed by good ones by value & contrarian investment • “Reversion to the mean” – But…

The “Inefficient Markets Hypothesis” The “Inefficient Markets Hypothesis”

• Individual investors don’t necessarily do this • Professional managers: – Self-defeating (“irrational”?) behaviour as well… – “choose portfolios that are excessively close to the • “follow the advice of financial gurus, benchmark they are evaluated against … • Fail to diversify, • To minimise the risk of underperforming this • Actively trade stocks and churn their portfolios, benchmark… • Sell winning stocks and hold on to losing stocks – Herd and select stocks that other managers select, thereby increasing their tax liabilities…” (Shleifer • Again to avoid falling behind and looking bad… 2000 p. 10) – Add … stocks that have recently done well, and – Undermines both EMH and possible gains from – Sell stocks that have recently done poorly, market inefficiency • To look good to investors who are getting end of • Also partly explains market inefficiency year reports on portfolio holdings…” (Shleifer 2000 • As does behaviour of money managers… pp. 12-13)

The “Inefficient Markets Hypothesis” Haugen’s Research

• “Bottom line” of IEM • Main proponent of IEM is Robert (Bob) Haugen – Two major groups of investors – Academic till mid-90s • Fund Managers – Resigned to apply ideas for profit – Short-term horizon forces index following – Published several books between academic & • Individuals commercial career – Behavioural herding causes following of fads • The Inefficient Stock Market – Market inefficiency generates opportunities • The Beast on Wall Street • Fund managers can’t pursue because of short-term monitoring • The New Finance • Individuals tend to miss by “following the crowd” • All detail – Opportunities to profit from “contrarian” investing – Empirical failings of CAPM • Buy high B/M, out of favour sectors, low volatility – Ways to profit from market systematic mispricing • Worse performance over short term possible – All are “contrarian” strategies: • Better performance over medium-long term likely • Buy “out of favour” stocks & profit

3 Haugen’s Research "Excellent" versus "Unexcellent" Companies (76-80)

• Famous book “In Search of Excellence” studied • Excellent companies looked much better than unexcellent companies regarded as excellent in terms of 6 ones: 24 characteristics as of 1980: Asset Growth; Growth; 21.78 22 Market to Book Ratio (favouring high over low); Return on 19.05 20 18.43 Capital, Equity, & Sales 18 16.04 – Ranked companies from “Excellent” to “Unexcellent” on 16 weighted scale of these factors 14 12 • Clayman (1987) checked subsequent performance of two 10 8.62 groups 8 7.09 5.93 6 4.88 • Both excellent and unexcellent reverted to mean… 3.76 4 – Better results from investing in “unexcellent” 2.46 2.49 2 0.62 companies than “excellent” ones: 0 Asset Equity Market- Return on Return on Return Growth Growth Book Ratio Total Capital Equity on Sales

"Excellent Companies" "Unexcellent Companies"

"Excellent" versus "Unexcellent" Companies (81-86) "Excellent" versus "Unexcellent" Companies (81-86)

• Results opposite of what fans of excellent companies • Clayman’s conclusion: expected: – “Over time, company results have a tendency to Cumulative Value of $100 Invested in 1981 regress to the mean

Unexcellent • As underlying economic forces attract new entrants Companies 297.5 • And encourage participants to leave low-return 280 businesses.

230 – Because of this tendency • Companies that have been “good” performers in the 181.6 180 past may prove inferior investments Excellent Companies • While “poor” companies frequently provide superior 130 investment returns in the future.” (1987, p. 63)

80 • Many other similar patterns uncovered by Haugen… 1981 1982 1983 1984 1985 1986

Source: M. Clayman, 1987, Financial Analysts Journal, “In Search of Excellence:The Investor’s Viewpoint.”

Haugen’s Research Haugen’s Research

• Future Returns to Stocks • Cumulative Value of $1 Invested in Various Forms of – Cheap Stocks vs Expensive Value and Growth High Yield • Relative to Current Earnings and Dividends Low Yield 100 High P.E. d d High yield and low price e e t – Stocks ranked and re-ranked t 90 Low P.E. s s to earnings ratio shares do better

e to earnings ratio shares do better e

v S&P 500 v 80 S&P 500 n

• by earnings n I I

1 1 70 $ $

• and as of April of each year. f f o o

60 e – Subsequent performance of cheapest and most e u u l l 50 a a V expensive quartiles then monitored. V

40 e 40 e v v i i t t

a 30 a 30 l l u u

m 20 m 20 u u C C 10 0 0 2 4 0 4 0 2 4 0 6 2 4 8 8 2 6 0 2 8 6 8 2 8 0 6 4 8 6 4 6 7 7 7 8 9 7 7 7 8 8 9 9 6 7 8 8 9 9 6 7 7 8 8 9 7 8 8 9 8 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

Year

4 Haugen’s Research Haugen’s Research

• Relative Performance of Portfolios Equally-weighted in • The Effect of Moving to Lower and Higher Risk the Cheap and Expensive Quartiles Portfolios of NYSE Stocks - 1928-1992

• Difference in cumulative return is measured over rolling 12% 5-year periods. n n Efficient Portfolio r r 11% u u t t

– Relative performance appears to cycle over time. e e R R

10% l 10% l a – But cheap stocks out-perform more often than not a u u S&P 500 n n n • In following graphs, “efficient” means “what works” n 9% A A

d d e – CAPM idea of “efficient portfolio” e 8% z z i i l l a a Inefficient Portfolio e • Efficient means risk-return tradeoff e 7% R – Higher return necessitates higher v R 6% – Actual investing experience 16% 18% 20% 22% 24% 26% 28% 30% • Efficient means lower volatility and higher return Risk

Haugen’s Research Haugen’s Research

• The Effect of Moving to Lower and Higher Risk • The Effect of Moving to Lower and Higher Risk Portfolios (1979-1992) Portfolios of Large and Small Value Stocks (1979-1992)

20% 20% n n r r Efficient Version n n u u r r

t 19% Efficient t 19% u u e 18% Efficient Version e t t R R e e

Version

Small Value Small Value l l R R

18% a 18% a l l Stock Index u a u a Efficient

Small Firm Stock Index n Small Firm Stock Index n u u n n n n 16% Version n n A Large Firm A 17%

A A

d d Stock Index e e d d z z e e i i 16%

l Large Value l Large Value z z i i a

l 14% a l 14%

e Stock Index e Stock Index a a Inefficient Version e Inefficient Version e R R 15% Inefficient R Inefficient R Inefficient Version Version 12% 14% Inefficient Version 12% 14% 16% 18% 20% 22% 24% 26% 28% 10% 12% 14% 16% 18% 20% 22% 24% 26% 28% Risk Risk

Haugen’s Research Haugen’s Research

• Effect of Moving to Lower and Higher Risk Portfolios of • The Relative Performance of Low- and High-volatility Large and Small Growth Stocks (1979-1992) Stock Portfolios

18% Efficient Version – What has been the performance 17% – Over overlapping 5-year periods n n r r 16% Efficient Version u u • Of low- and high-volatility portfolios relative to the t t e e R R 15% Large Growth 15% S&P 500 (positioned at the origin of the graph)?

l l

a Stock Index a Stock Index u u 14% – Risk-return tradeoff idea of CAPM implies n n n n A A 13%

Small Growth Stock Index Small Growth Stock Index • Higher volatility portfolio would have higher return d d Inefficient e e 12% z z i i Version l l • Lower volatility, lower return a a e e 11% R R • Data should “tilt up” in scatter plot 10% Inefficient – But instead… 9% Version 8% 14% 16% 18% 20% 22% 24% 26% 28% 30% Risk

5 Haugen’s Research Haugen’s Research

• Test of 5 Year Horizon Performance of Low and High • Over-estimation of Short-run Volatility Portfolios using S&P 500 Stocks (1972-1992) – Relationship Between Perceived and True Growth

10% Horizon and Average Growth Rates

8% Low Volatility Portfolio • Growth horizon: length of time a typical stock takes 0 0 0 0 5 5 High Volatility Portfolio

6%

to mean-revert to the average rate of earnings P P Da & & t ta 4% growth. S S t

il

ts

o o d t t o 2%

w – Perceived horizon is longer than the true horizon

n:

e e low v

v w

i 0% i 0%er

t r t v • Reversion to the mean cuts in ahead of expectations

a o a -10% -8% -6% -4% -2% 0% o 2% 4% 6% 8% 10%

l l l lat e e -2% ility R R y,

, • “Growth” stocks

h hig n n -4% he r r er u u re t – Perform well in short term t tu e e -6% rn! R R -8% – Disappoint in medium term

-10%

Volatility Relative to S&P 500

Haugen’s Research Haugen’s Research

• Over-estimation of Short-run • Relationship Between Perceived and True Growth Horizon and Average Growth Rates Relative Growth • Investors expect high – Investors over-estimate Above • average rate of growth Average performers will remain “ahead of the pack” for • And length of the growth horizon much longer than they do Average • OverpriceYears growth into Future stocks in interim • Reduce price in medium Below Average term as reversion to

True Growth Perceived Growthmean kicks in Horizon Horizon

Haugen’s Research Haugen’s Research

• Overestimation of Short Run and Average Growth • Tendencies identified in IMH – Explain “fractal” nature of stock market data • Investors • Initial under-reaction to good news Above – expect current growth • Then over-reaction to good news Average will be higher and last • Followed by disappointment by longer than proves to • Volatile “up-down” feedback – Give non-institutional investor opportunity to profit Perceived be the case… Average • Analyse stocks to identify – Over-estimate average – Low volatility industry growth as well – High Book to Market True Average Years into Future – Out of favour sectors etc. Below Average • Develop portfolio of such stocks • Adjust on quarterly basis (minimise transaction True Growth Perceived Growth costs) Horizon Horizon

6 Haugen’s Research Haugen’s Research

• An instance: exploit under-pricing of high Book to Market • $1 invested in 1926 in portfolio is worth in 2009: stocks – Negative B/M: $16 – Data from French (of Fama & French—once proponent – Low 30%: $1,074 of EMH): – Medium 40%: $2,415 • http://mba.tuck.dartmouth.edu/pages/faculty/ken. – High 30%: $14,507 Portfolio Performance by Book to Market Valuation Ratio french/data_library.html 100000 Negative Book Value Low 30% – Rank companies by Book to Market Value 10000 Medium 40% High 30% • Negative (negative book value—like Internet 1000 n r u t

startups 1996-2000) e R

100 e v i t a

• Low 30% (“Growth stocks”—expensive but expected l

u 10 m to grow above trend by market u C 1 • Medium 40% 0.1 • High 30% (“Value stocks”—Buffett-style buy) 0.01 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010

Years

Haugen’s Research Other approaches

• Haugen no • Not a product endorsement, but… longer academic – Shows theory of previous lecture used in (successful?) • Sells portfolio trading strategies management system based on IMH…

• Many other companies exploiting similar inefficiencies • Eg, fractal trading systems…

7