The Global Financial Crisis and the Efficient Market Hypothesis:
What Have We Learned?
by Ray Ball, University of Chicago*
he sharp economic downturn and turmoil in
the financial markets, commonly referred to
as the “global financial crisis,” has spawned an
impressive outpouring of blame. Free market
economics—the idea that coordinated political forces do not
improve on the “atomistic” actions of individuals—has come
under concerted attack. The Efficient Market Hypothesis
(EMH)—the idea that competitive financial markets ruthlessly
exploit all available information when setting security prices—
has been singled out for particular attention.
As one prominent example, market strategist Jeremy
Grantham has called the EMH “responsible for the
current financial crisis” because of its role in the “chronic
underestimation of the dangers of asset bubbles” by financial
executives and regulators.1 And in the prologue and epilogue
to his meticulously researched, well-written, and best-selling
history of modern financial economics, The Myth of the
Rational Market, Justin Fox appears to say much the same
thing.2 The reasoning boils down to this: swayed by the
notion that market prices reflect all available information,
investors and regulators felt too little need to look into and
verify the true values of publicly traded securities, and so
failed to detect an asset price “bubble.” The Turner Report
by the UK’s market regulator (discussed more fully below)
reaches a similar conclusion. And in a fit of soul-searching,
the University of Chicago Magazine asks: “Is Chicago School
Thinking to Blame?”3 These are but a handful of the many
accusations that have been heaped on the EMH.
I have argued in the past and will argue below that
the EMH—like all good theories—has major limitations,
even though it continues to be the source of important
and enduring insights.4 Despite the theory’s undoubted
limitations, the claim that it is responsible for the current
worldwide crisis seems wildly exaggerated.
If the EMH is responsible for asset bubbles, one wonders
how bubbles could have happened before the words “efficient
market” were first set in print—and that was not until 1965,
in an article by Eugene Fama.5 Economic historians typically
point to the 1637 Dutch tulip “mania” as the first such
event on record, followed by episodes like the 1720 South
Sea Company Bubble, the Railway Mania of the 1840s, the
1926 Florida Land Bubble, and the events surrounding the
market collapse of 1929. But all of these episodes occurred
well before the advent of the EMH and modern financial
economic theory. As the above list suggests, unusually large
price run-ups followed by unusually large drops have occurred
throughout the recorded history of organized markets. It’s
only the idea of market efficiency that is relatively new to
the scene.
Further, the argument that a bubble occurred because the
financial industry was dominated by EMH-besotted “pricetakers”—that is, by people who viewed current prices as
correct and so failed to verify true asset values—seems wildly
at odds with what we see in practice. Almost all investment
money is actively managed, despite all the evidence of
academic and industry studies showing that active managers
fail to beat the market in an average year.6 Money flows
into mutual funds strongly follow past performance, as if
individual managers consistently beat the market over time,
and despite the evidence that the past performance of most
* Ball is a trustee of Harbor Funds and serves on the Shadow Financial Regulatory
Committee and FASB’s Financial Standards Advisory Council, but the views expressed
here are his own.
1. Cited (with apparent approval) in a widely read New York Times business column,
Joe Nocera, “Poking Holes in a Theory on Markets,” New York Times, June 5, 2009.
www.nytimes.com/2009/06/06/business/06nocera.html?scp=1&sq=efficient%20
market&st=cse. See also Grantham’s foreword in Andrew Smithers, Wall Street
Revalued: Imperfect Markets and Inept Central Bankers (Chichester, UK: Wiley,
2009).
2. Justin Fox, The Myth of the Rational Market: A History of Risk, Reward, and
Delusion on Wall Street (New York: HarperCollins, 2009), page 320. See also: George
Cooper, The Origin of Financial Crises: Central Banks, Credit Bubbles, and the Efficient
Market Fallacy (New York: Vintage Books, 2008); Richard A. Posner, A Failure of
Capitalism: The Crisis of ‘08 and the Descent into Depression (Boston: Harvard University
Press, 2009); George Soros, The Crash of 2008 and What it Means: The New Paradigm
for Financial Markets (New York: Perseus, 2009); and Andrew Smithers, op. cit.
3. Cover story, University of Chicago Magazine, Vol. 102 No. 1, September-October
2009.
4. Ray Ball, “The Theory of Stock Market Efficiency: Accomplishments and
Limitations,” Journal of Applied Corporate Finance 8, Spring 1995, pp. 4-17, and “On
the Development, Accomplishments and Limitations of the Theory of Stock Market
Efficiency,” Managerial Finance 20 (issue no.2/3), 1994, pp. 3-48.
5. Fama referred to “an ‘efficient’ market for securities, that is, a market where, given
the available information, actual prices at every point in time represent very good
estimates of intrinsic values.” Eugene F. Fama, “The Behavior of Stock-Market Prices”
The Journal of Business, Vol. 38, No. 1 (January 1965), p. 90. The idea did not become
known outside of narrow academic circles until the 1970s. It was not an easy sell to
practitioners at the time.
6. The first of the many studies reaching this finding is Michael C. Jensen, “The
Performance of Mutual Funds in the Period 1945–1964,” Journal of Finance 23 (May
1968), pp. 389–416. A recent Morningstar report concludes that only 37% of managed
funds outperformed their respective Morningstar style indexes over the past three years,
adjusting for risk, size and style. Similar numbers were observed for five and 10-year
returns.
See: http://news.morningstar.com/newsnet/viewnews.aspx?article=/dj/2009100713
14dowjonesdjonline000480_univ.xml.
B
T
8
Journal of Applied Corporate Finance • Volume 21 Number 4
A Morgan Stanley Publication • Fall 2009
money managers is a poor predictor of future performance.7
Much of the enormous losses by banks and investment
banks in 2007-2008 originated in their trading desks and
proprietary portfolios, whose strategies and very existence
were premised on making money from market mispricing.
Investors who poured money into the property market,
stock market, and other asset markets in the years while the
“bubbles” were forming seemed to do so in the belief that
prices would continue to rise, with the implication that they
believed current prices were incorrect. It seems inconsistent
to argue simultaneously that asset price “bubbles” occur and
that investors passively believe current asset prices are correct.
Yet this is precisely what many EMH critics have claimed.
But if more homeowners, speculators, investors, and banks
had indeed viewed current asset prices as correct, they might
not have bid them up to the same extent they did, and the
current crisis might have been averted.
The related argument that when asset prices are rising
rapidly their level is not subject to scrutiny by investors also
seems wildly at variance with the facts. Take the case of
then Fed Chairman Alan Greenspan’s 1996 use of the words
“irrational exuberance.” Despite its seemingly innocuous
nature and positioning in a long and otherwise unheralded
speech, the reference received widespread media coverage
both at the time, and more or less continuously during the
decade before the financial crisis.8 When my recent Google
search of “Alan Greenspan irrational exuberance speech”
yielded over six million hits,9 I had to ask myself: Can we
really believe that investors were not aware of the possibility
of a stock market bubble?
Perhaps it is not surprising that blame for the crisis has
been leveled at the EMH. Many investors and employees
have incurred considerable losses, regulators have lost face,
and scapegoats are needed. The EMH is a natural candidate. It
sounds academic. It is not welcomed by most money managers
because it states what they are not honest enough to admit to
their clients: that they operate in a fiercely competitive world,
populated by a large number of capable and ambitious people
just like themselves, and thus superior investment returns are
generally (though not exclusively) attributable more to luck
than insight. To justify their fees, active money managers have
to argue they are “above average” and consistently beat the
market, but the EMH—and the body of empirical studies
supporting it—suggests otherwise. The theory is also viewed
with skepticism by many (if not most) of the large number
of MBA students who launch forth into the world every
year, each believing—as the behavioral studies tell us—that
he or she is substantially above average, even though they
are their own future competition. The idea that it is hard to
earn excess returns in a competitive market also threatens the
lucrative market for an astonishing range of “get-rich-quick”
consultancies and treatises. In my experience, people whose
living derives from commenting authoritatively on the actions
of others—notably, academics, financial advisers, consultants,
journalists, and book authors—are more inclined than most
to view others as less rational than themselves.10 So the notion
of market efficiency is a natural target for blame.
Asset bubbles are not a well-understood phenomenon in
general. Many serious economists have challenged the use of
the term, other than in the ex post sense of denoting episodes in
which prices rose and then fell by substantial amounts. Trying
to pin such episodes on the EMH therefore does not strike
me as a very constructive exercise. To my mind there is less
drama, but more insight, to be gained by examining what the
crisis tells us about the efficient markets theory. Does the rapid
and substantial fall in prices that occurred across countries and
asset classes invalidate the notion of market “efficiency”? Or
does it merely serve to remind us of its considerable limitations
as a theory to help us understand the behavior of asset prices?
If so, then what are those limitations?
7. E. Sirri and P. Tufano, “Costly Search and Mutual Fund Flows,” Journal of Finance,
53 (1998), pp. 1589-1622.
8. The complete reference is: “But how do we know when irrational exuberance has
unduly escalated asset values, which then become subject to unexpected and prolonged
contractions as they have in Japan over the past decade? And how do we factor that
assessment into monetary policy?” The Challenge of Central Banking in a Democratic
Society: Remarks by Chairman Alan Greenspan at the Annual Dinner and Francis Boyer
Lecture of The American Enterprise Institute for Public Policy Research, Washington,
D.C., December 5, 1996 (Washington, DC: The Federal Reserve Board).
9. Visited October 18, 2009.
10. It also seems worth pointing out that such people, having chosen these
occupations and careers, have incentives, monetary and otherwise, to view themselves
as more rational than their audiences.
11. Eugene F. Fama, “Mandelbrot and the Stable Paretian Hypothesis,” Journal of
Business, Volume 36, Issue 4 (October 1963), pp. 420-429.
Journal of Applied Corporate Finance • Volume 21 Number 4
What Does the EMH Say?
The basic idea behind the EMH is deceptively simple. It
merges two insights. The first is one of the simplest and most
powerful insights of economics, the notion that competition
enforces a correspondence between revenues and costs. If
profits are excessive, new entry reduces or eliminates them.
The second insight, which is Gene Fama’s, is to view changes
in asset prices as a function of the flow of information to
the marketplace. Putting these two insights together leads to
the EMH, which I interpret as saying just this: competition
among market participants causes the return from using
information to be commensurate with its cost.
This fundamental idea leads directly to a startling—and
testable—prediction about financial markets’ reactions to
publicly released and widely-disseminated information,
such as corporate quarterly earnings reports. In competitive
equilibrium, the gains from exploiting public information
should correspond to the cost of exploiting it. But to a first
approximation, public information is costless to obtain, and
A Morgan Stanley Publication
•
Fall 2009
9
hence the gains from its use should be competed away to
zero. From this comes the prediction that one cannot expect
to earn above-normal returns from using publicly available
information because it already is reflected in prices.
Simple as it might seem in hindsight, this reasoning was
revolutionary at the time. While it was not by any means
a complete description of how security prices behave, and
its deficiencies became more apparent over time, the EMH
irreversibly changed the thinking of not only economists—
but of a great many practitioners—about how securities
markets behave.
What Doesn’t the EMH Say?
The EMH has been the subject of so much misunderstanding
that outlining some of the things the hypothesis does not say
occupies considerably more space than what it does say.
1. No one should act on information.
What would happen if all investors passively indexed
their portfolios? Obviously, the market would cease to
be efficient, because no investors would be acting to
incorporate information into prices. This has been a source
of misunderstanding from the outset. The misunderstanding
arises from confusing efficiency as a statement about the
equilibrium resulting from investors’ actions with the actions
themselves. Hair salons operate in a fiercely competitive
market, and the average salon is not expected to make
abnormal returns. That does not say all salons should stop
cutting hair. Investors act on information in a fiercely
competitive market, and the average investor is not expected
to make abnormal returns. That does not say all investors
should stop acting on information.
Yet this is the essence of the claim that market participants
were seduced into believing that since market prices already
reflected all available information, there was nothing to gain
from producing information and, as a consequence, security
prices were allowed to deviate substantially from their true
values. The critique confuses a statement about an equilibrium
“after the dust settles” and the actions required to obtain that
equilibrium.
and Benoît Mandelbrot on so-called “Paretian return”
distributions—that is, distributions of possible outcomes
that have “fat tails,” or more frequent extreme observations
than expected from the more-familiar bell-shaped “normal
curve.”11 Under the EMH, then, one can predict that large
market changes will occur, but one can’t predict when.
3. The stock market should have known we were in an
asset “bubble.”
It is easy to identify bubbles after the fact, but notoriously
difficult to profit from them. For example, let’s go back to Alan
Greenspan’s famous reference to “irrational exuberance.” The
speech was given on December 5, 1996, a day on which the
Dow Jones closed at 6437. If that statement is taken to mean
that prices were too high at the time, the clear implication
is that by today—when we all know how inefficient the
market is and how irrationally exuberant we were 13 years
ago, and after we have had ample opportunity to change our
behavior in response to that knowledge—there should have
been a substantial price correction. But at the time of this
writing, the Dow is near 10,000, a full 50% higher than when
Greenspan spoke.12 In other words, after 13 years to reflect
on Greenspan’s warning, investors are not acting as if there
was a bubble when he sounded the warning.
Asset price bubbles, or episodes in which prices rise and
then fall by substantial amounts, are much easier to spot
using hindsight than they are to predict. I like to ask a simple
question of people who believe that most stock market
investors ignored a pre-crisis bubble that burst in 2007-08.
My question is whether, prior to the crisis, they personally
had withdrawn from the stock and real estate markets and put
their wealth into cash instead. To my mind, this is the only
reliable test of whether they believed there was a bubble and
distrusted market prices at the time. In my limited experience
few withdrew much. By this test, I—a financial economist
skeptical about the possibility of identifying asset bubbles
except in hindsight—seem to have been more wary of a
bubble than the people who blame “the market” (but not
themselves) for creating it.
2. The market should have predicted the crisis.
4. The collapse of large financial institutions indicates the
market is inefficient.
The EMH does not imply that one can—or should be able
to—predict the future course of stock prices generally, and
crises in particular. Exactly the opposite: if anything, the
hypothesis predicts we should not be able to predict crises. If
we could predict a market crash, current market prices would
be inefficient because they would not reflect the information
embodied in the prediction.
Furthermore, the existence but unpredictability of large
market events is consistent with the work of Fama himself
George Soros, in his most recent book, has opined: “On a
deeper level, the demise of Lehman Brothers conclusively
falsifies the efficient market hypothesis.”13 I would have
thought the opposite. To me, Lehman’s demise conclusively
demonstrates that, in a competitive capital market, if you take
massive risky positions financed with extraordinary leverage,
you are bound to lose big one day—no matter how large and
venerable you are. Market efficiency does not predict there will
be no spectacular failures of large banks or investment banks.
12. The increase exceeds the 35% consumer price inflation over the period.
10
Journal of Applied Corporate Finance • Volume 21 Number 4
13. George Soros, op. cit, p. 165.
A Morgan Stanley Publication • Fall 2009
If anything, it predicts the opposite—that size and venerability
alone will not guarantee you positive abnormal returns, and
will not protect you from the forces of competition.
5. The EMH assumes that return distributions do not change
over time.
The EMH is completely silent about the shapes of the
distributions of securities’ returns. Indeed, I will argue below
that this is one of the principal limitations of the theory.
This has been a glaring hole in “modern” financial economic
theory in general, going all the way back to Irving Fisher’s
work on the discounted present value model a century ago.
The EMH does not imply that past return distributions—
including statistics such as means, variances, skewness, and
correlation matrices—will mechanically repeat themselves in
the future. What the EMH does say about return distributions
is that, given a certain amount and kind of publicly available
information, security prices are “efficient” in the statistical
sense that they are “minimum-variance” forecasts of future
prices. In other words, to the extent that a price has already
adjusted to the available information, no future price reaction
to that information is necessary, and the investor is not exposed
to future price variability arising from that source. By contrast,
a market that adjusts only partially to information when it
arrives leaves the investor exposed to further reaction at a
later date, thereby resulting in excessive variability, and hence
“inefficient” prices.
In sum, the EMH says nothing about the stationarity over
time of return distributions. There is no deus ex machina in
securities markets that ensures the stability of such variables,
no economic forces that mechanically draw security returns
like lottery numbers every day from the same barrel. Quite
the contrary: there is considerable evidence that risk in
particular is “non-stationary” to an important degree. So
if financial economists—or math and physics majors with
little appreciation of long-term economic history posing as
financial economists—calculate future risks entirely from
recent historical data, they do so as an act of belief rather
than theory, and they ignore evidence contrary to that belief.
One cannot blame the EMH for such practices.
Yet the EMH is cited as playing a major role in the crisis
in the Turner Review, a post mortem report issued by the
U.K.’s market regulator at the request of the Chancellor of the
Exchequer. Consider this summary of the report’s conclusions
on market efficiency:14
At the core of these assumptions has been the theory of
efficient and rational markets. Five propositions with
implications for regulatory approach have followed:
14. The Turner Review: A Regulatory Response to the Global Banking Crisis (London,
UK: The Financial Services Authority, March 2009, page 39).
15. Consistent with Peltzman’s theory that increased regulation generally is a political
attempt to escape blame for crises. See Sam Peltzman, “Toward a More General Theory
Journal of Applied Corporate Finance • Volume 21 Number 4
(i) Market prices are good indicators of rationally
evaluated economic value.
(ii) The development of securitised credit, since based
on the creation of new and more liquid markets,
has improved both allocative efficiency and
financial stability.
(iii) The risk characteristics of financial markets
can be inferred from mathematical analysis,
delivering robust quantitative measures of
trading risk.
(iv) Market discipline can be used as an effective tool
in constraining harmful risk taking.
(v) Financial innovation can be assumed to be
beneficial since market competition would
winnow out any innovations which did not
deliver value-added.
Each of these assumptions is now subject to extensive
challenge on both theoretical and empirical grounds,
with potential implications for the appropriate design of
regulation and for the role of regulatory authorities.
Only the first of these five propositions bears any
resemblance to the simple notion of efficient price responses
to information. The third proposition—that market efficiency
implies there are “robust quantitative measures of trading
risk”—involves a considerable exaggeration of the theory’s
prescriptive import.
6. Financial regulators mistakenly relied on the EMH.
The crisis has prompted many to conclude that financial
regulators were excessively lax in their market supervision, due
to a mistaken belief in the EMH. This conclusion is made
explicit in the UK’s Turner Review. Perhaps not surprisingly, the
report advocates more regulation.15 It reasons as follows:16
e predominant assumption behind financial market
Th
regulation—in the US, the UK and increasingly across
the world—has been that financial markets are capable
of being both efficient and rational and that a key goal of
financial market regulation is to remove the impediments
which might produce inefficient and illiquid markets….
In the face of the worst financial crisis for a century,
however, the assumptions of efficient market theory have
been subject to increasingly effective criticism.
This characterization of what the EMH implies for
regulators makes sense in one respect. If the market does
a good job of incorporating public information in prices,
of Regulation,” Journal of Law and Economics 19 (1976), pp. 211-240.
16. Ibid, pages 39-40.
A Morgan Stanley Publication
•
Fall 2009
11
regulators can focus more on ensuring an adequate flow
of reliable information to the public, and less on holding
investors’ hands. Consistent with this view, in recent decades
there does appear to have been increased emphasis by
regulatory bodies worldwide on ensuring adequate and fair
public disclosure.
Otherwise, the characterization of the role of the EMH
in the crisis falls short of the mark. If regulators had been
true believers in efficiency, they would have been considerably
more skeptical about some of the consistently high returns
being reported by various financial institutions. If the capital
market is fiercely competitive, there is a good chance that
high returns are attributable to high leverage, high risk,
inside information, or dishonest accounting. True believers in
efficiency would have looked more closely at the leverage and
risk-taking positions of Lehman Brothers, Bear Sterns, AIG,
Freddie Mac and Fannie Mae, and banks and investment
banks generally. They might have questioned the source of the
trading profits of hedge funds like Galleon, and discovered
some using inside information. And they would have been
exceptionally skeptical of the surreally high and stable returns
reported over an extended period by Bernie Madoff.17
to come up with a more specific and implementable pricing
model. It therefore is less robust than the basic idea of correct
pricing. People who take models literally are in for a very big
disappointment.
No theory can explain everything. This is a central
point in Kuhn: anomalies abound in all theories, but we are
prepared to live with them if we find the theory to be more
useful than the best alternative. In other words, it takes a
theory to beat a theory—a theme I come back to later when
discussing the contributions of “behavioral finance.”
An analogy might be helpful here. One can view the
proposition “man is moral” as a useful way of thinking about
the world, without taking it to mean that no person ever
has acted or will act immorally, or without implying any
of the following: (1) that one knows exactly what “moral”
means; (2) that there are no logical inconsistencies in one’s
views about what constitutes “moral” behavior under different
circumstances; or (3) that one cannot design an experiment in
which people act inconsistently with a particular definition of
“morality.” The same is true of market “efficiency.”
Some Lessons from the Financial Crisis
So, what have we learned about market efficiency from the
financial crisis? The short answer is: some things we should
have known beforehand.
At a theoretical level, the EMH has many obvious limitations.
The most important of these limitations stems from the fact that
EMH is a “pure exchange” model of information in markets.
What this means is that the theory makes no statements
whatsoever about the “supply side” of the information
market: about how much information is available, whether
it comes from accounting reports or statements by managers
or government statistical releases, what its reliability is, how
continuous it is, the frequency of extreme events, and so forth.
The theory addresses only the demand side of the market. The
EMH says only that, given the supply of information, investors
will trade on it until in equilibrium there are no further gains
from trading. Consequently, the EMH is silent about the
shapes of return distributions and how they evolve over time.
An almost exclusive focus on the demand side is perhaps
the single biggest weakness of “modern” financial economics
generally. The discounted present value, or NPV, model for
valuation and capital budgeting states that, given an expected
stream of future cash flows, those cash flows are priced so as
to provide investors a given return. The Miller-Modigliani
theorems state that, given corporate investment decisions and
the earnings from that investment, pure exchange among
1. A Theory is Just a Theory.
First and foremost, the episode highlights that a theory is
just that—a theory. It is not a fact. It is an abstraction from
reality. It is an abstraction that we hopefully find useful
when organizing our thoughts and actions, but no theory
is perfect. As Thomas Kuhn, the well-known historian of
science, reminds us, all theories have “anomalies”—facts or
findings that the theories cannot explain.18 No theory can or
should totally determine our thoughts or our actions. People
who take theories literally are in for a disappointment.
Further, specific models of a theory are even greater
abstractions. They are ways of implementing the basic ideas in
a theory, using more detailed and more specific assumptions
that adapt the theory for particular purposes. They cannot
and should not be taken literally. For example, the Capital
Asset Pricing Model takes the basic concept of correct pricing
and adds a number of assumptions about return distributions
17. SEC Inspector General David Kotz concludes his report on the Commission’s
handling of the case as follows: “Despite numerous credible and detailed complaints, the
SEC never properly examined or investigated Madoff’s trading and never took the
necessary, but basic, steps to determine if Madoff was operating a Ponzi scheme.” Three
of the six complaints came from Harry Markopolos, who raised 30 “red flags.” These
included his calculation that there simply were not enough S&P-100 options in existence
to generate the dollar returns Madoff was reporting from trading in them, and his analysis
that Madoff’s claimed return stream was consistent with no known investment strategy.
In Congressional testimony, Markopolos attributed the SEC’s lack of adequate action to
it having “too many attorneys and too few professionals with any sort of financial
12
Journal of Applied Corporate Finance • Volume 21 Number 4
2. There are Limitations to the EMH as a Theory of
Financial Markets.
background” to understand his calculations (Bloomberg, February 4, 2009). This
episode does not smack of regulators following any type of financial theory, let alone
possessing the skepticism of Madoff’s returns that is implied by a genuine belief in the
EMH. See United States Securities and Exchange Commission, Office of Inspector
General: Report Of Investigation, Case No. OIG-509, Investigation of Failure of the SEC
to Uncover Bernard Madoff’s Ponzi Scheme, August 31, 2009.
18. Thomas S. Kuhn, The Structure of Scientific Revolutions. 2nd ed. Chicago:
University of Chicago Press, 1970. The Times Literary Supplement listed this work
among “The Hundred Most Influential Books since the Second World War.”
A Morgan Stanley Publication • Fall 2009
investors makes the value of the firm independent of and
unaffected by differences in capital structure and financing
policies generally. The CAPM states that, given the variancecovariance matrix of future returns and the pricing of two
benchmark efficient portfolios, pure exchange among investors
determines the risk-return relation. The Black-Scholes
option pricing model states that, given the share price, price
volatility, and several other variables, pure exchange among
investors determines the price of an option on the share. These
theoretical milestones all have been achieved at the expense of
ignoring the real sector—that is, where the cash flows come
from for discounting, what projects companies invest in, what
determines security risk, and so on.19
As a consequence, when households suddenly decide to
stop adding to the real housing stock, modern finance theory
is silent about the implications. For example, the CAPM
takes the riskless rate, market risk premium, and individualsecurity betas as given. But in the event of a large shock in
a real asset market, what values of these CAPM parameters
would be consistent with efficient pricing of securities? Most
empirical tests of market efficiency typically avoid this issue,
and implicitly assume that the observed values for riskless
rates, market risk premiums, and betas are correct.
Real factors obviously matter but, by focusing almost
exclusively on monetary exchange, modern financial theory
has made its major breakthroughs by ignoring them. An
equivalent problem is faced by those who assume the crisis
originated in the financial sector and then spread to the real
sector, reducing economic output and raising unemployment.
Indeed, the popular term financial crisis takes this assumption
as a given. My own view is that the problems originated in
the real asset markets (chiefly in real estate), but was first
reflected in the financial markets—precisely because those
markets are more efficient. The general public might have
first learned of the collapse in real asset prices from the credit
market liquidity problems and widening spreads that emerged
in the summer of 2007, or from the collapse of Bear Sterns or
Lehman Brothers, or from the fall in stock prices generally.
But that does not mean that the problems originated in, or
were “caused by,” the financial markets. They were just the
proverbial canary in the coal mine.
In addition to these limitations of EMH that stem from
ignoring the supply side of the information, there are a
number of others worth noting:
• Information is modeled in the EMH as an objective
commodity that has the same meaning for all investors. In
reality, investors have different information and beliefs. The
actions of individual investors are based not only on their
own beliefs, but beliefs about the beliefs of others—that is,
their necessarily incomplete beliefs about others’ motives for
trading. This likely becomes most important during periods
of rapid price changes, such as October 1987. Unlike more
stable periods, when an investor can wake up and read or
listen to some thoughtful analysis of the prime movers of
prices on the previous day, this kind of information is not
available in a timely fashion during periods of rapid price
change
• Information processing is assumed in the EMH to be
costless, and hence information is incorporated into prices
immediately and exactly. While it seems reasonable to assume
that the cost to investors of acquiring public information is
negligible, information processing (or interpretation) costs are
an entirely different matter. They have received surprisingly
little attention.
• The EMH assumes the markets themselves are
costless to operate. Generally speaking, stock markets are
paradigm examples of low-cost, high-volume markets, but
they are not entirely without costs. This limitation raises the
following conundrum: if there are pricing errors that are not
eliminated because they are smaller than the transaction costs
of exploiting them, is the market judged to be efficient—
because of the absence of profits from exploitable errors—or
inefficient—because of price errors that persist because of
transactions costs? The role of transaction costs in the theory
of market efficiency is unclear.
• Similarly, the EMH implicitly assumes continuous
trading, and hence ignores liquidity effects. There is evidence
that illiquidity is a “priced” factor—that is, higher returns
compensate for lower liquidity—though how to measure
liquidity is unclear.20 Few would take the fact that markets
are closed on weekends or overnight as a serious violation of
market efficiency, but episodes of heightened illiquidity are
another matter. Starting in the summer of 2007, illiquidity
was an extremely important feature of many credit markets
and real asset markets.
• The EMH also is silent on the issue of investor taxes.
In reality, many investors pay taxes on dividends and capital
gains, with some offsets for capital losses. The effects of
investor taxation on security prices and expected returns are
potentially large, but not well understood.
From the above, it should be apparent that the EMH
adopts a simplified view of markets. To those who take
theories literally—not as useful abstractions—the combined
effect of these simplifications could well be to encourage a
deus ex machina view of securities markets, as discussed above.
But that is a problem of the musician, not the instrument.
19. The rare breakouts from pure-exchange thinking have provided important insights,
notably the Jensen and Meckling proof that the Miller-Modigliani theorems do not hold
with positive agency costs, because firms’ investment decisions then depend on their
financing policies. See Jensen, M.C., and W.H. Meckling, 1976, “Theory of the Firm:
Managerial Behavior, Agency Costs and Ownership Structure,” Journal of Financial
Economics 3, pp. 305–360.
20. Y. Amihud and H. Mendelson, “Asset Pricing and the Bid-ask Spread,” Journal of
Financial Economics 17 (1986), pp. 223–249. R.A. Korajczyk and R. Sadka, “Pricing
the Commonality across Alternative Measures of Liquidity,” Journal of Financial
Economics 87 (2008), pp. 45–72.
Journal of Applied Corporate Finance • Volume 21 Number 4
A Morgan Stanley Publication
•
Fall 2009
13
3. There are limitations to tests of the EMH.
Many problems also arise in testing market “efficiency,”
including the following:
• Just as a test of the proposition “man is moral” requires
an operational definition of what constitutes “moral,” a test of
efficiency requires a precise specification of what constitutes
an “efficient” price response to information. Normally this is
done by comparing the returns earned from trading on the
information with the returns otherwise expected from passive
investing. But implementing the “counterfactual” in this way
suffers from what Fama describes as the “bad model” problem:
we do not have a perfect theory of the returns to be expected
from passive investment. Early empirical work relied on the
Capital Asset Pricing Model to estimate expected returns,
but it does a poor job of predicting returns on average: high
beta-risk stocks cannot be shown to earn higher average
returns than low beta-risk stocks. This finding could be
caused by the CAPM being a bad model, or by betas being
difficult to measure accurately; but either explanation causes
problems in testing market efficiency. Later work employs
the Fama-French three-factor model, which does a better job
of predicting returns but is cobbled together based more on
a foundation of empirical correlations than on solid asset
pricing theory. Because tests of market efficiency are “joint
tests” of the market’s ability to incorporate new information
in prices and a particular model of asset pricing, any flaws in
the model affect the reliability of the test of efficiency.
• Tests of the EMH involve studying the flow of
information into market prices. Many types of information
could be expected to change—or at least not be independent of
changes in—important asset pricing parameters such as interest
rates, risk, risk premiums, and securities’ risks. Consider the
information contained in variables like Federal Reserve policy,
tax rates, investor demographics, technological change, and
labor productivity. We know little about how such variables
evolve over time, or about the implications of their evolution
for the time series behavior of expected returns in an efficiently
priced market. Some of these variables will be subject to
long-term secular change. For example, a secular aging of the
population increases the ratio of retirees living off investment
income to workers, and could bring about a secular decline in
real interest rates. What sequence of aggregate (market-wide, or
index) security returns would then be consistent with efficient
pricing? Other variables will experience transitory shocks. As
was made painfully clear during 2008, aggregate system-wide
risk accelerates rapidly when correlations across asset returns
rise. Investors are unable to diversify system-wide risk, so
market indexes fall sharply to yield a substantially increased
risk premium in the form of increased expected returns (that
is, expected price reversals). What exact sequence constitutes
an efficient price reaction? Was the size of the fall, and thus the
size of the expected recovery, too large? Too small? The EMH
is silent on these issues.
14
Journal of Applied Corporate Finance • Volume 21 Number 4
• At the individual security level, important parameters
like risk are difficult to model and estimate. Risk is clearly
not a constant. Companies, like market indexes, can be
expected to experience occasional episodes of heightened
uncertainty—for example, during periods of strikes, antitrust
or other legal action, or major moves by competitors. Equity
betas can be expected to change in response to changes in
companies’ stock prices, which cause changes in marketvalued financial leverage. They can also be expected to vary
with major announcements, including earnings. What level
of risk makes security efficiently priced? Does the market
over- or under-assess risk? Many research designs estimate
and control for the realized, or ex post level of risk, with the
implicit assumption that the observed level is the correct level.
A similar observation can be made about securities’ loadings
on the three factors in the Fama-French model.
4. The real world is complex.
One of the important lessons from the global financial crisis
is that the world is more complex than many thought, and
certainly is more complex than many or most pricing models
used in practice. Many derivative securities that PhDs in
financial economics or physics were employed to model in
one clean, crisp equation were discovered to be—in practice if
not in theory—the creations of complicated legal documents.
For many mortgage backed securities, it is difficult to sort out
what the underlying cash flow rights of the investor actually
are when defaults start occurring.
Did the simplicity of the models employed by researchers,
from 1970 Fama’s formulation of EMH through to specific
pricing models, lull people into thinking that the EMH meant
the same things as the models? Maybe. But one can’t blame a
theory for people misusing it. Every theory is an abstraction;
no theory can be taken literally.
Anomalies, Behavioral Finance, and the Future of
“Market Efficiency”
By now, it should be clear that anomalies in the theory
of market efficiency abound. The long list includes price
overreactions and excess volatility; price underreactions
and momentum, particularly in relation to earnings
announcements; seasonal patterns in returns; and the relation
between future returns and many variables such as market
capitalization, market-to-book ratios, price-earnings ratios,
accounting accruals, and dividend yields.
No theory can explain all the data it is asked to explain:
there are always anomalies. Only in the next world are we
promised perfect comprehension. What is never totally clear
is whether the market anomalies are due to imperfections in
the markets themselves, imperfections in market efficiency
as a way of thinking about how competitive markets behave,
or defects in the research itself. One suspects the answer is
(d): all the above.
A Morgan Stanley Publication • Fall 2009
Where does behavioral research come into this picture?
In one sense, the “behavioralists” in finance merely jumped
on the bandwagon that started when the early financial
economists started observing and reporting anomalies. The
first discussion of an anomaly in the market reaction to public
information that I’m aware of is in my 1968 study with Philip
Brown of the market reaction to earnings announcements.
We observed that the market response to the announcements
persisted for several months, a phenomenon that later
became known as “post earnings announcement drift” or
“earnings momentum.”21 By the mid-1970s this pattern had
been observed in several studies, and I used Kuhn’s word
“anomaly” to describe it.22 Basu’s discovery of abnormal
returns by companies with low P/E ratios was published in
1977, and Banz’s finding of the same for small firms was
published in 1981.23 But the genesis of the behavioral finance
literature is generally identified as the publication of two
famous papers by Werner DeBondt and Richard Thaler, one
in 1985 and the second in 1987.24 Since then, behavioral
research has succeeded in poking many more holes in the
theory of efficient markets.
Has behavioral finance supplanted EMH as the prevailing
theory of financial markets? The question assumes that it is
a theory, as distinct from a collection of ideas and results.
As I see it, the behavioral literature relies on the theory of
efficient markets. By that I mean the following. A revealing
fact is that the behavioral finance literature contains no
references that I can find to anomalies in behavioral finance. I
reviewed six compendiums of behavioral finance and searched
all issues of the Journal of Behavioral Finance for references
to “anomalies.”25 The only references I could find were to
anomalies in the theory of efficient markets. Does the absence
of reference to its own anomalies mean that behavioral theory
is perfect?
Kuhn tells us that to discover anomalies one first must
have a theory that is capable of being contradicted. One of
the strengths of the EMH is its refutability: it can be tested.
One gets the impression that behavioral finance, taken as a
whole, consists of a set of disjointed and inconsistent ideas,
some of which are rationalizations of the anomalies of others.
If all theories are abstractions and all theories have anomalies,
but behavioral finance has no anomalies, the implication is
that it is not a theory.
However, behavioral finance does have its own anomalous
evidence, even if it does not receive prominent treatment
in its own literature. For example, it is a widely held belief
among fund managers that it is easier to earn fees from
selling and managing funds that trade on anomalies and
behavioral strategies than it is to earn abnormal returns on
the funds. Moreover, this belief was tested in a recent study
of 16 mutual funds whose stated investment strategy is to
trade on behavioral financial research ideas. In that study,
Wright, Banerjee and Boney conclude that, while there is
considerable variation across the funds, taken as a group
they attracted more investment dollars than comparable
non-behavioral managed funds, but without earning higher
risk-adjusted returns (which were roughly the same between
the two groups).26
None of the above is meant to detract from the
contribution of behavioral financial research, which has
widened our knowledge of how financial markets behave and
has demonstrated major holes in the efficient markets theory.
Nor does it imply that behavioral finance has replaced (or
will replace) market efficiency as the fundamental construct
underlying how we think about financial markets. Kuhn
reminds us that anomalies abound in all theories, but we
live with them if we find the theory useful: it takes a theory
to beat a theory. Despite its limitations, the notion that
prices efficiently incorporate information is an indispensable
foundation for how we organize the world. Three examples
can be used to illustrate this point.
The first example is the concept and method of discounted
present value. There is no evidence that anomalous evidence
for the efficient market theory has led to a wholesale
abandonment of present value. It continues to be widely
used in law, economics, business, finance and accounting. It
underlies how we think about the value of income streams,
and is used for valuation calculations in a variety of contexts.
21. Ray Ball and Philip Brown, “An Empirical Evaluation of Accounting Income
Numbers,” Journal of Accounting Research Vol. 6, 1968, pp.159‑178, discussion at pp.
173-174.
22. Ball, Ray, 1978, “Anomalies in relationships between securities’ yields and yieldsurrogates,” Journal of Financial Economics Vol. 6, pp. 103-126.
23. S. Basu, “Investment Performance of Common Stocks in Relation to Their PriceEarnings Ratios: A Test of the Efficient Market Hypothesis,” Journal of Finance, Vol. 32,
No. 3 (June, 1977), pp. 663-682. Rolf W. Banz , “The Relationship between Return and
Market Value of Common Stocks,” Journal of Financial Economics Vol. 9 (1981), pp.
3–18.
24. DeBondt, Werner, and Richard Thaler, 1985, “Does the Stock Market Overreact?”
Journal of Finance 40, pp. 793-805. DeBondt, Werner, and Richard Thaler, 1987,
“Further Evidence on Investor Overreaction and Stock Market Seasonality,” Journal of
Finance 42, pp. 557-581.
25. Richard H. Thaler, editor, Advances in Behavioral Finance (New York: Russell
Sage Foundation, 1993). Hersh Shefrin, editor. Behavioral Finance (Cheltenham, UK
and Northampton, MA: Edward Elgar, 2001). Hersh Shefrin, Beyond Greed and Fear:
Understanding Behavioral Finance and the Psychology of Investing. Andrei Shleifer,
Inefficient Markets: An Introduction to Behavioral Finance. (New York: Oxford University
Press, 2000). Edward. J. McCaffery and Joel Slemrod, editors. Behavioral Public
Finance (New York: Russell Sage Foundation), 2006. Mark Schindler, Rumors in
Financial Markets: Insights into Behavioral Finance (Chichester, England and Hoboken,
NJ: John Wiley & Sons Inc., 2007).
26. See Colby Wright, Prithviraj Banerjee and Vaneesha Boney, “Behavioral Finance:
Are the Disciples Profiting from the Doctrine?” The Journal of Investing, Vol. 17, No. 4,
Winter 2008, pp. 82-90.
In another study, Chan, Frankel and Kothari conducted direct tests of behavioral
hypotheses about investor biases in revising their priors in response to new financial
information. Using historical stock return data, they simulated investment portfolios that
trade on the different predictions implied by the two behavioral hypotheses. The study
found no evidence to support Tversky and Kahneman’s “representativeness” bias, and
only weak evidence for Edwards’ “conservatism” bias. See Chan, W., R. Frankel and S.P.
Kothari, 2004, “Testing Behavioral Finance Theories Using Trends and Sequences in
Financial Performance.” Journal of Accounting & Economics 38, 3-50; A. Tversky and D.
Kahneman, “Judgment under Uncertainty: Heuristics and Biases,” Science 185 (1974),
pp. 1124–1131; and W. Edwards, “Conservatism in Human Information Processing,” in
B. Kleinmutz, editors, Formal Representation of Human Judgment, Wiley, New York
(1968).
Journal of Applied Corporate Finance • Volume 21 Number 4
A Morgan Stanley Publication
•
Fall 2009
15
Yet the present value rule is a very strict application of the Law
of One Price, a fundamental economic principle that states:
“In an efficient market all identical goods sell for an identical
price.” In the words of Lamont and Thaler, this principle
“is the basis for much of financial economic theory.”27 The
present value rule assumes that the discount rate—that
is, the price per dollar of a cash flow at a future date—is
independent of the amount of the cash flow, its sign (inflow
or outflow), and the way it is packaged with cash flows at
other dates—for example, as part of a multi-period financial
instrument or a long-term investment project.28 That is, it
states that there is one price for future money, no matter how
it arises. The rule assumes efficient pricing, and it has not been
abandoned presumably because it is a useful—though clearly
not a perfect—guide for our thinking and calculations when
valuing assets, liabilities, and entire companies.
The second example is the legal theory of “fraud on the
market” that underlies the majority of U.S. securities class
action fraud cases. This legal theory states that investors
trading in an efficient market are implicitly relying on stock
prices that are assumed to incorporate all public information.
Because of this assumption, individual plaintiffs do not have
to prove that they directly received and relied on the false
information they allege was fraudulently provided; instead
they are assumed to have indirectly relied on such fraudulent
information when they traded at a particular market price
that incorporated it.29 In some respects, behavioral financial
theory is inconsistent with this legal theory, but it is difficult
to see how it could replace it.
The third example is the use of market prices for valuation
purposes in a wide variety of contexts. Consider the daily
calculation of net asset values (NAVs) of shares in mutual
funds. Fund NAVs are based on the prices of the securities
in which the fund is invested. The “fair value” of each
security is determined daily, and aggregated across securities
to determine the fund’s NAV—and as the price paid and
received by investors buying and selling fund shares, the NAV
is an important calculation. If it is traded on a liquid market,
the closing price of a security (or the price from the most
recent trade) is normally taken to be its fair value.
But now consider what often happens when securities are
traded in low-volume emerging markets, in which the last
trade price may well be “stale.” If the market index has fallen
since the last trade, the price at which it last traded is often
adjusted downward to an estimated fair value, with the aim
of treating investors buying and selling fund shares fairly. This
kind of adjustment of the last trade price can be viewed as
consistent with EMH in the sense that, as discussed earlier,
the theory is silent on questions of liquidity.
What we do not observe in practice is more illuminating.
We do not see prices being adjusted downward by large
amounts during asset bubbles because they are wildly in
excess of fair values, or upward during troughs. This is not the
practice of banks, investment banks, insurance companies,
private equity funds, mutual funds, or in any context in
which market pricing is important. There does not seem to be
a market for an adjusted index fund that makes anti-cyclical
price adjustments, subtracting from prices at the peak and
adding at the trough. There does not seem to be any legal
support for the argument that because prices prove to be too
high at the height of bubbles, and so unfairly advantage those
who sell mutual fund shares over those who buy, the prices
themselves should be adjusted substantially downward. My
conjecture is that investors and courts would not trust the
fund management to know ex ante when they are at a peak
or at a trough.
In other words, as a practical matter bubbles may only
exist in hindsight. Contemplating the fanciful nature of the
counterfactual helps us to understand why the practice of
relying on actual security market prices is so entrenched
in commercial practice, in law, and in regulation. They are
“efficient” enough, despite anomalous evidence against the
EMH. When push comes to shove, what is the practical
alternative?
27. Lamont, O.A. and Thaler, R.H. (2003), “Anomalies: The Law of One Price in
Financial Markets,” Journal of Economic Perspectives 17 (Fall 2003), pp. 191-202, at
page 191.
28. For example, see Thomas S. Y. Ho and Sang Bin Lee, The Oxford Guide to
Financial Modeling Applications for Capital Markets, Corporate Finance, Risk
Management and Financial Institutions (Oxford: 2004), page 54.
29. See Basic Inc. v. Levinson, 485 U.S. 224 (1988).
16
Journal of Applied Corporate Finance • Volume 21 Number 4
Closing Thoughts
Fama’s 1965 insight—combining simple competitive
economic theory with an information-based view of security
prices—irreversibly changed the way we look at financial
markets. Like all important insights, this is the case even if
it is not a complete representation of how markets behave.
The impact of the theory of efficient markets has proven to
be durable, and seems likely to continue to be so, despite its
inevitable and painfully obvious limitations.
ray ball is Sidney Davidson Professor of Accounting at the University
of Chicago’s Booth School of Business. He also is a trustee of Harbor
Funds and serves on the Shadow Financial Regulatory Committee and
the FASB’s Financial Standards Advisory Council.
A Morgan Stanley Publication • Fall 2009
Purchase answer to see full
attachment