Journal of Economic Perspectives—Volume 27, Number 2—Spring 2013—Pages 29–50

Finance: Function Matters, Not Size

John H. Cochrane

T
he US economy spends $170 billion a year on advertising, just to trick
people into buying stuff they don’t need. What a waste!
There are 2.2 people doing medical billing for every doctor that actu-
ally sees patients, costing $360 billion—2.4 percent of GDP. Talk about an industry
that is too big!
Wholesale and retail trade and transportation cost 14.6 percent of GDP, while
all manufacturing is only 11.5 percent of GDP. We spend more to move goods
around than to make them!
My wife asked me to look at light fixtures. Do you know how many thousands of
different kinds of light fixtures there are? The excess complexity is insane. Ten ought
to be plenty.
It’s ridiculous how much people overpay for brand names when the generic is
so much cheaper. People are pretty naive.
Business school finance professors are horribly overpaid. Ask an anthropolo-
gist! We get paid almost a half a million bucks, and work a grand total of 10 weeks a
year, all to teach students that they can’t make money trading in the stock market.
It’s fun to pass judgment on waste, size, usefulness, complexity, naiveté, and
excessive compensation, isn’t it? But as economists, we have an analytical structure
for thinking about these questions. We start with supply, demand, and competition,

■ John H. Cochrane is the AQR Capital Management Distinguished Service Professor of

Finance, University of Chicago Booth School of Business, Chicago, Illinois. He is also a
Research Associate, National Bureau of Economic Research, Cambridge, Massachusetts;
Senior Fellow, Hoover Institution, Stanford, California; and Adjunct Scholar, Cato Institute,
Washington, DC. His email address is [email protected], and his website is
http://faculty.chicagobooth.edu/john.cochrane.
http://dx.doi.org/10.1257/jep.27.2.29. doi=10.1257/jep.27.2.29
30 Journal of Economic Perspectives

and with the suggestion of the first welfare theorem that these forces usually lead to
socially beneficial arrangements. When outcomes seem puzzling using this analysis,
we embark on a three-pronged investigation. First, we work harder to find how
supply and demand might really operate, in the humble knowledge that initially
puzzling institutions and outcomes have often taken us years to comprehend.
Second, maybe there is a “market failure”— an externality, public good, natural
monopoly, asymmetric information situation, or missing market—that explains our
puzzle. Third, we often discover a “government failure,” that the puzzling aspect of
our world is a consequence of laws or regulation, either unintended or the result
of capture.
Only then can we begin to diagnose a divergence between reality and socially
desirable outcomes, and only then can we start to think of how to improve reality.
“I don’t understand it” doesn’t mean “it’s bad,” or “regulation will improve it.” And
since that attitude pervades policy analysis in general and financial regulation in
particular, economists do the world a disservice if we echo it.
I belabor this point, because I do not offer a competing black box. I don’t
claim to estimate the socially optimal “size of finance” at, say, 8.267 percent of
GDP. It’s just the wrong question. Hayek and the failure of planning should teach
us a little modesty: Pronouncing on socially optimal industry size is a waste of time.
Is the finance industry functioning well? Are there identifiable market or govern-
ment distortions? Will proposed regulations help or make matters worse? These
are useful questions.
With a rather catastrophic failure behind us and other crises bubbling on the
back burner, it also seems a bit strange to be arguing whether 5 or 8 percent of GDP
is the right “size” of finance, and whether it needs to be nudged to become larger
or smaller. Many of us might happily accept an additional 3 percentage points of
GDP in the financial sector in return for a financial system that is not prone to runs
and crises. Our political system has accepted a big increase in resources devoted to
financial regulation and compliance, and a potentially larger reduction in the effi-
ciency, innovation, and competitiveness of financial institutions and markets, in the
quest—misguided or not—for stability. The run-prone nature of the US financial
system, together with its massive regulation, subsidies, government guarantees, and
regulatory capture, looks to be a more fertile fishing ground for trying to under-
stand market and government failures than does mere size.
Still, the size of finance represents a contentious issue, and my plea that we ask
different questions isn’t going to silence the debate, so let us think about it. Let us
use size as an organizing principle for studying function and dysfunction.
Greenwood and Scharfstein nicely review the key facts and ideas in their paper
in this issue. Their most basic story is: quantity increased a lot, but prices didn’t
fall. This description suggests a simple economic interpretation: The demand
for financial services shifted out. People with scarce skills supplying such services
made a lot of money. A system with proportional fees, which is a common struc-
ture in professional services, interacted with stock-price and home-price increases
(a different surge in demand) to produce increased financial sector revenue. Why
 John H. Cochrane 31

demand shifted out, and why house and stock prices rose (temporarily, it turns
out) are good questions—but they don’t have much to do with the structure of the
finance industry. This story also suggests that, like the weather, if you don’t like
the size of finance, just wait a while. Finance has contracted rather dramatically
since 2007.
Many puzzles remain, however, and the current academic literature paints an
interesting and quite novel picture of how the finance industry functions—and
maybe does not function.

The Controversy over Active Management Fees

Management fees are a big part of the “size of finance.” Fees aren’t GDP, of
course, but they are much more easily measured. The large overall rise in fee revenue
reflects several offsetting trends. Individuals moved investments from direct hold-
ings to mutual funds, and then to index funds or other passive funds. This trend
continues. New investors in defined contribution plans invest almost exclusively in
mutual funds or exchange-traded funds.
Mutual fund fee rates came down sharply, in part reflecting the slow shift
to very low-fee index and semi-passive funds, and in part reflecting competi-
tive pressure. French (2008) reports that the average actively managed equity
mutual fund fee fell from 2.19 percent in 1980 to 1 percent in 2007. Greenwood
and Scharfstein (2012) report that average bond fund fees fell from 2.04 to
0.75 percent. Some index funds charge as little as 0.07 percent. Fee-based advisers
and wealth managers are lowering fees, and bundling larger arrays of services,
including tax and estate planning.
Funds are far more efficient vehicles for individual investors than holding
individual stocks. The measured GDP of the fund industry is at least in part a
benefit rather than a cost, as it displaces inefficient and unmeasured home produc-
tion of financial management services. Hiring a (legal) house cleaner also raises
measured GDP.
Thus, mutual fund fee revenues reflect declining rates multiplied by a much
larger share of assets under management. This market does reflect sensible forces,
if one is willing to grant a rather long time span for those forces to affect industry
structure. But after all, the moves to low-cost airlines and big-box retailers took a
while too.
However, at the same time that individuals were moving to passive funds and
those funds were expanding, high-wealth individuals and institutions (pensions,
endowments, sovereign-wealth funds, and so forth) moved their investments
to hedge funds, private equity, venture capital, and other even higher-fee and
more-active investment vehicles. Hedge fund fee rates are reportedly stable over
time, and surprisingly large: managers charge 1.5 –2.5 percent of assets each
year, and also 15–25 percent of profits. This part of the market offers the more
puzzling behavior.
32 Journal of Economic Perspectives

The Traditional View

High-fee active management and underlying active trading have been deplored
by academic finance for a generation. French (2008) offers a comprehensive
summary. French estimates that equity investors in aggregate, between 1980 and
2006, paid 0.67 percent per year in active management fees, whose present value he
estimated to equal 10 percent of their investments. French eloquently conveys the
view that these investors wasted their money.
The standard analysis divides investment returns into “alpha” and “beta.” We
run a regression of a fund’s returns on the returns of a low-cost index, both returns
in excess of the risk-free rate. Beta is the slope coefficient. Beta times the index
return is the component of the fund’s return that is earned for passively shouldering
systematic risk, and can be synthesized by the investor without paying fees. Alpha is
the intercept in this regression, and gives the mean of that part of the fund’s return
that cannot be easily replicated. Alpha is conventionally interpreted as the extra
return that the fund earns, on average, from the manager’s talent or superior infor-
mation, and therefore potentially worth paying a fee to obtain. Both alpha and beta
are, conceptually, one’s best estimate of this decomposition of returns going forward,
of course. Estimated alphas from past history contain a great deal of luck.
The average alpha of all equity mutual funds, before fees, is very nearly zero.
This result follows almost by accounting, since the portfolio of equity mutual funds,
taken as a whole, is almost exactly the value-weighted market portfolio.
The evidence on hedge fund, private equity, venture capital, and other returns
is complicated by survivor bias (funds that perform badly tend to drop out of
the data) and by difficulties of calculating benchmarks that appropriately reflect
the risks, time horizons, and illiquidity of these investments. But the academic
argument over whether such funds as a class provide substantial alpha ends up
arguing over a few percentage points one way or the other—hardly the promised
gold mines.
Mediocre average results for actively managed investments might not be
surprising. Entry into the business is relatively free. The average artist isn’t that
good, either.
But one might expect that, as in every other field of human endeavor, the
good managers would be reliably good. Michael Jordan’s past performance was
a good forecast of what would happen in the next game. Yet the nearly universal
conclusion of the academic literature is that there are no reliably “good” managers.
To evaluate this question, we must separate skill from luck. “Why did Warren
Buffet earn so much money?” is not a productive question. The classic technique
is to examine rules by which one might have chosen funds in the past, and then
study the subsequent returns of all such funds. Study after study finds no reliable
rule that one can use to identify funds that will perform well in the future, after
controlling for betas. (Carhart 1997 is an excellent example.) Fama and French
(2010) pursue a clever measurement that does not require one to hypothesize such
a rule. They show that the distribution of estimated alpha across mutual funds is
only very slightly wider than what one would expect if sample alphas were just due
 Finance: Function Matters, Not Size 33

to luck. Fama and French estimate (p. 1935) that the distribution of true alpha
has a standard deviation of only 1.25 percent on an annual basis, meaning that
only about one-sixth of funds have true alphas (gross, before fees) of 1.25 percent
or greater—while another one-sixth have “true alphas” of negative 1.25 percent or
worse. (True negative alpha is a bit of a puzzling concept. You should not be able to
reliably underperform the market either, as all I have to do is short what you buy.)
And all of this before fees.

A Supply-and-Demand View of Active Management and Its Fees

It seems the average investor should save 60 basis points a year and just buy
a passive index such as Vanguard’s Total Stock Market Portfolio. It seems that the
stock pickers should do something more productive, like drive cabs. Active manage-
ment and its fees seem like a total private, and social, waste.
Yet his hallowed view—and its antithesis—do not completely make sense. After
all, active management and fees have survived 40 years of efficient-market disdain.
Economists who would dismiss “people are stupid” as an “explanation” for a pricing
anomaly that lasts 40 years surely cannot use the same “explanation” for the persis-
tence of active management. Economists who think the evidence favors lots of
“inefficiencies” in the market are even less well placed to deplore active management.
They should conclude that we need more, or at least better, active management to
correct the market’s inefficiencies. Their puzzle is the inability of existing managers
to pick low-hanging fruit.
Progress is being made at last. Berk and Green (2004) have created a supply-
and-demand economic model that explains many of the basic facts of mutual fund
performance, flows, and fees. (Berk 2005 offers a simple exposition.)
Suppose that some fund managers do have alpha. Alpha, however, has dimin-
ishing returns to scale. Traders report that many strategies apply only to smaller stocks
(see evidence in Fama and French 2006) or that prices move against them if they try
to execute trades that are too large. As an example, suppose that a manager can
generate 10 percent risk-free alpha with $10 million in assets under management.
Suppose also that the manager’s fees are 1 percent of assets under management, and
suppose that the market does not go up or down. Then, in his first year, the manager
makes $1 million abnormal return. The manager pockets $100,000 and investors in
the fund receive $900,000.
Seeing these good results, investors rush in. But the manager’s idea cannot
scale past $10 million of assets, so the manager invests extra money in an index.
With $20 million under management, the manager generates $1 million alpha on
the first $10 million and nothing on the rest. The manager again receives 1 percent
of assets under management, which is now $200,000. But investors still get $800,000
alpha. More investors pour in.
The process stops when the manager has $100 million under management.
The manager still generates $1 million alpha, but now he collects $1 million in
fees. His investors get exactly zero alpha, the competitive rate of return. Everyone
is acting rationally.
34 Journal of Economic Perspectives

Berk and Green’s (2004) model is much more sophisticated than this simple
example. They include uncertainty in returns and a signal extraction problem for
investors, which give rise to interesting dynamics. A large literature has followed.
This model explains many puzzling facts: In equilibrium, returns to investors
are the same in active and passively managed funds. Funds earn only enough alpha
to cover their fees. Good past fund returns do not forecast good future returns.
Investors chase managers with good past returns anyway, seemingly irrational
behavior and thus one of the most famous puzzles in the mutual fund literature (for
example, Chevalier and Ellison 1997). Returns to investors do not measure alpha.
Fees do. Managers with good track records get paid a lot.
This model is the focus of the current debate. Fama and French (2010) complain
that the average alpha before fees is nearly zero —and negative, not zero, after fees.
Berk and Van Binsbergen (2012) answer that Fama and French’s benchmarks are not
tradable, and skill should be measured as alpha times assets under management, as
0.1 percent alpha on a billion dollars is a lot. Using these measures, they find investors
just about breaking even, and a good deal of positive skill. (Fama and French’s Table AI
agrees.) The model needs to be brought to the data quantitatively: Does the magnitude
of fund flows following performance follow the model’s predictions? Does it describe
fund exit, the persistence of negative alpha, and the shift to passive management?
Like all models, one can explore deeper foundations. What is this alpha, anyway? Why
are fees a flat percentage of assets under management? If the manager could simply
charge a $1 million fee to start with, the fund would not need to expand.
And all that is how it should be. After 40 years, the research agenda is finally
about how to fit the facts into a supply and demand framework. Arguing about
benchmarks, calibration, and optimal contracts is a lot more productive than
deploring the financial industry as folly, or declaring that if it survives, markets must
be working. The answer will surely not end up all on one side or another: Surely
some investors have overpaid for pointless trading. Surely there is some durable
value in an industry that has lasted so long. Surely there are some understandable
distortions. On this path, we may finally understand how this market works, and
maybe, humbly, suggest some improvements. This is a great example of how the
economic framework operates—and a sobering reminder of how long it often takes
to see that a straightforward economic analysis is possible.

Is It Silly to Pay a Proportional Fee?

Much of the argument that “finance is too big” rests on the view that fees based
on a proportion of assets under management are a suboptimal contract. Assets
under management went up, fees went up, and managers laughed all the way to the
bank. This is a big part of Greenwood and Scharfstein’s story in this issue. On closer
examination, this argument seems awfully strained.
First, we have seen in fact a substantial decline in many fees and migration
to lower fee vehicles, in mutual funds, exchange traded funds, and many wealth
management services. Competition does seem to be working, though more slowly
than we may like.
 John H. Cochrane 35

Second, fee revenue is not a good measure of the “size” of finance. Fees are
a transfer, like gambling losses, not a measure of resources consumed or output
produced. Policy may and obviously does care a lot about transfers, but that is a
conceptually different question than worrying about wasted resources. Moreover,
fees vary based on outcomes. If the fund gains or loses money, fee income rises and
falls as well. Hedge fund fees, usually 2 percent of assets and 20 percent of profits,
vary enormously. The same fees that were puzzlingly high in 2006 were a lot lower
in 2008. Fees have much of the character of a risk-sharing arrangement among
co-investors, rather than an expense for professional services.
Third, if the fund doubles in value because everything else in the economy
doubles —capital stock, earnings, and so on—then surely by constant returns to
scale, the value of investment management (whatever that is) also doubles.
And finally, I’d like to see a specific claim as to what the alternative, realistic, and
privately or socially optimal contract is. Funds cannot bill by the hour, passing on
“cost” as lawyers do (or rather, used to do), for obvious monitoring and principal–
agent reasons. Should we agree to pay a fraction of initial investment, regardless
of subsequent performance? It’s obvious why we don’t do that. Accounting for
different vintages of investment would be a nightmare. It would also violate the
regulatory principle that all investors must be treated equally.
Proportional fees seem almost inescapable in funds that allow investors to with-
draw money and invest freely. Suppose funds charge 1 percent for new money, but
do not lower dollar fees after losses. Then after a fund has lost half its value, its inves-
tors face 2 percent fees going forward. They will quickly withdraw their money and
give it to a new fund. Funds that lost money would quickly spiral out of existence, or
investors would undermine the fee by withdrawing and reinvesting the next day as
new money. Venture capital, private equity, and some hedge funds do not allow free
withdrawal so for them, this argument does not apply as strongly—and they have
more complex fee structures.
Percentage fees pervade professional services. Real estate agents charge per-
centage fees, and do better when house prices rise. Architects charge percentage
fees. Contingency-fee lawyers take a percentage of winnings. Salesmen get percent-
age commissions. Even corrupt officials often take percentage bribes.
Perhaps the argument boils down to the claim that there is no alpha, so nobody
should pay any fees at all for active management. That’s a different question. If
there is alpha or some other function of active management, its optimal contract
is a difficult (and much-studied, though I do not review it here) principal–agent
problem. Skill is hard to measure, and a fund’s actions are hard to monitor. It seems
a big jump to conclude that percentage fees came into existence and have persisted
for decades, across a wide range of industries, while inflicting important private and
social costs, just because people are naive or irrational in some unspecified way.

Are Fee-Payers Naive?

Delegating active management and paying large fees is common and increasing
among large, completely unconstrained, and very sophisticated investors. For
36 Journal of Economic Perspectives

example, the Harvard endowment was in 2012 about two-thirds externally managed
by fee investors and was 30 percent invested in “private equity” and “absolute
return,” largely meaning hedge funds.1 The University of Chicago endowment is
similarly invested 2 in private equity and “absolute return.” Apparently, whatever
qualms some of its curmudgeonly faculty express about alphas, fees, and active
management are not shared by the endowment. Its most recent annual report states:
“The majority of TRIP’s [Total Return Investment Portfolio] assets are managed
by external managers specializing in a specific asset class, geography, or strategy.
These asset managers outperformed their respective benchmarks in every asset
class, adding over 500 basis points of performance versus the strategic benchmark.”
Five hundred basis points! Put that in your pipe and smoke it, efficient marketers.
At least we know one active manager’s perception of what they get for their fees.
These endowments’ approach to portfolio management is pretty much stan-
dard at endowments, nonprofits, sovereign wealth funds, family offices, pension
funds, and so forth—anywhere there is a big pot of money to invest. These investors
pay a lot of attention to allocation among name-based buckets, as represented in
the pie charts, “domestic equity,” “international equity,” “fixed income,” “absolute
return,” “private equity,” and the like. Then, they allocate funds in the buckets to
groups of fee-based active managers.
This approach bears no resemblance to standard portfolio theory, in which an
investor pays attention only to means and covariances, not buckets. And don’t even
ask how often hedge fund manager A is shorting what B is buying; what happens to
fees when you give a portfolio of managers 2 2++20 compensation and half of them
win and half lose; or why one would pay the manager of a growth-oriented fund to
buy the same stock that the manager of the value-oriented fund just sold.
Why have these decision procedures become standard practice? Vague refer-
ence to “agency problems” and “naiveté” seem unpersuasive. Harvard’s endowment
was overseen by a high-powered board, including its president Larry Summers,
possibly the least naive investor on the planet. The picture that Summers and his
board, or the high-powered talent on Chicago’s Investment Committee are simply
too naive to demand passive investing, or that they really want the endowments
to be invested in the Vanguard total market index, but some “agency problem”
with the managers they hire and fire with alacrity prevents that outcome from
happening, simply does not wash. (Yes, delegated portfolio management is a classic
principal-agent problem. But no, it’s hard to conceive that it produces this result.)
Perhaps instead, we should admit that standard portfolio theory is not much help
in situations of any real-world complexity, try to understand what these rough and
ready procedures achieve, and offer more helpful advice.
As for “excessive” compensation, in the first layer of fees (fees to the manager
who pays fees to the other managers) Harvard endowment’s CIO Jane Mendillo

1
See the Harvard Management Company website: http://www.hmc.harvard.edu/investment-management.
2
See the University of Chicago’s Annual Report, “The Endowment”: http://annualreport.uchicago.edu
/page/endowment.
 Finance: Function Matters, Not Size 37

was paid $4.7 million, most of which was straight salary.3 The University of Chicago’s
Mark Schmid gets only $1.8 million, though our measly $5.6 billion assets under
management relative to Harvard’s $27.6 billion may have something to do with it.
If major nonprofit university endowments are paying this much, is it really a puzzle
that pension funds do the same thing?

Finding Alpha? Implications for Active Trading

To justify fees for active management, one must explain why active trading is
worthwhile. The average investor theorem is an important benchmark: The average
investor must hold the value-weighted market portfolio. Alpha, relative to the
market portfolio, is by definition a zero-sum game. For every investor who over-
weights a security or invests in a fund that earns positive alpha, some other investor
must underweight the same security and earn the same negative alpha. Collectively,
we cannot even rebalance. And each of us can protect ourselves from being the
negative-alpha mark with a simple strategy: hold the market portfolio, buy or sell
only the portfolio in its entirety, and refuse to trade away from its weights, no matter
what price is offered. If every uninformed trader followed this strategy, informed
traders could never profit at our expense.

Alphas and Multiple Factors

Alpha seems a dicey proposition. But the last 20 years of finance research is as
clear as empirical research in economics can be: There is alpha relative to the market
portfolio—there are strategies that deliver average returns larger than the covariation
of their returns with the market portfolio justifies—lots of it, and all over the place.
In Cochrane (2011), I provide a summary of this huge literature; I won’t provide a
separate citation for each fact here.
Examples of such strategies include value (stocks with low market value relative
to accounting book value), momentum (stocks that have risen in the previous year),
stocks of companies that repurchase shares, stocks of companies with accounting
measures of high expected earnings, and stocks with low betas. The “carry trade”
in maturities, currencies and, credit—buy high-yield securities, sell low-yield
securities—and writing options, especially the “disaster insurance” of out-of-the-
money put options, all generate alpha. Expected returns on the market and most
of the anomaly strategies vary predictably over time, implying profitable dynamic
trading strategies.
Many of these anomalies lead to new “factors,” new dimensions of “systematic”
risk and rewards. For example, if one buys a large portfolio of “value” (low-price)
stocks, engineered to have zero correlation with the market, thinking that one
will reap the value-stock alpha and diversify away the risks, one soon discovers the

3
See “Chart: Top Paid CIOs of Tax-Exempt Institutions,” http://www.pionline.com/article/20111107
/CHART04/111109905.
38 Journal of Economic Perspectives

tendency of all value stocks to rise and fall together. The portfolio remains risky no
matter how many stocks one adds. In this way, pursuing the “value” alpha requires
one to take on this additional dimension of undiversifiable risk.
As formalized in Fama and French’s (1996) three-factor model and its larger
successors, the world appears to have many such “factors,” acting as the market return
factor did in our early understanding, each offering orthogonal dimensions of risk
and a return premium to those investors who are willing to take the risks. Those
“factor premiums” capture most of observed “alpha” relative to the market portfolio.
Large risk premiums opened up in the recent financial crisis, as prices of very
nearly identical securities diverged. For example, corporate bonds traded at lower
prices than their synthetic replication by a Treasury bond and a credit default swap.
The “covered interest parity” condition failed: You could earn money by borrowing
dollars, buying euros, investing in European money markets, and converting back
to dollars in the futures markets. If you could borrow dollars! These events and
other price movements in the crisis suggest to the researchers studying them “fire
sales,” “financial constraints,” “financial frictions,” “price pressure,” and “limits to
arbitrage”—all of which are ways of saying that the active managers of the time were
insufficient to equalize prices of nearly identical securities, and active traders could
have made alphas. Similar pricing divergences and insufficient arbitrage appeared
in the trading frenzies of the Internet boom (for example, Lamont and Thaler
2003; Cochrane 2003).
There are multiple dimensions of risk, and bearing these risks generates
expected-return rewards, rewards that change over time. These facts are not really
under debate. Their interpretation is. These alphas might represent imperfect risk
sharing and (often temporary) market segmentation, or “sentiment,” irrational
attachment or aversion to broad categories of securities. They might also reflect
a multidimensional and time-varying nature of risk premiums in a fully-integrated
and informationally efficient market. They certainly look less and less like “informa-
tion” about individual securities that is somehow improperly reflected in prices.
These facts and interpretations lend a quite new color to our central questions:
Is the financial sector too large or too small? How should investors behave in a world
with multiple dimensions of systemic risk? What is the economic function of active
management, and the economic value of management fees?

Multidimensional Risk-Sharing
The conventional disdain for active financial management is based on a conven-
tional perspective: The market portfolio is the one and only source of “systematic”
risk which generates a premium. It is accessible through low-cost passive invest-
ments. The investor understands this opportunity and knows how much market risk
he or she wishes to take. Alpha represents the trader’s knowledge of information
not reflected in market prices.
But the dozens of semi-passive strategies, each of which produce alpha (relative
to the market), each of which exposes the investor to new dimensions of undiversifi-
able risk, and many of which are poorly understood, changes the picture completely.
 John H. Cochrane 39

Each investor needs to decide which of the many sources of risk he or she is best
able to bear, or needs to avoid despite their attractive premiums.
Investors need to consider the even larger set of asset market risks that do not
bear premiums. Before chasing alphas, investors should hedge the risks of their
jobs, businesses, outside income streams, real estate, or peculiar liability streams
by setting up portfolios of assets whose returns are negatively correlated with those
risks. You should want a portfolio that rises when there is bad news about your
future income. Curiously, academic finance has done little to characterize these
nonpriced risks and prescribe hedging strategies.
One can see this process beginning. Many pension funds are moving towards
bond-like investments to match their liabilities. University endowments are begin-
ning to recognize how their liability streams affect investments. They thought of
themselves as “long term” investors able to reap the premiums of illiquid invest-
ments, and able to wait patiently through market downturns, until many in the
crisis realized they were supporting a bond-like liability stream in salaries of tenured
professors, and were leveraged by bond-financed construction. They found them-
selves trying to sell illiquid assets at the bottom like everyone else. Now, they are
thinking about matching endowment funding to projects that can bear risk and
adapting portfolios to their cash flows, including the implicit beta that alumni
donations rise when the stock market goes up. Endowments are recognizing that
their objectives include an important tournament relative to other universities
(Goetzmann and Oster 2012). The wealth-management arms of big banks help to
set up hedge portfolios for executives who have large unsaleable stock or option
positions, to help them come as close to shorting their own business as possible.
Websites available to individual investors are starting to emphasize intelligent and
individual-specific choice of “style” rather than promise generic “alpha.”
But none of this is easy. Merton (1971) described state-variable hedging
demands 40 years ago. Yet, with thousands of following papers, academic portfolio
theory still really does not offer clear-cut real-world advice (Cochrane forthcoming).
The nature and amount of multidimensional systematic risk one should take
is also much more nebulous and difficult to assess than the traditional question of
how much market risk one should take. Should you write put options, to earn the
premium? Or maybe you should buy put options as disaster insurance? Are you posi-
tioned to buy value stocks? To take on the credit risks of default? To take the risk
that high-interest rate foreign currencies depreciate against the dollar? Do the alpha
premiums these strategies offer compensate for the risks you will suffer when they
lose money? The whole alpha/beta definition is falling apart.
Even then, taking advantage of time-varying multidimensional risks requires
technical knowledge. Do you know how to write a credit default swap contract, how
to make stock momentum strategy work without drowning in transactions costs,
how to take advantage of temporarily high put option premiums in the euro-zone,
or even how reliably to buy a “value” portfolio? Because such questions are not easy,
portfolio problems like this might certainly benefit from professional and special-
ized management, and such management ought to be able to charge a fee.
40 Journal of Economic Perspectives

Perhaps some of the puzzling features of investment practice might be under-

stood as a rough and ready way of adapting to this more realistic portrait of risks and
returns. If so, some active management and dynamic trading represents a form of
socially beneficial insurance provision.
Hedge funds might make more sense in this investment world. They can move
to and from asset classes as risk premiums change, and by using leverage and deriva-
tives they can alter overall exposures quickly without incurring the transactions cost
of buying and selling large portfolios.
Many of these alpha-generating strategies and new “factors” suggest needed
institutional developments. As a concrete and recent example, consider the “betting
against beta” anomaly reexamined by Frazzini and Pedersen (2011a, b). They docu-
ment that low-beta stocks get higher average returns than they should, and high
beta stocks get lower returns than they should. Their interpretation is that many
investors want more risk than the market portfolio provides, yet leverage is costly
to obtain. These investors buy high-beta stocks instead of leveraging, driving up the
prices of high-beta stocks, and vice versa for low-beta stocks. In this setting, arbitra-
geurs cannot help. The problem is the price of risk, needing wider risk-sharing, not
an arbitrage (riskless profit) opportunity. To bring prices back to what they should
be, we need low-cost vehicles to bring leveraged low-beta investments to the part of
the investing public that wants them—which, perhaps not coincidentally, Frazzini
and Pedersen’s company provides.
We have seen this kind of institutional development before. Small stocks were
one of the first prominent anomalies, generating (it appeared) higher average returns
than their betas justified. But it was hard for individual investors to hold a diversified
portfolio of small stocks. Arbitrageurs could only do so much, because small stocks
move together, so a concentrated portfolio bears undiversifiable risk. Small stock
mutual funds were started, which allowed a mass of investors to participate. Fees and
expenses of those funds contributed to revenue and measured GDP, in a way that the
activities of individual investors holding small stocks did not. But they allowed the risk
of small stocks to be widely shared and the small stock premium to decline.
So far I have made no mention at all of informational inefficiency, exploiting
mispricings, superior information, or winning the zero-sum alpha game. I have not
violated the average investor theorem. Given the new facts of empirical finance, a
large role for active management exists without any of that at all. Of course, I do not
claim that current portfolio practice, and especially hiring many different high-fee
hedge funds, is an optimal strategy. But it isn’t necessarily as “naive” or “agency
conflicted” as it otherwise seems.

Marketing
In the quest to explain the persistence of active management and its fees,
one other analogy seems worth pursuing: marketing. Marketing and advertising
have long been a puzzle to economists, along with readers of Consumer Reports and
coupon-clippers everywhere. Why buy the brand name when the generic is nearly
identical, and costs a lot less?
 Finance: Function Matters, Not Size 41

The money-management industry is essentially a marketing industry. Its practi-

tioners take generic ingredients, package, label, advertise, and market them. Yes, it’s
puzzling that people don’t buy the generic at Vanguard. It’s puzzling that they don’t
buy the pieces and assemble their own, with E*TRADE. It’s puzzling that they pay
so much for the slight differences in ingredients that the active managers deliver.
And it is equally puzzling that they pay for Coke, Clorox, Bayer, or bottled water;
that they shop at Macy’s not Target, Whole Foods not Costco, and a hundred other
brand names.
This is not the place to digress into the “rationality” of marketing and adver-
tising. Simply dismissing centuries worth of branding and advertising as naiveté
and folly seems, well, its own form of naiveté. Perhaps by thinking of active fund
management as an instance of this larger pattern, we may make some progress to
understanding how it actually works.

Information Trading and Price Discovery

Much trading and active management, however, is clearly aimed at bringing

information to the market, not at better sharing of time-varying and multidimensional
risk. The first welfare theorem does not clearly apply to information production, so
we have little a priori reassurance that the quest for trading profits produces the
“right” amount—or, perhaps more importantly, the right kind— — of information.
It is possible that not enough social resources are devoted to trading, because
information is a public good. As French (2008) wrote, despite deploring the private
costs of alpha-chasing: “I offer no evidence on whether society is buying too little
or too much of this good. Price discovery, however, is an externality—each active
investor pays the full cost of his efforts but captures only a tiny slice of the benefit—
so there is no reason to think active investors purchase the optimal amount of
price discovery.”
The common complaints “the financial crisis proves markets aren’t efficient,”
or that tech and mortgages represented “bubbles,” are at heart complaints that
there was not enough active information-based trading. All a more “efficient” market
could have done is to crash sooner, by better expressing the pessimist’s views.
Remember, “efficiency” means that prices incorporate all available information, not
that markets are clairvoyant. The definition of “efficiency” is widely misunderstood.
I once told a newspaper reporter that I thought markets are pretty “efficient,” and
he quoted me as saying markets are “self-regulating!”
If information is not incorporated into market prices and to such an extent
that simple strategies with big alphas can be published in the Journal of Finance,,
there are not enough arbitrageurs. If asset prices fall in “fire sales,” only to rebound
later, there are not enough buyers following the fire trucks. If credit constraints are
impeding the flow of capital, there is a social benefit to loosening those constraints.
The literature on short-selling is revealing on this point. Short sellers uncover
far more financial fraud than the Securities and Exchange Commission. Conversely,
42 Journal of Economic Perspectives

some of the biggest alphas and “inefficiencies” occur when there is a technical or
regulatory impediment to short seller’s activities. Lamont (2012) finds 2.4 percent
monthly alpha to a portfolio of short-selling-constrained stocks, a large informa-
tional inefficiency. This is a concrete example of inadequate (because constrained)
information-based trading.
Information trading produces more informationally-efficient prices, which
are socially useful. With better market signals, companies raise capital more easily
for valuable projects, and are signaled not to invest in poor projects or at poor
times. True, the simple q theory, which predicts that corporate investment should
be a perfect function of stock price relative to book value, is formally rejected, but
its glass is also half full: There are strong correlations between stock prices and
investment, over time (through the tech boom and bust of the 1990s and through
the financial crisis (see Cochrane 1991; 2011, Figure 10)) and across industries
(Google versus, say, GM). When issuing stock generates a lot of money, companies
do it, and build factories or websites. Those who view asset market booms and
following busts as “irrational” or “bubbles” point to the consequent investment
booms and busts as examples of the social costs of inefficient markets, thereby
endorsing the social value of more efficient markets.
Even without investment, more efficient prices provide better risk sharing, as
in an endowment economy. If the owner of an apple tree and that of a pear tree
hedge their risks by trading stock in the other tree, their risk-sharing improves
when stock prices are more efficient. (Hirshleifer’s, 1971, famous analysis stating
that efficiency is only socially beneficial if production is involved did not treat such
risk-sharing).
Information trading is central to “liquidity provision” and thus the success of
markets for risk sharing. Markets such as Consumer Price Index, GDP futures or
hurricane catastrophe options failed because there was not enough information
trading. This is an important external benefit. Indeed, in the public forum, hedge
funds and high-frequency traders primarily defend their activities by touting their
“market making” and “liquidity provision” for small investors. (Of course, they are
also pandering to their regulators’ tastes here.)

The Puzzle of Information Trading

Still, the cacophony of trading seems like a lot of effort for these goals. The
classic theory of finance predicts that information is perfectly reflected in prices,
with no trading volume needed. Suppose Apple is trading at $500 per share, but you
know that the iPhone 6 will make Apple worth $1,000 per share. If you approach an
uninformed investor with an offer to buy Apple at $600 per share, the index investor
should answer: “No, you must know something I don’t know. I only buy and sell the
entire index, so I don’t lose to people like you.” If you offer $700, the index investor
answers: “I don’t think you heard me. I only buy and sell the entire index.” You keep
trying, bidding the price up all the way to $1,000 per share, at which point you give
up. The price rises, reflecting your information, but no trade occurred. This is a
colloquial version of Milgrom and Stokey’s (1982) famous no-trade theorem.
 John H. Cochrane 43

The theory that prices reflect information with zero trading volume is of
course dramatically at odds with the facts. The classic theory also ignores costs. If
information traders cannot earn positive alpha, and if producing information and
trading on it takes any time and resources, the information traders won’t bother,
and nobody is left to make prices reflect information. For this reason, as Grossman
and Stiglitz (1980) wrote, informationally efficient markets are impossible.
The standard compromise model (Grossman and Stiglitz 1980; Kyle 1985;
and a huge literature) posits “informed” traders who receive a signal about a firm’s
value, “liquidity” traders who for unspecified reasons must trade, and “market
makers” who intermediate, charging a bid-ask spread to defend themselves against
the informed traders.
Now, all current theories of trading rely on some sort of “irrationality” or other
artificial assumptions. “Liquidity traders” are the classic example. Other models,
like Scheinkman and Xiong (2003), posit slightly irrational dogmatic beliefs so each
information trader can believe he or she is smarter than average. Many models,
such as Acharya and Pedersen (2005), write down overlapping generations of agents
without bequests who die every week or so, forcing them to trade.
But these assumptions are convenient shortcuts for getting trading into the
model for other purposes, such as studying price discovery and liquidity. They are
not there to describe microfoundations of socially destructive trading that needs
remediation by policy. The “irrationality” that breaks the no-trade theorem, or
the irrationality of the liquidity traders, is not typically deeply micro-founded in the
psychology literature, as in true behavioral finance. People live more than a week,
and leave bequests.
The fact staring us in the face is that “price discovery,” the process by which
information becomes embedded in market prices, uses a lot of trading volume, and
a lot of time, effort, and resources. And we are only beginning to understand it.
The empirical literature offers tantalizing glimpses of this process. A very small
taste of this vast literature: The period after a news announcement often features
high price volatility and trading volume, in which markets seem to be fleshing
out what the news announcement actually means for the value of the security. For
example, Lucca and Moench (2012, Figure 6) show a spike in stock-index trading
volume and price volatility in the hours just after the Federal Reserve announce-
ments of its interest rate decisions. The information is perfectly public. But the
process of the market digesting its meaning, aggregating the opinions of its traders,
and deciding what value the stock index should be with the new information, seems
to need actual shares to trade hands.4 Perhaps the common model of informa-
tion— essentially, we all agree on the deck of cards, we just don’t know which one
was picked—is wrong.
Securities such as “on the run” or benchmark bonds, where “price discovery”
takes place, have higher prices than otherwise identical securities. Traders are

4
Banerjee and Kremer (2010) and Kim and Verrecchia (1991) offer models in which such disagreement
about public information leads to trading volume.
44 Journal of Economic Perspectives

willing to suffer lower average returns in order to participate in the information-

trading game, in much the same way as money holders suffer lower returns for the
transactions services money provides (see Cochrane 2003 and references therein).
Similarly, “liquidity” seems to be extremely valuable to investors and has been so for
a long time, even though none of us feel the need to trade every 10 minutes.
Markets in financial securities are set up, and exist, almost entirely to be markets
for information trading, and high-frequency “liquidity provision,” that we find hard
to fathom. They are not really markets for the securities themselves. We could easily
handle individuals’ lifetime saving and dissaving needs, and firms’ need to issue and
retire equity, with orders-of-magnitude less volume, in much sleepier bank-like insti-
tutions. Yes, we could each avoid being the negative-alpha part of price discovery by
only buying index funds. It’s a bit of a puzzle that we don’t. It’s also a good thing we
don’t, or there would be no traders making prices efficient.
But as with active management, perhaps we should work just a little harder
before dismissing the hundreds of years of trading activity, and the entire existence
of the New York Stock Exchange, Chicago Mercantile Exchange, and other markets,
as monuments to human folly, or before advocating regulations such as transactions
taxes —the perennial favorite answer in search of a question—to reduce trading
volume whose size, function, and operation we do not understand. Are we sure that
they should not be transactions subsidies?
And before we deplore, it’s worth remembering just how crazy passive indexing
sounds to any market participant. “What,” they might respond, “would you walk
in to a wine store and say ‘I can’t tell good from bad, and the arbitrageurs are out
in force. I sure won’t pay you 1 percent for recommendations. Just give me one
of everything’?”

High-Frequency Trading and Market-Making

It’s especially hard to see why high-frequency trading is needed. Price discovery
every millisecond doesn’t seem necessary to guide corporate investment or indi-
vidual risk sharing and hedging.
High-frequency trading reminds us in the extreme that the amount of trading
based on a well-understood or “fundamental” piece of information about a compa-
ny’s cash flow is minuscule. Models in which an informed trader possesses a “signal”
about the value of a liquidating dividend just don’t describe the vast majority of
trading. High-frequency traders do not trade on earnings reports 20 milliseconds
ahead of the market.
Instead, high-frequency traders—and even most “low-frequency” day and week
traders—look at patterns of prices, volumes, and past trading activity, not “informa-
tion” or opinion about firm fundamentals.
They may describe their strategy as “statistical arbitrage,” removing the small
predictability of high-frequency price movements (and grossly misusing the term
“arbitrage.”) Sometimes they defend their social function as “market makers” or
“liquidity providers.” If so, market making is a far more dynamic process than simply
posting bid-ask spreads, as the standard theory envisions! If you ask their critics, they
 Finance: Function Matters, Not Size 45

are artfully front-running demand from less-sophisticated investors, subtracting

“liquidity,” worsening “price impact,” choking bandwidth with quickly-canceled
orders and removing the economic rewards to genuine information trading. Their
activity may also answer the interesting question of how information spreads from
one informed trade to the whole market. Somebody has to notice the price pattern
and pile in.
However we come to understand these issues, the social costs and benefits of
high-frequency trading are clearly not at all related to the minor (as a fraction
of GDP) resources devoted to them—the cost of possibly useless fiber-optic cable,
co-located servers, and the time of smart programmers who could be developing
better iPhone games. The social question for high-frequency trading—like all of
finance, really—is whether it screws up markets or makes them more efficient
and “liquid.”
There isn’t yet much evidence or theory on this point, but isolated events
suggest doubts about liquidity-provision and efficiency. For example, in the May 6,
2010 “Flash Crash,” the Standard and Poor’s 500 fell 6 percent in a few minutes after
a large sell order arrived, and promptly recovered in less than an hour, only after a
five minute trading halt. Kirilenko, Kyle, Samadi, and Tuzun (2011) who study this
event (see their Figure 1) document that high-frequency traders absorbed demand
for about four seconds before turning around and selling along with everyone else.
On July 19, 2012, Coke, McDonalds, IBM, and Apple saw price sawtooths: sharp rises
exactly on each hour, reversed by the next hour. Vigna and Lauricella (2012) offer
some amazing graphs.5 These movements were widely attributed to an algorithm
placing big orders exactly on the hour—and other algorithms not picking up on the
inefficient signal abundantly obvious to the human eye. These palpable inefficien-
cies suggest a market with very little “liquidity provision,” not the opposite.
The structure of markets, with design and regulation stemming from the days
of human trading, could be at fault. Prices must jump in discrete intervals—once
1/8 dollar, now 1 cent. Limit orders must be filled in strict time priority: if order A
arrives before order B, order A must be filled completely and B gets nothing. Yet
time is continuous. A’s order need only arrive a millisecond before B’s, and A wins
the pot. (Traders report that the ability to quickly cancel limit orders that are in
the back of the line is another advantage of very high speed.) You can see an arms
race for speed emerge. It’s worth spending a lot on computers to speed up trades
by a few milliseconds.
If my hunch is correct, it suggests an obvious solution: Suppose that an
exchange operated on a discrete clock, as a computer does in order to let signals
settle down before processing them. The exchange could run a once-per-second, or
even once-per-minute, matching process, with all orders received during the period
treated equally. If there are more buy than sell at the crossing price, orders are

5
The website http://www.nanex.net/FlashCrash/OngoingResearch.html is devoted to weird behavior
in high-frequency markets.
46 Journal of Economic Perspectives

filled proportionally. Such an exchange would eliminate extremely high-frequency

trading, because there would be no gain or loss from acting faster than a minute.
Would this system be an improvement, to efficiency and liquidity? Would
exchanges choose such systems if they were allowed to do so? The Taiwan Stock
Exchange already matches limit orders once every 90 seconds (Barber, Lee, Liu,
and Odean 2008). Is its performance atrociously worse? These are all good ques-
tions! High-frequency trading is a ripe area of research.

Housing, Consumer Credit, and the Size of Regulated Finance

The growth of housing finance and consumer credit raises a different set of
issues. It’s useful to divide the mortgage business into three parts: mortgage origina-
tion, mortgage refinancing, and mortgage-backed securities.
The increase in fees for residential loan origination is easily digested as the
response to an increase in demand. The increase in housing demand may indeed
not have been “socially optimal” (!). There are plenty of government policies and
perhaps a few market dislocations to blame. But it doesn’t make much sense to
criticize growth in the financial industry for responding to this increase in demand,
whatever its source, or for passing along the subsidized credit—which was and
remains the government’s explicit intention to increase—with the customary fee.
The large fees collected for refinancing mortgages are a bit more puzzling. US
mortgages are strangely complicated, predominantly featuring fixed rates, no penalty
for prepaying when interest rates fall, limited recourse, and a complex refinancing
option. Other countries have gravitated to much simpler contracts. The now-familiar
structure of US mortgages emerged after only the Great Depression, when new federal
agencies started issuing them. Before the Great Depression, US mortgages lasted only
five to ten years and required only the payment of interest. The principal was due at
the end of the loan, and was typically refinanced (Green and Wachter 2005, p. 95).
Today, the structure of mortgage contracts is pretty much dictated by what the govern-
ment agencies that dominate the market will buy and guarantee.
These observations suggest that such complex contracts are not a market
necessity. However, a glance at my cellphone contract and frequent flyer miles rules
suggests to me that price discrimination by needless complexity might be part of
the story as well.
Still, collecting fees when interest rates decline or consumers refinance is not
conceptually part of GDP. They are state-dependent transfers dictated by the terms
of an option contract. And we are unlikely to see a lot of refinancing as interest rates
eventually rise.
There was a lot of financial innovation in mortgage-backed securities, some
of which notoriously exploded. But here again, whether we spend a bit of GDP
filling out forms or paying fees is clearly the least of the social benefit and cost ques-
tions. The “shadow banking” system was prone to a textbook systemic run, which
happened. This fragility, not the size or fraction of GDP, is the important issue.
 John H. Cochrane 47

A good part of this innovation, such as creating off-balance-sheet, special-

purpose vehicles and tailoring securities in order to game credit ratings, was
clearly designed to engineer around ill-conceived regulations. That part counts
as a regulatory failure needing reform, rather than a market failure needing
additional regulation.
Yet much of this financial innovation has the potential to be of large social
benefit. Suppose that mortgages were bundled into securities, intermediated by
mutual funds whose values float, just like those of equity mutual funds, and held
around the world in retirement accounts, pension funds, and our endowments’
portfolios, without government guarantees at every step. This would be a terrific
financial structure. Though mortgage-backed securities are a bit opaque, they are
nowhere near as opaque as the entire balance sheet of, say, Citigroup. Further-
more, such a structure would be immune to runs, bankruptcies, and bailouts, thus
requiring minimal regulation. And the fees required to fill out the mortgage-backed
security paperwork would surely be less than the bank and regulatory paperwork,
regulation, and compliance costs of the current system.

Concluding Remarks

The size and revenues of the finance industry increased because fee income for
refinancing, issuing, and securitizing mortgages rose along with the rise in housing
transactions and house prices, and because asset-management fee income rose
along with a shift to professional management from “roll-your-own” portfolios and a
rise in asset values. Compensation to employees with skills in short supply increased.
Fee schedules themselves declined a bit. These facts suggest “demand shifted out,”
not “something big changed in the structure of this industry.”
Demand that shifts out can shift back again. Demand for financial services
evaporated with the decline in housing and asset values in the 2008 recession and
subsequent period of sclerotic growth. Much of the “shadow banking system” has
disappeared. For example, asset-backed commercial paper outstanding rose from
$600 billion in 2001 to $1.2 trillion in 2007—and now stands at $300 billion. Finan-
cial credit market debt outstanding in the flow of funds rose from $8.6 trillion in
2000 to $17.1 trillion in 2008—and now stands at $13.8 trillion. Employment
in financial activities rose from 7.7 million in 2000 to 8.4 million in 2007—and
is now back to 7.7 million (according to the Bureau of Labor Statistics). Study
of “why is finance so big,” using data that stops in 2007, may soon take its place
alongside studies of “why are Internet stocks so high” in 1999 or studies of “why is
there a Great Moderation” in 2006.
An older literature on the size of the financial system, forgotten in the current
debate, studies the socially inefficient resources devoted to cash management in the
face of positive interest rates, measuring social costs as the area under the money
demand curve. Lucas (2000) concluded that finance was about 1 percent of GDP
too big by this measure. The fragility of those cash-management schemes can now
48 Journal of Economic Perspectives

be added to the list of social costs. Zero interest rates have eliminated these costs
for now, and if the Fed continues to pay market interest on reserves, those costs can
remain largely eliminated in the future.
The size question for the finance industry going forward, under the Dodd–Frank
regulatory structure, is likely to be how many resources are devoted to regulation,
regulatory compliance, lobbying to influence those regulations, and the distortions
they induce. The social cost question remains how to create a financial system that
is not prone to runs, crashes, and bailouts, even if that costs a few percentage points
of GDP. Unless sovereign debt bites us first.
Many puzzles remain in the structure of the finance industry. The persistence of
high-fee active management chosen by sophisticated institutional investors remains
a puzzle. To some extent, as I have outlined, this pattern may reflect insurance
provision, that is, the dynamic and multidimensional character of asset-market risk
and risk premiums. To some extent, this puzzle also goes hand in hand with the
puzzle of why price discovery seems to require so much active trading, and whether
and how information trading provides valuable “liquidity.” It is possible that there
are far too few resources devoted to price discovery and market stabilization. In the
financial crisis, we surely needed more pools of cash prepared to pounce on fire
sales, and more opportunities for negative long-term views to express themselves.
Surveying the current economic literature on these issues, it is certain that we
do not very well understand the price-discovery and trading mechanism, nor the
economic forces that allowed high-fee active management to survive so long.
Unless we adopt the arrogant view that what we don’t understand must be bad,
it is clearly far too early to make pronouncements such as “There is likely too much
high-cost, active asset management,” or “Society would be better off if the cost of
this management could be reduced.” Such statements are not supported by theory
or evidence. Nor is their not-so-subtle implication that resources devoted to greater
regulation—by politicians and regulators no less naive than current investors, no
less behaviorally-biased, armed with no better understanding than academic econo-
mists, and with much larger agency problems and institutional constraints—will
improve matters. This proposition amounts to Samuel Johnson’s dictum on second
marriages, “the triumph of hope over experience.”

References

Acharya, Viral V., and Lasse H. Pedersen. 2005. Individual Investors Lose by Trading?” Review of
“Asset Pricing with Liquidity Risk.” Journal of Finan- Financial Studies 22(2): 609– 632.
cial Economics 77(2): 375 – 410. Berk, Jonathan B. 2005. “Five Myths of Active
Banerjee, Snehal, and Ilan Kremer. 2010. Portfolio Management.” Journal of Portfolio Manage-
“Disagreement and Learning: Dynamic Patterns of ment 31(3): 27–31.
Trade.” Journal of Finance 65(4): 1269–1302. Berk, Jonathan B., and Richard C. Green. 2004.
Barber, Brad M., Yi-Tsung Lee, Yu-Jane Liu, “Mutual Fund Flows and Performance in Rational
and Terrance Odean. 2008. “Just How Much Do Markets.” Journal of Political Economy 112(6): 1269 –95.
 Finance: Function Matters, Not Size 49

Berk, Jonathan B., and Jules H. Van Binsbergen. http://www.people.hbs.edu/dscharfstein/Growth

2012. “Measuring Managerial Skill in the Mutual _of_Modern_Finance.pdf.
Fund Industry.” NBER Working Paper 18184. Grossman, Sanford G., and Joseph E. Stiglitz.
Carhart, Mark M. 1997. “On Persistence in 1980. “On the Impossibility of Informationally
Mutual Fund Performance.” Journal of Finance Efficient Markets.” American Economic Review
52(1): 57–82. 70(3): 393 – 408
Chevalier, Judith, and Glenn Ellison. 1997. Hirshleifer, Jack. 1971. “The Private and Social
“Risk Taking by Mutual Funds as a Response to Value of Information and the Reward to Inventive
Incentives.” Journal of Political Economy 105(6): Activity.” American Economic Review 61(4): 561–74.
1167–1200. Kim, Oliver, and Robert E. Verrecchia. 1991.
Cochrane, John H. 1991. “Production-Based “Trading Volume and Price Reactions to Public
Asset Pricing and the Link between Stock Returns Announcements.” Journal of Accounting Research
and Economic Fluctuations.” Journal of Finance 29(2): 302–21.
46(1): 207–234. Kirilenko, Andrei A., Albert S. Kyle, Mehrdad
Cochrane, John H. 2003. “Stock as Money: Samadi, and Tugkan Tuzun. 2011. “The Flash Crash:
Convenience Yield and the Tech-Stock Bubble.” The Impact of High Frequency Trading on an
Chap. 12 in Asset Price Bubbles edited by William C. Electronic Market.” http://papers.ssrn.com/sol3
Hunter, George G. Kaufman, and Michael Pomer- /papers.cfm?abstract_id=1686004.
leano. Cambridge: MIT Press. Kyle, Albert S. 1985. “Continuous Auctions and
Cochrane, John H. 2011. “Discount Rates.” Insider Trading.” Econometrica 53(6): 1315 –36.
Journal of Finance 66(4): 1047–1108. Lamont, Owen A. 2012. “Go Down Fighting:
Cochrane, John H. Forthcoming. “A Mean- Short Sellers vs. Firms.” Review of Asset Pricing
Variance Benchmark for Intertemporal Portfolio Studies 2(1): 1–30.
Theory.” Journal of Finance. Lamont, Owen A., and Richard H. Thaler. 2003.
Fama, Eugene F., and Kenneth R. French. 1996. “Can the Market Add and Subtract? Mispricing in
“Multifactor Explanations of Asset Pricing Anoma- Tech Stock Carve-Outs.” Journal of Political Economy
lies.” Journal of Finance 51(1): 55 – 84. 111(2): 227– 68.
Fama, Eugene F., and Kenneth R. French. 2006. Lucas, Robert E., Jr. 2000. “Inflation and
“Dissecting Anomalies.” Journal of Finance 63(4): Welfare.” Econometrica 68(2): 247–74.
1653 –78. Lucca, David O., and Emanuel Moench. 2012.
Fama, Eugene F., and Kenneth R. French. 2010. “The Pre-FOMC Announcement Drift.” http://www
“Luck versus Skill in the Cross-Section of Mutual .newyorkfed.org/research/staff_reports/sr512.html.
Fund Returns.” Journal of Finance 65(4): 1915 – 47. Merton, Robert C. 1971. “Optimum Consump-
Frazzini, Andrea, and Lasse Heje Pedersen. tion and Portfolio Rules in a Continuous Time
2011a. “Betting against Beta.” Paper available at Model.” Journal of Economic Theory 3(4): 373 – 413.
http://www.econ.yale.edu/~af227. Milgrom, Paul, and Nancy L. Stokey. 1982.
Frazzini, Andrea, and Lasse Heje Pedersen. “Information, Trade, and Common Knowledge.”
2011b. “Embedded Leverage.” Paper available at Journal of Economic Theory 26(1): 17–27.
http://www.econ.yale.edu/~af227. Pensions & Investments. 2011. “Chart: Top Paid
French, Kenneth R. 2008. “Presidential Address: CIOs of Tax-Exempt Institutions.” November 7.
The Cost of Active Investing.” Journal of Finance http://www.pionline.com/article/20111107
63(4): 1537–73. /CHART04/111109905.
Goetzmann, William N., and Sharon Oster. Scheinkman, Jose, and Wei Xiong. 2003. “Over-
2012. “Competition among University Endow- confidence and Speculative Bubbles.” Journal of
ments.” NBER Working Paper 18173. Political Economy 111(6): 1183 –1219.
Green, Richard K., and Susan M. Wachter. 2005. Vigna, Paul, and Tom Lauricella. 2012.
“The American Mortgage in Historical and Inter- “Sawtooth Trading Hits Coke, IBM, McDonald’s,
national Context.” Journal of Economic Perspectives and Apple Shares.” Wall Street Journal, July 19.
19(4): 93 –114. http://blogs.wsj.com/marketbeat/2012/07/19
Greenwood, Robin, and David Scharfstein. /sawtooth-trading-hits-coke-ibm-mcdonalds
2012. “The Growth of Modern Finance.” July, 2012. -and-apple-shares/.
50 Journal of Economic Perspectives
This article has been cited by:

1. Tarun Chordia, Bin Miao. 2020. Market efficiency in real time: Evidence from low latency activity
around earnings announcements. Journal of Accounting and Economics 70:2-3, 101335. [Crossref]
2. Ryan Erhard, Richard G. Sloan. 2020. Does low latency trading improve market efficiency? A
discussion. Journal of Accounting and Economics 70:2-3, 101342. [Crossref]
3. Elizabeth Blankespoor, Ed deHaan, Iván Marinovic. 2020. Disclosure processing costs, investors’
information choice, and equity market outcomes: A review. Journal of Accounting and Economics
70:2-3, 101344. [Crossref]
4. Zijun Wang. 2020. The high volume return premium and economic fundamentals. Journal of Financial
Economics . [Crossref]
5. Nikolai Roussanov, Hongxun Ruan, Yanhao Wei. 2020. Marketing Mutual Funds. The Review of
Financial Studies 132. . [Crossref]
6. D. P. Frolov. 2020. From transaction costs to transaction value: Overcoming the frictional paradigm.
Voprosy Ekonomiki :8, 51-81. [Crossref]
7. Yang Sun. 2020. Index Fund Entry and Financial Product Market Competition. Management Science
. [Crossref]
8. Zijun Wang. 2020. The High-Volume Return Premium and Economic Fundamentals. SSRN
Electronic Journal . [Crossref]
9. David Ardia, Laurent Barras, Patrick Gagliardini, Olivier Scaillet. 2020. Hedge Fund Performance
under Misspecified Models. SSRN Electronic Journal . [Crossref]
10. Magnus Dahlquist, Markus Ibert, Felix Wilke. 2020. Expected Mutual Fund Performance. SSRN
Electronic Journal . [Crossref]
11. Pablo Kurlat. 2019. The Social Value of Financial Expertise. American Economic Review 109:2,
556-590. [Abstract] [View PDF article] [PDF with links]
12. Richard M. Frankel, S.P. Kothari, Luo Zuo. 2019. Economics of Accounting Earnings. SSRN
Electronic Journal . [Crossref]
13. Elizabeth Blankespoor, Ed deHaan, Ivan Marinovic. 2019. Disclosure Processing Costs and Investors’
Information Choice: A Literature Review. SSRN Electronic Journal . [Crossref]
14. Zijun Wang, Yan Qian, Shiwen Wang. 2018. Dynamic trading volume and stock return relation: Does
it hold out of sample?. International Review of Financial Analysis 58, 195-210. [Crossref]
15. Daniel Fricke, Austin Gerig. 2018. Too fast or too slow? Determining the optimal speed of financial
markets. Quantitative Finance 18:4, 519-532. [Crossref]
16. Gunther Capelle-Blancard. 2018. What is the Point of (the Hundreds of Thousands of Billions of )
Stock Transactions?. Comparative Economic Studies 60:1, 15-33. [Crossref]
17. Emmanuel Carré, Guillaume L’œillet. 2018. The Literature on the Finance–Growth Nexus in the
Aftermath of the Financial Crisis: A Review. Comparative Economic Studies 60:1, 161-180. [Crossref]
18. Leonard Kostovetsky, Alberto Manconi. 2018. How Much Labor Do You Need to Manage Capital?.
SSRN Electronic Journal . [Crossref]
19. Francesco D'Acunto, Laurent Frrsard. 2018. Finance, Talent Allocation, and Growth. SSRN Electronic
Journal . [Crossref]
20. Richard M. Frankel, S.P. Kothari, Luo Zuo. 2018. Why Shareholder Wealth Maximization despite
Other Objectives. SSRN Electronic Journal . [Crossref]
21. Laurent Barras, Patrick Gagliardini, Olivier Scaillet. 2018. The Cross-Sectional Distribution of Fund
Skill Measures. SSRN Electronic Journal . [Crossref]
22. Jay M. Chung, Jaeuk Khil, Jhinyoung Shin. 2017. Performance of Active and Passive Management of
Korea's NPS Funds. Asia-Pacific Journal of Financial Studies 46:4, 535-557. [Crossref]
23. Jean-Edouard Colliard. 2017. Catching Falling Knives: Speculating on Liquidity Shocks. Management
Science 63:8, 2573-2591. [Crossref]
24. Kirill Shakhnov. 2017. The Allocation of Talent: Finance Versus Entrepreneurship. SSRN Electronic
Journal . [Crossref]
25. Gunther Capelle-Blancard. 2017. À quoi servent les (centaines de milliers de milliards de) transactions
boursières ?. Revue d'économie financière 127:3, 37. [Crossref]
26. Emmanuel Carré, Guillaume L’Œillet. 2017. Une revue de la littérature récente sur le  nexus finance-
croissance après la crise : apports, limites et pistes de recherche. Revue d'économie financière 127:3,
271. [Crossref]
27. Arvid O.I. Hoffmann, Thomas Post. 2016. How does investor confidence lead to trading? Linking
investor return experiences, confidence, and investment beliefs. Journal of Behavioral and Experimental
Finance 12, 65-78. [Crossref]
28. Anna Agapova, James E. McNulty. 2016. Interest rate spreads and banking system efficiency: General
considerations with an application to the transition economies of Central and Eastern Europe.
International Review of Financial Analysis 47, 154-165. [Crossref]
29. Maya Eden. 2016. Excessive Financing Costs in a Representative Agent Framework. American
Economic Journal: Macroeconomics 8:2, 215-237. [Abstract] [View PDF article] [PDF with links]
30. Leonard E. Burman, William G. Gale, Sarah Gault, Bryan Kim, Jim Nunns, Steve Rosenthal. 2016.
Financial Transaction Taxes in Theory and Practice. National Tax Journal 69:1, 171-216. [Crossref]
31. Tobias Dieler. 2016. (In)Efficient Asset Trade, Differential Borrowing Cost and Central Bank Policy.
SSRN Electronic Journal . [Crossref]
32. Maximilien Demarquette. 2016. Strategic Trading with Endogenous Fundamental and Non-
Fundamental Information Acquisition. SSRN Electronic Journal . [Crossref]
33. David Easley, Maureen O'Hara. 2016. How Much Trading Volume is Too Much?. SSRN Electronic
Journal . [Crossref]
34. Wolfgang Drobetz, Dirk C. Schilling, Henning Schröder. 2015. Heterogeneity in the Speed of Capital
Structure Adjustment across Countries and over the Business Cycle. European Financial Management
21:5, 936-973. [Crossref]
35. Jarko Fidrmuc, Philipp Schreiber, Martin Siddiqui. 2015. The Transmission of Bank Funding to
Corporate Loans: Deleveraging in Germany. Open Economies Review 26:3, 581-597. [Crossref]
36. David P. McCaffrey. 2015. Private and Public Controls in the Over-the-Counter Derivatives Market,
1984-2015. SSRN Electronic Journal . [Crossref]
37. Yang Sun. 2014. The Effect of Index Fund Competition on Money Management Fees. SSRN
Electronic Journal . [Crossref]
38. Kei Kawakami. 2014. Excessive Dynamic Trading: Propagation of Belief Shocks in Small Markets.
SSRN Electronic Journal . [Crossref]
39. Rodney Maddock. 2013. Is the Australian Financial Sector Too Big?. SSRN Electronic Journal .
[Crossref]
40. Arvid O. I. Hoffmann, Thomas Post. 2013. How Does Investor Confidence Lead to Trading? Theory
and Evidence on the Links between Investor Return Experiences, Confidence, and Investment Beliefs.
SSRN Electronic Journal . [Crossref]
41. Austin Gerig, Daniel Fricke. 2013. Liquidity Risk, Speculative Trade, and the Optimal Latency of
Financial Markets. SSRN Electronic Journal . [Crossref]
42. Jean-Edouard Colliard. 2012. Catching Falling Knives: Speculating on Market Overreaction. SSRN
Electronic Journal . [Crossref]