"2-:)67-8=3*)227=0:%2-%"2-:)67-8=3*)227=0:%2-%
',30%60=311327 ',30%60=311327
-2%2')%4)67 #,%6832%'908=)7)%6',
-59-(-8=-7/%2(<4)'8)( 83'/)89627-59-(-8=-7/%2(<4)'8)( 83'/)89627
Ľ9&3šA7836
3&)68 8%1&%9+,
"2-:)67-8=3*)227=0:%2-%
3003;8,-7%2(%((-8-32%0;36/7%8,88476)437-836=94)22)(9*2')$4%4)67
%683*8,)-2%2')%2(-2%2'-%0%2%+)1)28311327
)'311)2()(-8%8-32)'311)2()(-8%8-32
A7836Ľ 8%1&%9+,-59-(-8=-7/%2(<4)'8)( 83'/)89627
3962%03*30-8-'%0
'3231=
,884(<(3-36+
!,-74%4)6-74378)(%8 ',30%60=311327,88476)437-836=94)22)(9*2')$4%4)67
36136)-2*361%8-3240)%7)'328%'86)437-836=43&3<94)22)(9
-59-(-8=-7/%2(<4)'8)( 83'/)89627-59-(-8=-7/%2(<4)'8)( 83'/)89627
&786%'8&786%'8
!,-7789(=-2:)78-+%8)7;,)8,)61%6/)8;-()0-59-(-8=-7%78%8):%6-%&0)-14368%28*36%77)846-'-2+#)
?2(8,%8)<4)'8)(783'/6)89627%6)6)0%8)('6377‐7)'8-32%00=838,)7)27-8-:-8-)73*6)8962783@9'89%8-327
-2%++6)+%8)0-59-(-8=961328,0=0-59-(-8=1)%796)%2%:)6%+)3*-2(-:-(9%0‐783'/1)%796)7)78-1%8)(
;-8,(%-0=(%8%6)0-)7328,)46-2'-40)8,%836()6@3;-2(9')7+6)%8)66)89626):)67%07;,)20-59-(-8=-7
03;)66318,639+,8,)%:)6%+)6)896232783'/7;-8,,-+,7)27-8-:-8-)7830-59-(-8=)<'))(7
8,%8*36783'/7;-8,03;7)27-8-:-8-)7&=4)6')28%229%00=%(.978)(*36)<43796)7838,)1%6/)86)8962
%7;)00%77->):%09)%2(131)2891*%'8367968,)6136)%0-59-(-8=6-7/*%'836%''39287*36,%0*3*8,)
463?8783%131)2891786%8)+=3:)68,)7%1)‐=)%64)6-3(
-7'-40-2)7-7'-40-2)7
-2%2')%2(-2%2'-%0%2%+)1)28
!,-7.3962%0%68-'0)-7%:%-0%&0)%8 ',30%60=311327,88476)437-836=94)22)(9*2')$4%4)67
642
[Journal of Political Economy, 2003, vol. 111, no. 3]
䉷 2003 by The University of Chicago. All rights reserved. 0022-3808/2003/11103-0006$10.00
Liquidity Risk and Expected Stock Returns
L
ˇ
ubosˇPa´stor
University of Chicago, National Bureau of Economic Research, and Centre for Economic Policy
Research
Robert F. Stambaugh
University of Pennsylvania and National Bureau of Economic Research
This study investigates whether marketwide liquidity is a state variable
important for asset pricing. We find that expected stock returns are
related cross-sectionally to the sensitivities of returns to fluctuations
in aggregate liquidity. Our monthly liquidity measure, an average of
individual-stock measures estimated with daily data, relies on the prin-
ciple that order flow induces greater return reversals when liquidity
is lower. From 1966 through 1999, the average return on stocks with
high sensitivities to liquidity exceeds that for stocks with low sensitiv-
ities by 7.5 percent annually, adjusted for exposures to the market
return as well as size, value, and momentum factors. Furthermore, a
liquidity risk factor accounts for half of the profits to a momentum
strategy over the same 34-year period.
Research support from the Center for Research in Security Prices and the James S.
Kemper Faculty Research Fund at the Graduate School of Business, University of Chicago,
is gratefully acknowledged (Pa´stor). We are grateful for comments from Nick Barberis,
John Campbell, Tarun Chordia, John Cochrane (the editor), George Constantinides, Doug
Diamond, Andrea Eisfeldt, Gene Fama, Simon Gervais, David Goldreich, Gur Huberman,
Michael Johannes, Owen Lamont, Andrew Metrick, Mark Ready, Hans Stoll, Dick Thaler,
Rob Vishny, Tuomo Vuolteenaho, Jiang Wang, and two anonymous referees, as well as
workshop participants at Columbia University, Harvard University, New York University,
Stanford University, University of Arizona, University of California at Berkeley, University
of Chicago, University of Florida, University of Pennsylvania, Washington University, the
Review of Financial Studies Conference on Investments in Imper fect Capital Markets at
Northwestern University, the Fall 2001 NBER Asset Pricing meeting, and the 2002 Western
Finance Association meetings.
liquidity risk 643
I. Introduction
In standard asset pricing theory, expected stock returns are related cross-
sectionally to returns’ sensitivities to state variables with pervasive effects
on investors’ overall welfare. A security whose lowest returns tend to
accompany unfavorable shifts in that welfare must offer additional com-
pensation to investors for holding the security. Liquidity appears to be
a good candidate for a priced state variable. It is often viewed as an
important feature of the investment environment and macroeconomy,
and recent studies find that fluctuations in various measures of liquidity
are correlated across assets.
1
This empirical study investigates whether
marketwide liquidity is indeed priced. That is, we ask whether cross-
sectional differences in expected stock returns are related to the sen-
sitivities of returns to fluctuations in aggregate liquidity.
It seems reasonable that many investors might require higher ex-
pected returns on assets whose returns have higher sensitivities to ag-
gregate liquidity. Consider, for example, any investor who employs some
form of leverage and faces a margin or solvency constraint, in that if
his overall wealth drops sufficiently, he must liquidate some assets to
raise cash. If he holds assets with higher sensitivities to liquidity, then
such liquidations are more likely to occur when liquidity is low, since
drops in his overall wealth are then more likely to accompany drops in
liquidity. Liquidation is costlier when liquidity is lower, and those greater
costs are especially unwelcome to an investor whose wealth has already
dropped and who thus has higher marginal utility of wealth. Unless the
investor expects higher returns from holding these assets, he would
prefer assets less likely to require liquidation when liquidity is low, even
if these assets are just as likely to require liquidation on average.
2
The well-known 1998 episode involving Long-Term Capital Manage-
ment (LTCM) seems an acute example of the liquidation scenario above.
1
Chordia, Roll, and Subrahmanyam (2000), Lo and Wang (2000), Hasbrouck and Seppi
(2001), and Huberman and Halka (2002) empirically analyze the systematic nature of
stock market liquidity. Chordia, Sarkar, and Subrahmanyam (2002) find that improvements
in stock market liquidity are associated with monetary expansions and that fluctuations
in liquidity are correlated across stocks and bond markets. Eisfeldt (2002) develops a
model in which endogenous fluctuations in liquidity are correlated with real fundamentals
such as productivity and investment.
2
This economic story has yet to be formally modeled, but recent literature presents
related models that lead to the same basic result. Lustig (2001) develops a model in which
solvency constraints give rise to a liquidity risk factor, in addition to aggregate consumption
risk, and equity’s sensitivity to the liquidity factor raises its equilibrium expected return.
Holmstro¨m and Tirole (2001) also develop a model in which a security’s expected return
is related to its covariance with aggregate liquidity. Unlike more standard models, their
model assumes risk-neutral consumers and is driven by liquidity demands at the corporate
level. Acharya and Pedersen (2002) develop a model in which each asset’s return is net
of a stochastic liquidity cost, and expected returns are related to return covariances with
the aggregate liquidity cost (as well as to three other covariances).
644 journal of political economy
The hedge fund was highly levered and by design had positive sensitivity
to marketwide liquidity, in that many of the fund’s spread positions,
established across a variety of countries and markets, went long less
liquid instruments and short more liquid instruments. When the Russian
debt crisis precipitated a widespread deterioration in liquidity, LTCM’s
liquidity-sensitive portfolio dropped sharply in value, triggering a need
to liquidate in order to meet margin calls. The anticipation of costly
liquidation in a low-liquidity environment then further eroded LTCM’s
position. (The liquidation was eventually overseen by a consortium of
14 institutions organized by the New York Federal Reserve.) Even though
exposure to liquidity risk ultimately spelled LTCM’s doom, the fund
performed quite well in the previous four years, and presumably its
managers perceived high expected returns on its liquidity-sensitive
positions.
3
Liquidity is a broad and elusive concept that generally denotes the
ability to trade large quantities quickly, at low cost, and without moving
the price. We focus on an aspect of liquidity associated with temporary
price fluctuations induced by order flow. Our monthly aggregate li-
quidity measure is a cross-sectional average of individual-stock liquidity
measures. Each stock’s liquidity in a given month, estimated using that
stock’s within-month daily returns and volume, represents the average
effect that a given volume on day d has on the return for day d ⫹ 1,
when the volume is given the same sign as the return on day d. The
basic idea is that, if signed volume is viewed roughly as “order flow,”
then lower liquidity is reflected in a greater tendency for order flow in
a given direction on day d to be followed by a price change in the
opposite direction on day Essentially, lower liquidity correspondsd ⫹ 1.
to stronger volume-related return reversals, and in this respect our li-
quidity measure follows the same line of reasoning as the model and
empirical evidence presented by Campbell, Grossman, and Wang
(1993). They find that returns accompanied by high volume tend to be
reversed more strongly, and they explain how this result is consistent
with a model in which some investors are compensated for accommo-
dating the liquidity demands of others.
We find that stocks’ “liquidity betas,” their sensitivities to innovations
in aggregate liquidity, play a significant role in asset pricing. Stocks with
higher liquidity betas exhibit higher expected returns. In particular,
between January 1966 and December 1999, a spread between the top
and bottom deciles of predicted liquidity betas produces an abnormal
return (“alpha”) of 7.5 percent per year with respect to a model that
accounts for sensitivities to four other factors: the market, size, and value
factors of Fama and French (1993) and a momentum factor. The alpha
3
See, e.g., Jorion (2000) and Lowenstein (2000) for accounts of the LTCM experience.