CampbellGrossmanWang93 TRADING VOLUME

TRADING VOLUME AND SERIAL CORRELATION IN
STOCK RETURNS*
This paper investigates the relationship between aggregate stock market

trading volume and the serial correlation of daily stock returns. For both stock
indexes and individual large stocks, the first-order daily return autocorrelation
tends to decline with volume. The paper explains this phenomenon using a model in
which risk-averse "market makers" accommodate buying or selling pressure from
"liquidity" or "noninformational" traders. Changing expected stock returns re-
ward market makers for playing this role. The model implies that a stock price
decline on a high-volume day is more likely than a stock price decline on a
low-volume day to be associated with an increase in the expected stock return.
There is now considerable evidence that the expected return

on the aggregate stock market varies through time. One interpreta-
tion of this fact is that it results from the interaction between
different groups of investors. Suppose that some investors,
"liquidity" or more generally "noninformational" traders, desire
to sell stock for exogenous reasons. Other investors are risk-averse
utility maximizers; they are willing to accommodate the selling
pressure, but they demand a reward in the form of a lower stock
price and a higher expected stock return. If these investors
accommodate the fluctuations in noninformational traders' de-
mand for stock, then they can be thought of as "market makers" in
the sense of Grossman and Miller [1988], even though they may
hold positions for relatively long periods of time and may not be
specialists on the exchange. l
It is hard to test this view of the stock market using data on
stock returns alone, because very different models can have similar
*We thank J. Harold Mulherin and Mason Gerety for providing data on daily
NYSE volume, G. William Schwert for providing daily stock return data, Martin
Lettau for correcting an error in the theoretical model, Ludger Hentschel for able
research assistance throughout this project, and four anonymous referees and
seminar articipants a t Princeton University and the 1991 NBER Summer
Institute for helpful comments. Campbell acknowledges financial support from the
National Science Foundation and the Sloan Foundation. Wang acknowledges
support from the Nanyang Technological University Career Development Assistant
Professorship a t the Sloan School of Management.
1. See also Campbell and Kyle 119931; De Long, Shleifer, Summers, and
Waldmann 11989, 19901;Shiller 119841;Wang [1993a, 1993131;and others.
o 1993 by the President and Fellows of Harvard College and the Massachusetts Institute of
Technology.
The Quarterly Journal of Economics, November 1993
906 QUARTERLY JOURNAL OF ECONOMICS
implications for the time-series behavior of returns. In this paper

we use data on stock market trading volume to help solve this
identification problem. The simple intuition underlying our work is
as follows. Suppose that one observes a fall in stock prices. This
could be due to public information that has caused all investors to
reduce their valuation of the stock market, or it could be due to
exogenous selling pressure by noninformational traders. In the
former case, there is no reason why the expected return on the
stock market should have changed. In the latter case, market
makers buying stock will require a higher expected return, so there
will tend to be price increases on subsequent days. The two cases
can be distinguished by looking at trading volume. If public
information has arrived, there is no reason to expect a high volume
of trade, whereas selling pressure by noninformational traders
must reveal itself in unusual volume. Thus, the model with
heterogeneous investors suggests that price changes accompanied
by high volume will tend to be reversed; this will be less true of
price changes on days with low volume.
Shifts in the demand for stock by noninformational traders
can occur at low frequencies or at high frequencies. Daily trading
volume is a signal for high frequency shifts in demand. Changes in
demand that occur slowly through time are harder to detect using
volume data because there are trends in volume associated with
other phenomena such as the deregulation of commissions and the
growth of institutional trading. We therefore focus on daily trading
volume and the serial correlation of daily returns on stock indexes
and individual stocks. Daily index autocorrelations are predomi-
nantly positive [Conrad and Kaul, 1988; Lo and MacKinlay, 19881,
but our theory predicts that they will be less positive on high-
volume days.
The literature on stock market trading volume is extensive,
but is mostly concerned with the relationship between volume and
the volatility of stock returns. Numerous papers have documented
the fact that high stock market volume is associated with volatile
returns; October 19, 1987, is only the best known example of a
pervasive p h e n ~ m e n o nIt
. ~has also been noted that volume tends
to be higher when stock prices are increasing than when prices are
falling.
2. See Gallant, Rossi, and Tauchen [1992];Harris [19871; Jain and Joh [19881;
Jones, Kaul, and Lipson [19911; Mulherin and Gerety [19891; Tauchen and Pitts
[1983]; and the survey in Karpoff [19871. Larnoureux and Lastrapes [I9901 ar e
that serial correlation in volume accounts for the serial correlation in volaticy
which is often described using ARCH models.
VOLUME AND SERIAL CORRELATION IN STOCK RETURNS 907
In contrast, there is almost no work relating the serial

correlation of stock returns to the level of volume. One exception is
Morse [1980], who studies the serial correlation of returns in
high-volume periods for 50 individual securities. He finds that
high-volume periods tend to have positively autocorrelated re-
turns, but he does not compare high-volume with low-volume
period^.^ Several recent papers study serial correlation in relation
to volatility: LeBaron [1992al and Sentana and Wadhwani [19921,
for example, show that the autocorrelations of daily stock returns
change with the variance of returns. Below, we compare the effects
of volume and volatility on stock return autocorrelations.
The organization of our paper is as follows. In Section I1 we
conduct a preliminary exploration of the relation between volume,
volatility, and the serial correlation of stock returns. In Section I11
we present a theoretical model of stock returns and trading
volume. In Section IV we show that the model can generate
autocorrelation patterns similar to those found in the actual data.
We use both approximate analytical methods and numerical simu-
lation methods to make this point. Section V concludes.
A. Measurement Issues
The main return series used in this paper is the daily return on
a value-weighted index of stocks traded on the New York Stock
Exchange and American Stock Exchange, measured by the Center
for Research in Security Prices (CRSP) at the University of
Chicago over the period 7/3/62 through 12/30/88. Results with
daily data over this period are likely to be dominated by a few
observations around the stock market crash of October 19, 1987.
For this reason, the main sample period we use in this paper is a
shorter period running from 7/3/62 through 9130187 (1962-1987
for short, or sample A). We break this period into two subsamples:
7/3/62 through 12/31/74 (1962-1974 for short, or sample B),
which is the first half of the shorter sample and which excludes the
3. Some other papers have come to our attention since the first draft of this
paper was written. Duffee [I9921 studies the relation between serial correlation and
tradingvolume in aggregate monthly data, while LeBaron [1992bl uses nonparamet-
ric methods to characterize the aggregate daily relation more accurately. Conrad,
Hameed, and Niden [I9921 study the relation between individual stocks' return
autocorrelations and the tradingvolume in those stocks.
period of flexible commissions on the New York Stock Exchange;

and 1/2/75 through 9130187 (1975-1987 for short, or sample C),
which is the remainder of the shorter sample period. Finally, we
use the complete data set through the end of 1988 in order to see
whether the extreme movements of price and volume in late 1987
strengthen or weaken the results we obtain in our other samples.
We call this long sample 1962-1988 for short, or sample D.
We also study the behavior of some other stock return series.
For the period before CRSP daily data begin, Schwert [I9901 has
constructed daily returns on an index comparable to the Standard
and Poors 500. We use this series over the period 112126-6/29/62.
The behavior of large stocks is of particular interest, since mea-
sured returns on these stocks are unlikely to be affected by
nonsynchronous trading. The Dow Jones Industrial Average is the
best known large stock price index, and so we study its changes
over the period 1962-1988. Individual stock returns also provide
useful evidence robust to nonosynchronous trading, so we study
the returns on 32 large stocks that were traded throughout the
1962-1988 period and were among the 100 largest stocks on both
7/2/62 and 12/30/88.4
Stock market trading volume data were kindly provided to us
by J. Harold Mulherin and Mason S. Gerety. These researchers
collected data from The Wall Street Journal and Barron's on the
number of shares traded daily on the New York Stock Exchange
from 1900 through 1988. They also collected data on the number of
shares outstanding on the New York Stock Exchange. For a
detailed description of their data, see Mulherin and Gerety [19891.
The ratio of the number of shares traded to the number of
shares outstanding is known as turnover, or sometimes as relative
volume. Turnover is used as the volume measure in most previous
studies (for example, Jain and Joh [I9881 and Mulherin and Gerety
[19891). Since the number of shares outstanding and the number of
shares traded have both grown steadily over time, the use of
turnover helps to reduce the low-frequency variation in the series.
It does not eliminate it completely, however, as can be seen from
the plot of the series presented in Figure I. Turnover has an
upward trend in the late 1960s and in the period between the
4. The 32 stocks are American Home Products, AT&T, Amoco, Caterpillar,

Chevron, Coca Cola, Commonwealth Edison, Dow Chemical, Du Pont, Eastman
Kodak, Exxon, Ford, GTE, General Electric, General Motors, ITT, Imperial Oil,
IBM, Merck, 3M, Mobil, Pacific Gas and Electric, Pfizer, Procter and Gamble, RJR
Nabisco, Royal Dutch Petroleum, SCE, Sears Roebuck, Southern, Texaco, USX, and
Westinghouse.
R a w Turnover
Date
FIGUREI
Level of Stock Market Turnover, 1960-1988
elimination of fixed commissions in 1975 and the stock market

crash of 1987. The growth of turnover in the 1980s may be due in
part to technological innovations that have lowered transactions
costs. In addition, the variance of turnover seems to increase with
its level during the 1980s.
In our empirical work, we want to work with stationary time
series. When we relate our empirical results to our theoretical
model, we want to measure trading volume relative to the capacity
of the market to absorb volume. For both these reasons we wish to
remove the low-frequency variations from the level and variance of
the turnover series. To remove low-frequency variations from the
variance, we measure turnover in logs rather than in absolute
units. To detrend the log turnover series, we subtract a one-year
backward moving average of log turnover. This gives a triangular
moving average of turnover growth rates, similar to the geometri-
cally declining average of turnover growth rates used by Schwert
[I9891 to explain stock return volatility. We explore some alterna-
tive detrending procedures below.
Our detrended volume measure is plotted in Figure 11. The
figure shows no trends in mean or variance, but it does show
considerable persistence. The first daily autocorrelation of de-
trended volume is about 0.7, and the fifth daily autocorrelation is
D e v i a t i o n s f r o m 1 y r MA of Log T u r n o v e r
.--.
'4
I
Date
FIGUREI1
Detrended Log Turnover, 1960-1988
still about 0.5. The standard deviation of the series is close to 0.25;
all these moments are stable across subsamples.
We also need a measure of stock return volatility. We take the
conditional variance series estimated by Campbell and Hentschel
[I9921 using daily return data over the period 1926-1988. Camp-
bell and Hentschel used a quadratic generalized autoregressive
heteroskedasticity (QGARCH) model with one autoregressive term,
two moving average terms, and a mean return assumed to change
in proportion to volatility. The QGARCH model is very similar to
the standard GARCH model of changing volatility, but it allows
negative returns to increase volatility more than positive returns
do.
B. Forecasting Returns from Lagged Returns, Volatility, a n d
Volume
Table I summarizes the evidence on the first daily autocorrela-
tion of the value-weighted index return. For each of our four
sample periods, the table reports the autocorrelation with a
heteroskedasticity-consistent standard error, and the R2 statistic
for a regression of the one-day-ahead return on a constant and the
return. This statistic, which we write as R2(1) in the table, is just
the square of the autocorrelation. The remarkable fact is that the
TABLE I
THE FIRSTAUTOCORRELATION OF STOCKRETURNS
(1) rt+l = cu + Prt
(2) rt+l = a + (Z:=l P,Di)rt
Sample period
autocorrelation exceeds 0.15 in every sample period; it is about 0.2

over the full sample and nearly 0.3 in the 1962-1974 period.
Table I also shows the improvement in R 2 that can be obtained
by allowing the first autocorrelation to vary with the day of the
week. A regression of the one-day-ahead return on the current
return interacted with five day-of-the-week dummies has an R 2
statistic, labeled R 2 ( 2 )in the table, that is at least 0.5 percentage
points larger than the R 2 of the basic regression. The increase in R 2
is even greater in the 1962-1988 period, but much of this is due to
the single week of the stock market crash. The day-of-the-week
dummies are significant enough that we include them in all our
subsequent regressions.
Table I1 looks at the relationship between volume and the first
autocorrelation of the value-weighted index return. We regress the
one-day-ahead stock return on the current stock return interacted
not only with day-of-the-week dummies but also with volume.
Alternatively, we interact the current return with dummies and
with estimated conditional variance. Finally, we report a regres-
sion in which the current return is interacted with dummies,
conditional variance, volume, and volume squared. The last of
these variables is included to capture any nonlinearity that may
exist in the relation between volume and autocorrelation.
Panel A of Table I1 uses the sample period 1962-1987. Over
this period Table I showed that 5.7 percent of the variance of the
one-day-ahead value-weighted index return can be explained by a
TABLE I1
VOLUME, VOLATILITY, AND THE FIRSTAUTOCORRELATION
r t + l = cu + ( I L 1 PiDi + ylVt + y z ~ f +
2 y3(1000uf2))rt
- -
Sample period Y1 Yz Y' 3

and specification (s.e.1 (s.e.1 (s.e.1 R2
A: 713162-9130187
Volume
Volatility
Volume and volatility
B: 713162-12131174
Volume
Volatility
C: 112175-9130187
Volume
Volatility
D: 713162-12130188
Volume
Volatility
regression on current return interacted with day-of-the-week

dummies. The first row of panel A shows that this R2 statistic can
be increased to 6.5 percentage points by interacting the regressor
with dummies and detrended trading volume. The coefficient on
the product of volume and the stock return is -0.33 with a
heteroskedasticity-consistent standard error of 0.06. This is eco-
nomically as well as statistically significant. The standard devia-
tion of detrended volume is about 0.25. Thus, as volume moves
from two standard deviations below the mean to two standard
deviations above, the first-order autocorrelation of the stock return
is reduced by about 0.3.
These strong results for volume are not matched by our

volatility measure. When volume is excluded from the regression,
volatility enters negatively, but it is statistically and economically
insignificant. When volume and volume squared appear in the
regression, volatility enters positively but is again insignificant.
The quadratic term on volume is positive and not quite significant
at the 5 percent level. Thus, in panel A there is only weak evidence
for any specification more complicated than the linear volume
regression reported in the second row.
In panels B and C of Table 11, we break the 1962-1987 period
into subsamples 1962-1974 and 1975-1987. The strongest results
come from the earlier subsample 1962-1974. In this period the
average first-order autocorrelation of the stock return is almost
0.3, and a regression of the one-day-ahead return on the current
return interacted with day-of-the-week dummies gives an R2
statistic of 8.4 percent. This can be increased by more than a
percentage point by taking account of a linear relationship between
the autocorrelation and trading volume. Once again volatility and
quadratic volume terms add little. In the later subsample, 1975-
1987, the first-order autocorrelation is much smaller on average.
Volume raises the regression R2 from 4.3 percent only to 4.6
percent, although the linear volume term is still statistically
significant with a t-statistic of 2.9. In this period there is a stronger
negative relationship between volatility and autocorrelation, al-
though volume is still slightly superior to volatility when both are
included in the regression.
Finally, in panel D of Table I1 we ask whether the addition of
the stock market crash period to the sample weakens or reinforces
our results. It turns out that the most recent data weaken the effect
of volume on the first-order autocorrelation of returns. Even in the
1962-1988 period, however, volume remains significant at the 5
percent level. We note also that the 1962-1988 period is the only
one for which day-of-the-week dummies make a major difference to
the results. When these dummies are excluded, the volume effect
becomes much stronger in the 1962-1988 period than in the
1962-1987 period. This is because the stock price reversals of the
week of October 19, 1987, are captured by day-of-the-week dum-
mies when these are included, or by volume when dummies are
omitted.
Table I11 has exactly the same structure as Table I, but now
the dependent variable is the two-day-ahead stock return so the
table describes the second-order autocorrelation of the return. The
9 14 QUARTERLY J O U R N U OF ECONOMICS
TABLE I11
THE SECOND
AUTOCORRELATION OF STOCKRETURNS
(1) rt+z = cu + prt
(2) rt+z = a + (Zf=l PiDi)rt
P
Sample period (s.e.1 R2(1) R2(2)
average second-order autocorrelation is small and statistically

insignificant in every sample period. Even when day-of-the-week
dummies are interacted with the current return, the R2 statistic of
the regression is less than 1.5 percent.
Table IV, which has the same structure as Table 11, shows
some evidence for volume effects on the second autocorrelation.
However, the evidence is much weaker than that for volume effects
on the first autocorrelation. Over the 1962-1987 sample period
(panel A) we find that volume enters the regression significantly
only when it is included in quadratic form. The linear coefficient is
-0.23 with a standard error of 0.08, while the quadratic coefficient
is 0.55 with a standard error of 0.15. These coefficients imply that
the second-order autocorrelation falls with volume until volume
reaches 0.2, about two-thirds of a standard deviation above its
mean. At higher levels of volume the positive quadratic term
dominates, and the autocorrelation starts to increase again. Look-
ing at subsamples in panels B and C, we find that the evidence for
volume effects on the second autocorrelation comes entirely from
the 1962-1974 period. Finally, in panel D we see that the addition
of the stock market crash period leads to stronger evidence for a
volume effect on the second autocorrelation.
One might ask whether higher-order autocorrelations also
change with trading volume. As a crude way to answer this
question without having to look at each autocorrelation individu-
VOLUME AND SERIAL CORRELATION I N STOCK RETURNS 915
TABLE IV
VOLUME, VOLATILITY, AND THE SECOND
AUTOCORRELATION
rt+z = a + (ZL1 PiDi + ylVt + yzV; + y3(1000u;))rt
Sample period YI Yz Y3
and specification (s.e.1 (s.e.1 (s.e.1 R2
A:713162-9130187
Volume -0.028
(0.060)
Volatility
Volume and volatility -0.233 0.550 0.071 0.008

(0.079) (0.146) (0.276)
B: 713162-12131174
Volume -0.186 0.011
(0.115)
Volatility -0.012 0.009
(0.338)
Volume and volatility -0.390 1.149 -0.039 0.021
(0.113) (0.345) (0.326)
C: 112175-9130187
Volume 0.074 0.003
(0.069)
Volatility 0.613 0.003
(0.391)
(0.109) (0.175) (0.400)
D: 713162-12130188
Volume -0.178 0.016
(0.089)
Volatility 0.007 0.011
(0.105)
Volume and volatility -0.024 -0.241 0.072 0.019
(0.087) (0.119) (0.102)
ally, we have run regressions of stock returns on moving averages

of past stock returns and on moving averages of past stock returns
interacted with trading volume. Regressions of this sort (where the
lags in the moving averages run from 1 to 5 or from 2 to 6) yield
results similar to those reported in Tables I1 and IV, with
somewhat reduced statistical significance. This suggests that the
main volume effects are in the first couple of autocorrelations, but
that there are at least no offsetting effects in higher autocorrela-
tions out to lags 5 or 6.
QUARTERLY JOURNAL O F ECONOMICS
TABLE V
VOLUME,VOLATILITY,AND THE FIRST AUTOCORRELATION:
ALTERNATIVE VOLUMEMEASURES
r t + ~= + (2:=1P,Di + y~Vt+ yzMAVt + ys(Vt + MAVt))rt
Sample period and YI Yz Y3

specification (s.e.1 (s.e.1 (s.e.1 R2
A: 713162-9130187
Detrended volume -0.328 0.065
(0.060)
Total volume -0.156 0.064
(0.028)
Detrended and -0.313 -0.090 0.066
trend volume (0.061) (0.037)
B:713162-12131174
(0.114)
(0.081)
Detrended and -0.417 0.292
trend volume (0.108) (0.141)
C: 112175-9130187
Detrended volume -0.214
(0.073)
(0.040)
Detrended and -0.218 -0.090
trend volume (0.073) (0.046)
D:713162-12130188
(0.080)
(0.050)
Detrended and -0.134 -0.065 0.064
trend volume (0.066) (0.059)
C. Alternative Volume and Volatility Measures

So far we have worked exclusively with detrended volume. It is
natural to ask whether similar results could be achieved without
detrending. To answer this, in Table V we run similar regressions
to those in Table 11, but using total volume instead of detrended
volume. We also run regressions including both the detrended
series and the trend. The general pattern is that detrended volume
has superior explanatory power to total volume, although this is
not true in 1975-1987. When both detrended and trend volume are
included, the coefficient on detrended volume is always negative

and significant, whereas the coefficient on the trend switches sign
from positive in 1962-1974 to negative in 1975-1987.
In an earlier version of this paper, we also used an unobserved
components model to stochastically detrend volume. The resulting
series was much less persistent than the detrended series used
here, having a positive first-order autocorrelation and than a series
of negative higher-order autocorrelations. The stochastically de-
trended volume series gave results similar to but systematically
weaker than those in Table II.5 This can be interpreted in terms of
our theoretical expectation that the serial correlation of stock
returns declines when volume increases relative to the ability of
market makers to absorb volume. A one-year backward moving
average of past volume, which reacts sluggishly to changes in
volume, seems to be a better measure of market making capacity
than an estimated random walk component from an unobserved
components model, which reads very quickly to changes in v01ume.~
To check the robustness of this finding, we have also tried
measuring volume as the deviations of log turnover from three-
month and five-year backward moving averages. Both these alter-
native moving average measures gave results similar to those
reported in Table 11. Thus, it seems to be important to measure
volume relative to a slowly adjusting trend, but the exact details of
trend construction are not crucial.
The results reported above also use a single measure of
volatility, the fitted value from a QGARCH model. The choice of
this particular model in the GARCH class is not critical since all
models in this class give very similar fitted variances. Nelson
[I9921 shows that high-frequency data can be used to estimate
variance very precisely, even when variance is changing through
time and the true model for variance is unknown.
It could be objected, however, that estimated conditional
volatility cannot compete equally with trading volume because
each day's conditional variance uses information only through the
5. LeBaron [I992131 builds on the work of this paper to explore the relation
between autocorrelation and high-frequency volume movements more thoroughly.
6. Grossman and Miller [I9881 model the long-run determination of market-
making capacity and show that in steady state the constant negative autocorrelation
of stock price changes is determined by the cost of maintaining a market presence
and the risk aversion of market makers. We believe that market-making capacity
adjusts slowly to the steady state, and this is consistent with our empirical results.
For simplicity, our theoretical model assumes that market-making capacity is fixed;
it is a short-run counterpart to Grossman and Miller's long-run model.
TABLE VI
THEFIRSTAUTOCORRELATION
OF STOCK RETURNS: ALTERNATIVE
SAMPLE
PERIODS
(1) rt+l = a + Prt
(2) rt+l = a + (Z:=, PiDiIrt
P
Sample period (s.e.1 R2(1) R2(2)
previous day. A simple way to respond to this is to add the current

squared return to the regression, since in any GARCH model the
squared return is the innovation in conditional variance. When we
do this, we find that the current squared return sometimes enters
significantly but does not have any important effect on the
estimated volume effect. To save space, we do not report results for
this specification.
D. Evidence from Earlier Periods
As a further check on the robustness of our results, in Tables
VI and VII we look at Schwert's [I9891 daily stock index
return over the period from 1926. The tables use five different
samples: the full pre-1962 data set 112126-6/29/62 (sample E);
decadal subsamples 112126-12130139, 1/2/40-12/31/49, and
113150-6/29/62 (samples F, G, and H, respectively); and a long
sample splicing together Schwert's series with the CRSP value-
weighted index over the period 112126-91 30187 (sample I).7
Table VI shows that the average first autocorrelation of stock
returns has varied considerably over the decades. In the 1930s it
was very small at 0.015, but it increased to above 0.1 in the 1940s
and 1950s. Table VII shows that the effect of volume on autocorre-
7. Sample I is long enough that adding the stock market crash period has
very little effect on the results, so to save space we do not report results for
112126-12130188.
TABLE VII
VOLUME, VOLATILITY,
AND THE FIRSTAUTOCORRELATION:

ALTERNATIVE SAMPLEPERIODS
rt+l = a + (CL, PiDi + ylVt + yt + y3(1000u:))rt
Sample period and 71 Yz '~3

specification be.) be.) be.) R2
E:112126-6129162
Volume
Volatility
F: 112126-12130139
Volume
Volatility
G:1/2140-12/31/49
Volume
Volatility
H: 113150-6129162
Volume
Volatility
I: 112126-9130187
Volume
Volatility
QUARTERLY JOURNAL OF ECONOMICS
TABLE VIII
THE FIRSTAUTOCORRELATION OF STOCKRETURNS:
THE DOWJONES INDUSTRIAL AVERAGE
( 1 )rt+l = a + Prt
( 2 )rt+l = a + (Zf=l PLDi)rt
P
Sample period be.) R2(1) R2(2)
A: 713162-9130187 0.141 0.020 0.027

(0.016)
lation has always been negative, although in many sample periods

it is statistically significant only when squared volume and volatil-
ity are also included in the r e g r e s s i ~ n Volatility
.~ is statistically
significant only in the period 1950-1962.
E. Nonsynchronous Trading
All the empirical results so far have used the return on a
value-weighted stock index. It could be objected that the serial
correlation of the index return is mismeasured because the indi-
vidual stocks in the index are not all traded exactly at the close. In
principle, nonsynchronous trading can lead to spurious positive
autocorrelation in an index return, although Lo and MacKinlay
[I9901 have shown that this effect is very small unless stocks fail to
trade for implausibly long periods of time.
As one way to respond to this objection, in Tables VIII and IX
we repeat our basic regressions using price changes of the Dow
Jones Industrial Average. Although this series omits dividends,
this has only a minimal effect on daily autocorrelations. Nonsyn-
chronous trading should also have only a trivial effect on the
behavior of the Dow Jones. The average autocorrelation of the Dow
Jones is much smaller than the average autocorrelation of the
8. An anonymous referee has objected that we report results with squared
volume only because they are statistically significant. We note, however, that we
reported the same squared volume regressions in the first version of this paper
which did not look at the older data.
TABLE IX
VOLUME, VOLATILITY, AND THE FIRSTAUTOCORRELATION:
THE DOWJONES INDUSTRIAL AVERAGE
rt+l = a + (Zf=, PiDi + ylVt + yzV; + y3(1000u;))rt
Sample period '~1 Yz Y3

and specification be.) be.) (s.e.1 R2
A: 713162-9130187
Volume -0.257 0.032
(0.061)
Volatility

(0.074) (0.147) (0.268)
B: 713162-12131174
Volume -0.359
(0.116)
Volatility -0.128 0.046
(0.304)
Volume -0.157 0.025

(0.073)
Volatility
Volume and volatility -0.177 0.138 - 1.036 0.027

(0.102) (0.174) (0.372)
D: 713162-12130188
Volume -0.127 0.037
(0.094)
Volatility

(0.159) (0.142) (0.132)
value-weighted portfolio, only one-half as large in some periods.

However, there is still a highly significant estimated effect of
volume on the autocorrelation.
Another way to respond to the nonsynchronous trading con-
cern is to use data on individual stock returns. Nonsynchronous
trading creates spurious positive autocorrelation in an index
return because today's market return is measured contemporane-
ously for those stocks that trade today, but only with a lag for
nontraded stocks. However, nonsynchronous trading has only a
trivial effect on measured individual stock return autocorrelations
[Lo and MacKinlay, 19901. Even when one uses individual stock
TABLE X
VOLUME
AND THE FIRSTAUTOCORRELATION:
INDMDUAL
STOCK
RETURNS
Equal-weighted index regression:
~ E W , ~ +=I LYEw + (C;=lP~w,iDi+ Y E W ~ ~ ) ~ E W , L
Pooled regression:
rj,t+l + clp + (C5=1 Pp,iDi + ypVt)rjjt, j = 1, . . . , 32
Individual stock regressions:
rj,t+1= aj + (CL1 Pj,iDi + yjVt)rj,t, j = 1, . . . , 32
7 = (1132) C g l yj, f., = (1132) ZE1t,j
-
YEW YP Y a,
Sample period be.) (s.e.1 (# < 0) (# < -1.64)
returns, aggregate volume is probably a better variable than

individual volume because idiosyncratic buying or selling pressure
does not create systematic risk for market makers. Accordingly, we
combine individual stock returns with the single aggregate volume
series.
Table X summarizes results for 32 large stocks that were
traded throughout the period 1962-1988 and were among the 100
largest at both the beginning and end of the period. The first
column of the table reports the volume effect on the autocorrela-
tion of an equally weighted index of these stocks. The index is very
similar to the Dow Jones, having an almost identical first autocor-
relation and a correlation with the Dow Jones of about 0.95. Not
surprisingly, therefore, the effect of volume on the index autocorre-
lation is similar to the effect reported in Table IX. The second
column of Table X shows the volume effect on the correlation of
each stock return with its own first lag, where the individual
returns are stacked together in a single pooled regression. The
standard error is corrected for heteroskedasticity and for the
contemporaneous correlation of individual stock returns, using the
method of White [1984].The volume effect on the own autocorrela-
tion of each stock return is smaller than the volume effect on the
index return, but the statistical significance of the effect is not
much reduced.
The third column of Table X shows the average effect of
volume on the first autocorrelation across 32 separate OLS regres-
sions, one for each individual stock. Not surprisingly the cross-
sectional average effect is close to the effect in the pooled regres-
sion. The number of negative individual coefficients is also reported;
at least 25 of these coefficients are negative in every sample period.
Finally, the fourth column of Table X shows the cross-sectional
average t-statistic for the effect of volume on the autocorrelation,
and the number of individual t-statistics that are less than - 1.64
(the 5 percent level for a one-tailed test, or the 10 percent level for a
two-tailed test). The cross-sectional average t-statistic is less than
- 1in every period except 1975-1987, and as many as a third of the
individual t-statistics are less than - 1.64.
We interpret these results as strong evidence that nonsynchro-
nous trading is not solely responsible for the phenomena we have
described.
In this section we present a model of noninformational trading

that can account for the empirical relationship between trading
volume and the serial correlation of stock returns. Several authors,
including Campbell and Kyle [19931, De Long, Shleifer, Summers,
and Waldmann [1989, 19901, Grossman and Miller [19881, and
Shiller [19841, have developed models in which expected stock
returns vary through time as some investors accommodate the
shifting stock demands of other investors. But none of these
authors explicitly work out the implications of their models for
trading volume.
Most previous work has modeled noninformational trading as
an exogenous process. De Long, Shleifer, Summers, and Waldmann
[1989,1990] derive noninformational trading from shifting misper-
ceptions of future stock payoffs. Here we derive noninformational
trading from shifts in the risk aversion of some traders. We do this
because we find it natural to relate changing demands to chang-
ing taste^,^ but the basic intuition of our model carries through
9. For simplicity, we treat shifts in investors' risk aversion as exogenous. More

generally, investors' attitudes toward risk may depend on wealth and other state
variables. This can lead them to follow dynamic hedging strategies even when they
face a constant investment opportunity set [Grossman and Zhou, 19921.
regardless of how noninformational trading is introduced. Any

such trading will give the same qualitative relation between
volume and the serial correlation of returns.
We consider an economy in which there exist two assets: a
risk-free asset and a risky asset ("stock"). We assume that
innovations in the stock price are driven by three random vari-
ables: (i) the innovation to the current dividend, (ii) the innovation
to information about future dividends, and (iii) the innovation to
the time-varying risk aversion of a subset of investors. Shock (i)
causes the payoff to the stock to be stochastic so that a premium is
demanded by investors for holding it. Shock (iii) generates changes
in the market's aggregate risk aversion, which cause the expected
return on the stock to vary. Shock (ii) is in the model so that prices
and dividends do not fully reveal the state of the economy and
volume provides additional information.
The properties of our model can be understood as follows. If a
large subset of investors becomes more risk averse, and the rest of
the economy does not change its attitudes toward risk, then the
marginal investor is more risk averse, and in equilibrium, the
expected return from holding the stock must rise to compensate
the marginal investor for bearing the risk. Simultaneously, risk is
reallocated from those people who become more risk averse to the
rest of the market. The reallocation is observed as a rise in trading
volume. Note that the rise in expected future returns is brought
about by a fall in the current stock price that causes a negative
current return. Therefore, a large trading volume will be associated
with a relatively large negative autocorrelation of returns.
A. The Economy
Our model further specifies the economy as follows. The
risk-free asset, which is in elastic supply, guarantees a rate of
+
return R = 1 r with r > 0. We assume that there is a fixed supply
of stock shares per capita, which is normalized to 1. Shares are
traded in a competitive market. Each share pays a dividend in
period t of D, = D + D,. D > 0 is the mean dividend, while D, is
the zero-mean stochastic component of the dividend. (We use
similar notational conventions for other variables below.) D,
follows the process:
We assume that the innovation ug,, is i.i.d. with normal distribu-

-
tion ug,, M(O, a:).
There are two types of investors in the economy, type A and

type B. Both types of investors have constant absolute risk
aversion. The type A investors' risk aversion parameter is a
constant a , while the type B investors' risk aversion parameter is
b,, which may change over time. Let w be the fraction of type A
investors.
Each period, investors solve the following problem:
subject to
where W, is wealth, X,is the holding of the risky asset, and Pt is the
ex dividend share price of the stock, all measured at time t. E, is the
expectation operator conditioned on investors' information set Yt
at time t.
The set Y ,contains the stock price P, and the dividend D,. It
also contains a signal, St,which all investors receive at time t about
the future dividend shock UD,,+I:
For simplicity, we assume that Stand E g , t + ~are jointly i.i.d. normal,

- -
E [ u , , + ,I St] = St, ED,, M(O,u?),and St M(O,ui).
B. The Equilibrium Price of the Risky Asset
Let Ft(the "cum-dividend fundamental value" of the stock) be
the present value of the expected future cash flow from a share of
the stock, including today's dividend, discounted at the risk-free
rate. It is easy to show that
and that the innovation variance of F,, a;, is given by
In the case that investors are risk neutral, Ft - Dt gives the

equilibrium ex-dividend price of the stock. When investors are risk
averse, however, the equilibrium price will depend on the risk
aversion of the market.
Define a variable 2,that can be interpreted as the risk aversion
of the marginal investor in the market:

abt
z, = (1- o)a + ob, '
Let Z, = Z + z,. We assume that Z,follows an AR(1) process:
We also assume that uz,,is independent of other shocks and is i.i.d.

-
normal: uz,, ~ ( 0 , u g )This
. assumption allows Z,, and thus b,, to
be negative. This could be avoided, however, if we replaced the
exponential utility assumption (1)by the assumption that inves-
tors' have mean-variance preferences; that is, they maximize the
objective function E,Wt+l - ~ a r , W , + ~ lAll
2 . the results in the
paper would follow, and we could restrict the Z, process to be
bounded away from zero.
Finally, we assume that ug I ug2 = (R - CQ)~/~U;. This
assumption is used to derive an equilibrium price function where
the price of the stock is a decreasing function of the aggregate risk
aversion 2,.
THEOREM1. For the economy defined above, there exists an
equilibrium price of the stock that has the followingform:
(7) Pt = Ft - Dt + (Po + PzZt),
where pz = -((R - aZ)l2$)[l - and PO =

(1 - c ~ ~ )< ~0.
~ / r
Proof of Theorem 1. See Appendix A.
C. Excess Stock Returns and Trading Volume

The excess return per share on the stock realized at time t + 1
is written as = P,+ + D,+l - RP,. Given the equilibrium price,
the expected excess return anticipated by investors in period t,
denoted bye,, is
(8) e, = EIQ,+lJY,l= u$Zt, u2Q - UF + pgui,
where u$ = var [Q,+ I Y ,I. Then, we have
where K = p Z l u ~Equation
. (8) states that the unexpected excess
stock return per share has two components: innovations in ex-
pected excess returns per share and innovations in expected future
cash flows per share. Given the return process (9), the serial
correlation in returns can be easily calculated:
is positive if az > 1/R and is negative if az < 1/R.

Clearly, PQ,,Q,+,
Let Xp and Xt be, respectively, the optimal stock holdings of type A
and type B investors. The solution of the optimization problem (1)
yields
E[Qt+lIPt,Dt,Stl 1
= - Zt,
Xp = a var [Qt+ 1 Pt,Dt,StI a
(11)
E[Qt+lIPt,Dt,Stl 1
= - 2,.
X: = 6, var [Qt+lIPt,Dt,StI bt
Changes in investors' preferences relative to one another
generate trading. Xp (and x:) change as Zt changes:
Trading volume is then
Given the Zt process, mean trading volume is = EIVtl = (2ouz)I

(ad m . Equation (13) completes the solution of the model for
the joint behavior of volume and stock returns.
IV. IMPLICATIONS
OF THE MODELFOR VOLUME
AND SERIAL
CORRELATION
Investors in the economy have perfect information about the
current level of 2,. They can use Zt to predict future excess returns
as shown by equation (7).When Zt is high, the type B investors are
highly risk averse and less willing to hold the stock. The price of the
stock has to adjust to increase the expected future excess return so
that the type A investors are induced to hold more of the stock.
We, as econometricians, do not directly observe Zt or St. We
observe only realized excess returns and trading volume.1° How-
ever, these variables do provide some information about the
10. We could actually use a finer information set containing dividends, prices,
and volume. This would improve our inferences about 2,.For simplicity, however,
we use only excess returns and volume in this paper.
current level of Zt and can help predict future returns. A low return
due to a drop in the price could be caused either by an increase in Zt
or by a low realization of St,i.e., bad news about future cash flow.
However, changes in Zt will generate trading among investors,
while public news about future cash flows will not. Therefore, low
returns accompanied by high trading volume are more likely due to
increases in Zt while those accompanied by low trading volume are
more likely due to low realizations of St.In the case of an increase
in Z,, the expected excess return for next period will be high, while
for the case of low St,it will not. Thus, the autocorrelation of the
stock return should decline with trading volume.
A. Analytical Results
In this subsection we use analytical methods to develop this
intuition more formally. In the next subsection we use simulation
methods to a similar end.
We want to calculate the predictable component in the excess
return based on the current return and volume: EIQt+lI Qt,Vtl =
ui E [ZtI Qt,Vtl.The following theorem holds.
THEOREM 2. Under the assumptions we have made about the
structure of the economy and the distribution of shocks, we
have
A quadratic approximation to equation (14) is
(15) 1 Qtl = (+Q - O+vV?)&t,

where (O+v) > 0.
Proof of Theorem 2. See Appendix B.
In order to understand the results in Theorem 2, first consider
the case where volume Vt = 0. In this case, there is no change in
the investors' relative risk aversion (i.e., Zt has remained the
same). Hence, there should be no change in the expected excess
return from the previous period. The realized excess return approx-
imates the expected excess return in the previous period. Thus,
E[Qt+iI Qt,Vt = 01 = +Q Qt .
Now consider the case where volume is not zero. This implies
that a risk preference shock has occurred. Note that if Q, = 0 (i.e.,
there were no unusual date t returns), then E[Q,+~ I Qt= O,Vtl = 0,
independent of the value of V,. Although volume implies that 2, is

different from it does not reveal the direction of the change. If
Q, is negative, however, we can infer that 2, is more likely to have
increased than decreased, and thus the expected value of 2, is high.
Given a negative Q,, the higher is V,, the higher is the implied value
of 2,.
We can re-express equation (15) in a form that looks more
similar to the regression equations used in the previous section:
where is positive while the sign of 4, is ambiguous. (See

Appendix B. )
In Theorem 2 we only consider how current volume in addition
to the current return can help in predicting future returns. In
principle, we could use the whole history of returns and volume to
forecast future returns. Let Y/e:= {Q,,V,: T It ]be the information
set that contains the history of excess returns and volume up to
and including period t. The forecasting problem faced by an
econometrician is to calculate the conditional expectation:
E[Q,+ lY:l =@[Z, JYt*l.This is a nonlinear filtering problem for
which there is no simple solution. We could calculate the condi-
tional expectation iteratively: having calculated the expectation
conditional on the return and volume in the current period, we
could calculate the expectation conditional on the return and
volume in the current and the previous period, and so on [Wang,
1993al. This process would reveal higher-order dynamic relations
between return and volume, which could be related to empirical
work like that of Brock, Lakonishok, and LeBaron [19921. How-
ever, this is outside the scope of the current paper.
Theorem 2 provides some justification for the exploratory
regressions we reported in the empirical section of the paper. The
theorem states that aggregate risk aversion (and hence the ex-
pected stock return) is related to the lagged stock return and to the
lagged return interacted with volume. The coefficient on the
volume-weighted lagged return should be negative, as we found in
the data. Note that there is some slippage between the theoretical
variables in our model and the variables measured in our empirical
work. The model generates predictions about the level of turnover
and the serial correlation of returns per share, while our empirical
work concerns the detrended log of turnover and the serial
correlation of log returns per dollar invested.
B. Simulation Results
Although the analysis of the previous subsection makes our
basic point, that volume and serial correlation should be negatively
related in our heterogeneous-agent model, it is not clear whether
this effect is quantitatively important for plausible parameter
values. In this subsection we run some simple simulations to
address this question. The model of Section 111, with normal
driving processes, is straightforward to simulate because it is a
linear model conditional on investors' information. It only becomes
nonlinear when we condition on the smaller information set
containing volume and returns alone. The key question is how to
calibrate the parameters of the model.
We begin by describing the riskless and risky assets in the
economy. We set the riskless interest rate R equal to 1.01 at an
annual rate, or 1.00004 at a daily rate assuming that there are 250
trading days in a year. We set the autoregressive parameter for the
stock dividend, aD,equal to one. This makes the dividend a random
walk. In daily data any plausible dividend process will have CYDvery
close to one, and the model is simplified by setting it equal to one.
Next we normalize the stock price so that it equals one when all the
stochastic terms equal zero, and set stochastic terms to zero at the
beginning of our simulations. This normalization means that the
average stock price should not be too far from one during our
simulation periods, although the stock price process has a unit root
so there is no fixed mean. The normalization makes absolute price
variability close to percentage price variability, and it ensures that
the coefficient of absolute risk aversion and the coefficient of
relative risk aversion are similar if initial riskless asset holdings are
small.l1
The next step is to pick a plausible value for the innovation
variance u i of the stock's fundamental value F,. We choose ug =
(0.0U2,so that the standard deviation of the daily stock return (in
the absence of shifting risk aversion) is 1 percent. This is a little
higher than the average in postwar data. Equation (4) gives the
implications of this choice for the variances of the dividend signal
St and the contemporaneous dividend innovation E,. If there is no
dividend signal, then a : = 0, and the implied variance of the
dividend innovation E, is uf = (R - ~ ) ~ u i l R If ~all. dividend
11. Note that riskless asset holdings of the agents are not identified by the
model. With exponential utility, these holdings do not affect demand for the risky
asset.
information is received one day in advance, then at = 0, and a: =

(R - I ) ~ U ~The
. simulation results are only trivially affected by
varying the relative importance of the dividend signal and the
contemporaneous dividend innovation.
We now turn to the specification of the two groups of investors.
Suppose initially that both groups have the same average risk
aversion coefficient. Then this coefficient can be identified from the
stock demand equation (11).When all investors have constant risk
aversion a, then (11) implies that a = E [Q,+Jlvar If stock
price equals fundamental value, then var [Q,+J = a: = (0.01)2.
Setting a = 3 gives a reasonable value for E [Q,+J of 0.0003, or 7.5
percent at an annual rate. This procedure for estimating average
risk aversion is a variant of that proposed by Friend and Blume
[19751.
Next we consider o, the proportion of market-making agents.
Given the Z, process, this parameter plays two roles. First, in
equation (13) the trading volume generated by a given shift in Z, is
proportional to ola. Second, in equation ( 5 ) the mapping between
Z, and the risk aversion of liquidity traders b, is determined by o.
When o is small, Z, moves almost one-for-one with b,; when o is
large, on the other hand, large shifts in b, result in smaller changes
in 2,.
It turns out that if we set a = 3, then we must also pick a very
small value of o, 0.0005. Figure I shows that turnover is typically
0.5 percent or less. In exploratory simulations with larger values of
o , the variation in Z, required to explain the effect of volume on
autocorrelation generates too much trading volume when market-
makers have risk aversion a = 3. We can, however, increase the
fraction of market-makers if we also make market-makers more
risk averse, since volume is determined by the ratio ola. We obtain
almost identical simulation results, for example, if we set o = 0.005
and a = 30, while keeping the mean of 2, equal to 3 to match the
mean stock return.
The trickiest part of the calibration is to specify the dynamics
of the Z, process. We would like to pick a process that generates
realistic stock price behavior. Equation (6) gives the price innova-
tion variance aspgag + a;. Unfortunately, the coefficientpz is itself
a function of a; and the other parameters of the model. When a; =
0, however, Appendix A shows that pz = pz(0) = -uil(R - az).
The coefficient pz(0) is the value of pz that obtains when Z is
deterministic. As a simple way to calibrate the model, we define a
coefficient A equal to the standard deviation of price innovations
caused by randomness in Z divided by the standard deviation of

innovations in fundamental value, evaluated usingpz(0):
Solving this equation for a;, we find that

(18) ug = A2(R - ~ i ~ ) ~ / u i .
This equation can be substituted into the condition that ug Iuz2,
which guarantees a real solution for the coeffi~ientp~. We can then
restate that condition in the simple form A I 0.5. Thus, only
limited extra stock price variability can be generated by shifting
risk aversion.
In preliminary simulations, we varied A over the permissible
range from 0 to 0.5, while at the same time varying the persistence
parameter over its permissible range from 0 to 1. We found a
strong negative relation between trading volume and the first
return autocorrelation only for A values above about 0.2, and %
values below about 0.5. With smaller values of A, shifting risk
aversion did not have a sufficient effect on stock price behavior to
be readily detectable, even with very large numbers of observa-
tions. With larger values of az, price changes caused by changing
risk aversion are largely permanent so trading volume does not
strongly signal that price movements will be reversed. For the final
simulations reported below, we picked A = 0.25 and = 0, 0.25,
and 0.5.
Once we have chosen parameter values, we can solve for the
price coefficients p o and pz. The final step is to choose an initial
dividend Do to meet our requirement that the initial price equal
one. We then draw normal innovations with the appropriate
variances and create artificial data on stock prices and trading
volume. We create series that have 3000 observations (roughly the
number of observations in our 1962-1974 and 1975-1987 sub-
samples) after discarding the first 100 observations.
Illustrative simulation results are reported in Table XI. The
table shows regression results for a standard AR(1) return model
and for our model interacting the return with trading volume. All
parameters are fixed as described above, except for the parameter
% which describes the persistence of shifts in risk aversion. This
parameter is 0 in panel A of Table XI, 0.25 in panel B, and 0.5 in
panel C. In panel A we find a strong effect of volume on the first
autocorrelation of returns. The t-statistic on volume is 3.67, and
TABLE XI
SIMULATIONS
OF VOLUME AND THE FIRSTAUTOCORRELATION
rt+l = a + (P + ylVt)rt
Specification
Volume
B:a,= 0.25
AR(1)
Volume
C:a,= 0.5
AR(1) -0.005 0.000
(0.261)
Volume 0.024 - 14.07 0.001
(0.840) (1.26)
the addition of volume to the regression increases the R 2 statistic

by more than 50 percent (although of course the R remains very
low in absolute terms). The coefficient on volume is -16.2, while
the standard deviation of volume (not shown in the table) is 0.0036.
Thus, when volume moves from two standard deviations below the
mean to two standard deviations above, the autocorrelation of the
stock return falls by 0.23, an economically significant amount.
As the persistence of risk aversion increases, the relation
between volume and autocorrelation weakens. Results in panel B
are only slightly weaker than those in panel A, but in panel C the
coefficient on volume is statistically insignificant although the
point estimate is still negative. Even with % = 0.5, the half-life of a
shift in risk aversion is only one trading day, so it is clear that risk
aversion shifts must be highly transitory for our model to fit the
data.
A related problem for our model is that the parameter values
in Table XI imply extreme movements in average risk aversion 2,.
The simulation reported in panel A has a sample average for 2, of
3.42, close to the population value of 3. The sample standard
deviation is 25.2, with a minimum of -89 and a maximum of 88.
Given that market-makers are assumed to be a very small fraction
of the market, the implied movements of liquidity traders' risk

aversion b, are almost equal to those of 2,. (As noted above, a larger
value of o would imply larger movements in b, relative to Z,,
worsening this problem.) As % increases, the movements in Z, are
slightly dampened, but they remain extreme even when % = 0.5.
The sample average Z, in panel C is 2.80, with a standard deviation
of 14.4, a minimum of -46, and a maximum of 57.
This difficulty arises for the following reason. Persistent shifts
in Z, have large effects on prices, but as noted above, they do not
generate a strong high-frequency relationship between volume and
serial correlation. Volume interacted with the lagged stock return
helps to identify the recent change in the expected stock return;
but this is not a good guide to the current level of the expected stock
return when the expected return follows a persistent time series
process. Transitory shifts in Z,, on the other hand, have small
effects on prices because small temporary price movements can
create large temporary changes in expected returns. Equation (18)
shows that as the persistence parameter CYZ falls, Z, must become
more variable for any given price impact parameter A. Thus, to get
a strong effect of volume on serial correlation, we need very large
transitory shifts in risk aversion. This is an example of the
well-known fact that high-frequency predictability in asset returns
is hard to explain using a frictionless model with utility-
maximizing risk-averse agents. Our model has an advantage in
that it allows for heterogeneous and time-varying risk aversion,
but it does not entirely escape this problem.
Our model has another empirical difficulty related to persis-
tence. We have found that the autocorrelation of stock returns
depends on a detrended volume measure that is fairly persistent,
having a first daily autocorrelation of about 0.7 and a fifth daily
autocorrelation that still exceeds 0.5. When one extracts the
high-frequency component of volume by using an unobserved
components model or subtracting a few days' moving average of
volume [LeBaron, 1992131, the relation between volume and auto-
correlation becomes much weaker. This contradicts the implication
of our model that volume is an MA(1)process when market average
risk aversion Z, is white noise (and close to an MA(1) process when
risk aversion is a transitory AR(1) process). It should, however, be
possible to generalize the model to mitigate this problem. Since
volume depends on the absolute value of the change in market
average risk aversion Z,, a conditionally heteroskedastic process for
Z, could produce persistent movements in volume.
In this paper we have documented a striking fact about

short-run stock market behavior: the first daily autocorrelation of
stock returns is lower on high-volume days than on low-volume
days. This phenomenon appears even in very large stock indexes
and individual stock returns, so that it is unlikely to be due to
nonsynchronous stock trading. We have proposed an alternative
explanation relying on the idea that trading volume occurs when
random shifts in the stock demand of noninformational traders are
accommodated by risk-averse market-makers. If we allow large
transitory shifts in noninformational demand, then our model fits
many of the features of the data.
The proof of Theorem 1 follows a fairly standard pattern.

First, we conjecture that the equilibrium price function has the
given form. Second, we solve the optimization problem of both type
A and type B investors given the conjectured price function.
Finally, we impose the market-clearing condition to verify the
conjectured price function.
If the price function takes the conjectured form, the excess
return per share of the stock, denoted by &,+I = Pt+1+ Dt+1- RP,,
can be expressed as
The conditional distribution of the future excess return is normal

and has the following moments:
E[Qt+l141= -r(po + PZZ)+ PZ(% - R)z,,
var [Qt+llA]I u i = u i + pgui.
Given the price function, the solution to the optimization
problem (1) gives the optimal holdings of type A and type B
investors:
EL&,+,IPt,Dt,Stl
(A.2a)
X4 = a var [Q,+IIP~,D~,S~I
The market-clearing condition states that

(A.3) OX: + (1 - O)X; = 1.
Hence,
Since ola + (1- o)lb, = l/Zt,we have

(A.5) (OIZ- R)pz = a:, -r(po + pZZ) = u:Z.
Under the condition that a; Iuj2,we have two real roots forpz:
For Q < 1,both roots are negative. We choose the root that gives
the right limit when a: goes to zero. In the case that u -+ 0,pz :
should go to zero. This leads to the solution forpz which is the root
with the positive sign. po is then given bypo = (1 - %)pZElr.
Define At = (ola)(Z,- Zt-l).Thus, Vt = I A t 1. Also, define EF,,

= Ft - Et-l [Ftl.E F , ~gives the innovation process to F,. Then,
(A.7) &t+l = Pz(Zt+l- 2,) + E ~ , t + l .
Let 2 be the covariance matrix of (&,,A,).
LEMMA.Given that Q, +,, Q, and A t are jointly normal, we have
(~4.8) E [Q ,+ 11 Q,, Vtl = $g Q t - $v tanh (OV~Q

t ) ~
t ,
where
and
Proof of Lemma. See Wang [1993al.

It is easy to show that
--o (1 + R ) P Z U ~
uQt+l,At - a 1+ % a
Hence,
To a quadratic approximation, equation (A.8) can be re-ex-

pressed as
Clearly, O+v > 0. This completes the proof of Theorem 2. We can

further write Vt = V + Vt, where V = E[V,] is the mean volume. To
the same order of approximation, equation (A.9) becomes
(A.10) a&,+,
I Q,,V,I = (4, - +,Vt)&,,
where 4o = - (04v)pand = > 0.
QUARTERLY JOURNAL OF ECONOMICS
REFERENCES
Brock, William A,, Josef Lakonishok, and Blake LeBaron, "Simp,l,e Technical
Trading Rules and the Stochastic Properties of Stock Returns, Journal of
Finance, XLVII (19921,1731-64.
Campbell, John Y., and Ludger Hentschel, "No News Is Gvyd News: An Asymmet-
ric Model of Changing Volatility in Stock Returns, Journal of Financial
Economics, XXXI (19921,281-318,
Campbell, John Y., and Albert S. Kyle, "Smart Money, Noise Trading, and Stock
Price Behavior," Review of Economic Studies, LX (19931, 1-34.
Conrad, Jennifer, Allaudeen Hameed, and Cathy M. Niden, "Volume and Autoco-
variances in Short-Horizon Individual Security Returns," unpublished paper,
University
* nn"
of North Carolina at Chapel Hill and University of Notre Dame,
IYYL.
Conrad, Jennifer, and Gautam Kaul, "Time-Variation in Expected Returns,"
Journal of Business, LXI (19881,409-25.
De Long, J. Bradford, Andrei Shleifer, Lawrence H. Summers, and Robert J.
Waldmann, "The Size and Incidence of the Losses from Noise Trading,"
Journal ofFinance, XLIV (19891,681-96.
De Long, J. Bra;$ford, Andrei Shleifer, Lawrence H. Summers, and Robert J.
Waldmann, Noise Trader Risk in Financial Markets," Journal of Political
Economy, XCVIII (19901,703-38.
Duffee, Gregory, "Trading Volume and Return Reversals," Finance and Economics
Discussion Series No. 192, Board of Governors of the Federal Reserve System,
1992.
Friend.
- .-
. .. Irwin. and
., -- ---> Marshall
.- - .
- -- ~- - - - -
-- - E. Blume. "The Demand for Riskv Assets." American
- - ~ ~ - - -
Economic Review, LXV (19751,900-22.~

Gallant, A. Ronald, Peter E. Rossi, and George Tauchen, "Stock Prices and
Volume," Review ofFinancia1 Studies, V (19921, 199-242.
Grossman, Sanford J., and Merton H. Miller, "Liquidity- - and Market Structure,"
~ o u r n hofFinance,
l XLIII (19881,617-33.
Grossman, Sanford J., and Zhongquan Zhou, "Optimal Investment Strategies for
Controlling Drawdowns," unpublished paper, University of Pennsylvania,
1992.
Harris, Lawrence, "Transactions Data Tests of the Mixture of Distributions
Hypothesis," Journal of Financial and Quantitative Analysis, XXII (19871,
127-41.
Jain, Prem C., and Gun-Ho Joh, "The Dependence Between Hourly Prices and
Trading Volume," Journal of Financial and Quantitative Analysis, XXIII
(19881, 269-83.
Jones, Charles M., Gautam Kaul, and Marc L. Li son "Transactions, Volume, and
Volatility," unpublished paper, University o t ~ i c h i g a n1991.
,
Karpoff, Jonathan M., "The Relation Between Price Changes and Trading Volume:
A Survey," Journal of Financial and Quantitative Analysis, XXII (19871,
109-26.
Lamoureux, Christopher G., and William D. Lastrapes, "Heteroskedasticity in
Stock Return Data: Volume versus GARCH Effects," Journal of Finance, XLV
(19901, 221-29.
LeBaron, Blake, "Some Relations Between Volatility and Serial Correlation in
Stock Market Returns," Journal of Business, LXV (1992a1,199-219.
, "Persistence of the Dow Jones Index on Rising Volume," unpublished paper,
University of Wisconsin, 1992b.
Lo, Andrew W., and A. Craig MacKinlay, "Stock, Prices Do Not Follow Random
Walks: Evidence from a Simple Specification Test," Review of Financial
Studies, 1(1988),41-66.
Lo, Andrew W., and A. Craig MacKinlay, "An Econometric Analysis of Nonsynchro-
nous Trading," Journal of Econometrics, XLV (19901,181-211.
Morse, Dale, "Asymmetrical Information in Securities Markets and Trading
Volume," Journal of Financial and Quantitative Analysis, XV (19801, 1129-
46.
Mulherin J. Harold, and Mason S. Gerety, "Daily Trading Volume on the NYSE
During the Twentieth Century," unpublished paper, Clemson University,
1989.
Nelson, Daniel B., "Filtering and Forecasting with Misspecified ARCH Models I:
Getting the Right Variance with the Wrong Model," Journal of Econometrics,
LII (1992). 61-90.
Schwert, G. ~ i l l i a m ,"Why Does Stock Market Volatility Change Over Time?"
J t u r n a l of Finance, XLIV (1989),1115-53.
, Indexes of U. S. Stock Prices from 1802 to 1987," Journal of Business, LXIII
(1990), 399-426.
Sentana, Enrique, and Sushi1 Wadhwani, "Feedback Traders and Stock Return
Autocorrelations: Evidence from a Century of Daily Data," Economic Journal,
CII (1992), 415-25.
Shiller, Robert J . , "Stock Prices and Social Dynamics," Brookings Papers on
Economic Activity (19841,457-98.
Tauchen, George E., and Mark Pitts, "The Price Variability-Volume Relationship
on Speculative Markets," Econometrics, LI (1983), 485-505.
Wan ,Jiang, "A Model of Competitive Stock Trading Volume," unpublished paper,
h o a n School, Massachusetts Institute of Technology, 1993a.
Wang, Jiang, "A Model of Intertemporal Asset Prices Under Asymmetric
Information," Review of Economic Studies, LX (1993b1, forthcoming.
White, Halbert, Asymptotic Theory for Econometricians (Orlando, FL: Academic
Press, 1984).

CampbellGrossmanWang93 TRADING VOLUME

Uploaded by

Copyright:

Available Formats

CampbellGrossmanWang93 TRADING VOLUME

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

CampbellGrossmanWang93 TRADING VOLUME

Uploaded by

Copyright:

Available Formats

TRADING VOLUME AND SERIAL CORRELATION IN

This paper investigates the relationship between aggregate stock market

There is now considerable evidence that the expected return

implications for the time-series behavior of returns. In this paper

In contrast, there is almost no work relating the serial

period of flexible commissions on the New York Stock Exchange;

4. The 32 stocks are American Home Products, AT&T, Amoco, Caterpillar,

Level of Stock Market Turnover, 1960-1988

elimination of fixed commissions in 1975 and the stock market

Detrended Log Turnover, 1960-1988

autocorrelation exceeds 0.15 in every sample period; it is about 0.2

Sample period Y1 Yz Y' 3

Volume and volatility

Volume and volatility

Volume and volatility

Volume and volatility

regression on current return interacted with day-of-the-week

These strong results for volume are not matched by our

average second-order autocorrelation is small and statistically

Volume and volatility -0.233 0.550 0.071 0.008

ally, we have run regressions of stock returns on moving averages

Sample period and YI Yz Y3

C. Alternative Volume and Volatility Measures

included, the coefficient on detrended volume is always negative

previous day. A simple way to respond to this is to add the current

AND THE FIRSTAUTOCORRELATION:

Sample period and 71 Yz '~3

Volume and volatility

Volume and volatility

Volume and volatility

Volume and volatility

Volume and volatility

QUARTERLY JOURNAL OF ECONOMICS

A: 713162-9130187 0.141 0.020 0.027

lation has always been negative, although in many sample periods

Sample period '~1 Yz Y3

Volume and volatility -0.358 0.278 0.080 0.033

Volume -0.157 0.025

Volume and volatility -0.177 0.138 - 1.036 0.027

Volume and volatility -0.251 0.103 0.081 0.039

value-weighted portfolio, only one-half as large in some periods.

returns, aggregate volume is probably a better variable than

In this section we present a model of noninformational trading

9. For simplicity, we treat shifts in investors' risk aversion as exogenous. More

regardless of how noninformational trading is introduced. Any

We assume that the innovation ug,, is i.i.d. with normal distribu-

There are two types of investors in the economy, type A and

For simplicity, we assume that Stand E g , t + ~are jointly i.i.d. normal,

and that the innovation variance of F,, a;, is given by

In the case that investors are risk neutral, Ft - Dt gives the

of the marginal investor in the market:

Let Z, = Z + z,. We assume that Z,follows an AR(1) process:

We also assume that uz,,is independent of other shocks and is i.i.d.

where pz = -((R - aZ)l2$)[l - and PO =

C. Excess Stock Returns and Trading Volume

where u$ = var [Q,+ I Y ,I. Then, we have

is positive if az > 1/R and is negative if az < 1/R.

Trading volume is then

Given the Zt process, mean trading volume is = EIVtl = (2ouz)I

A quadratic approximation to equation (14) is