Description
The purpose of this paper is to examine changes in short-sale transactions of target firms
and acquiring firms around merger and acquisition (M&A) announcements using daily short-sale
transaction data from the New York stock exchange and NASDAQ. The paper further aims to
investigate the link between short-sale transactions and trading costs.
Journal of Financial Economic Policy
Short sales around M&A announcements
Liuqing Mai Robert van Ness Bonnie van Ness
Article information:
To cite this document:
Liuqing Mai Robert van Ness Bonnie van Ness, (2009),"Short sales around M&A announcements", J ournal
of Financial Economic Policy, Vol. 1 Iss 2 pp. 177 - 197
Permanent link to this document:
http://dx.doi.org/10.1108/17576380911010272
Downloaded on: 24 January 2016, At: 21:35 (PT)
References: this document contains references to 28 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 399 times since 2009*
Users who downloaded this article also downloaded:
J oa˜o A.C. Santos, (2009),"Do markets “discipline” all banks equally?", J ournal of Financial Economic
Policy, Vol. 1 Iss 1 pp. 107-123 http://dx.doi.org/10.1108/17576380910962402
Bruno Funchal, Mateus Clovis, (2009),"Firms' capital structure and the bankruptcy law design", J ournal of
Financial Economic Policy, Vol. 1 Iss 3 pp. 264-275 http://dx.doi.org/10.1108/17576380911041737
Carlos A. Ulibarri, Ionut Florescu, J oel M. Eidsath, (2009),"Regulating noisy short-selling
of troubled firms?", J ournal of Financial Economic Policy, Vol. 1 Iss 3 pp. 227-245 http://
dx.doi.org/10.1108/17576380911041719
Access to this document was granted through an Emerald subscription provided by emerald-srm:115632 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for
Authors service information about how to choose which publication to write for and submission guidelines
are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well as
providing an extensive range of online products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee
on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive
preservation.
*Related content and download information correct at time of download.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Short sales around M&A
announcements
Liuqing Mai
College of Business Administration, University of Missouri,
Saint Louis, Missouri, USA, and
Robert van Ness and Bonnie van Ness
School of Business Administration, University of Mississippi,
Oxford, Mississippi, USA
Abstract
Purpose – The purpose of this paper is to examine changes in short-sale transactions of target ?rms
and acquiring ?rms around merger and acquisition (M&A) announcements using daily short-sale
transaction data from the New York stock exchange and NASDAQ. The paper further aims to
investigate the link between short-sale transactions and trading costs.
Design/methodology/approach – Two abnormal short-sale measures are developed. Two
regression models based on the two short-sale measures are constructed and ordinary least squares
is used to estimate the regressions. Two samples to test bid-ask spreads (BAS) before and after M&A
announcements t-test are used.
Findings – The paper ?nds that target ?rms experience signi?cant excess short sales (ES) from
day 2 1 to day þ 7; while acquiring ?rms experience signi?cant ES from day 0 to day þ 20. For
acquiring ?rms, the ?ve-day pre-announcement abnormal short sale is negatively related to the
announcement day return and is positively related to post-announcement return. Such a relationship
for target ?rms is not observed. For target ?rms, it is found that changes in short activity are not
signi?cantly related to changes in trading cost. For acquiring ?rms, short activity changes are
positively related to quoted spreads and percentage quoted spreads. The short-sale activity changes
are negatively related to effective spreads.
Research limitations/implications – The paper is a ?rst step to understanding whether short
sales affect market liquidity around M&A announcements; therefore restriction is necessary.
Additional research can be done which should extend the current study to include the options market.
Practical implications – From the results, the paper cannot conclude that short sellers are informed
traders around M&A announcements. Therefore restrictions on short sales around M&A
announcements may not be warranted.
Originality/value – The paper ?lls an important blank in the existing literature by examining
short-sale transactions around M&A announcements. Such an investigation is of particular interest to
market regulators as they try to update the short-sale rules.
Keywords Sales, Acquisitions and mergers, Stock exchanges
Paper type Research paper
1. Introduction
In this paper, we examine changes in short-sale transactions of target ?rms and
acquiring ?rms around merger and acquisition (M&A) announcements. Speci?cally,
we test whether the short-sale transactions before M&A announcements are driven
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1757-6385.htm
The authors would like to thank George Christodoulakis (the Editor) and an anonymous referee
for helpful comments and suggestions. All errors are the responsibility of the authors.
Short sales
around M&A
177
Journal of Financial Economic Policy
Vol. 1 No. 2, 2009
pp. 177-197
qEmerald Group Publishing Limited
1757-6385
DOI 10.1108/17576380911010272
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
mainly by informed traders and associated with post-announcement stock price
performance. Further, we investigate the link between short-sale transactions and
trading costs.
Such an investigation is of particular interest to market regulators. The securities and
exchange commission (SEC) notes the need to update its regulations on short sales and
seeks public comments[1]. One of the SEC’s questions has asked is whether speci?c
market events, such as mergers and acquisitions, make a security more vulnerable to
abusive short-sale activity. If so, short selling could be regulated, or even prohibited,
before and during such events[2]. To date, short-sale research has not offered an answer
to this question. Prior studies on short-sales focus on the pricing implications of
restrictions on short sales and the use of short selling to form investment strategies
(Miller, 1977; Ross, 1977; Diamond and Verrecchia, 1987; Jarrow, 1980; Dyvbig, 1984;
Figlewski, 1981). Researchers examine short-sale transactions around earnings
announcements (Woolridge and Dickson, 1994; Christophe et al., 2004), earnings
restatements (Desai et al., 2006), and seasoned equity offerings (SEO) (Genard and
Nanda, 1993). Short-sale transactions around mergers and acquisitions are yet to be
examined. This paper addresses this gap.
Further, this study differs fromprevious studies in that it examines the link between
short sales and trading costs around M&A announcements. One bene?t provided by
short sales is liquidity which can be measured by the bid-ask spreads (BAS) and depth.
However, howshort sales affect liquidity around M&Aannouncements is not clear. This
study analyzes the issue from a market microstructure perspective by examining the
link between short sales and BASs. Understanding how short sales affect market
liquidity is of particular importance to market regulators because of the need to balance
the costs and bene?ts of restrictions.
This study is also of academic interest. Diamond and Verrecchia (1987) argue that
short sellers are sophisticated investors and hypothesize a negative relation between
short selling and future declines in stock price. However, empirical results are mixed.
Senchack and Starks (1993), Aitken et al. (1998), Dechow et al. (2001), and Christophe
et al. (2004) ?nd supporting evidence for this hypothesis. Woolridge and Dickson (1994)
and Brent et al. (1990) do not. We believe that M&A announcements provide a good
environment to study this hypothesis since informed traders typically trade before
M&A announcements (Kweon and Pinkerton, 1981). Given the cost of short selling, the
proportion of informed short sellers around M&A announcements should increase, as
they are motivated by their information advantage. In particular, prior research ?nds
that, around M&A announcements, target ?rms achieve signi?cant positive abnormal
returns; acquiring ?rm returns are mixed depending on the mode of acquisition, the
method of payment, and on average, acquiring ?rms earn zero abnormal returns
(Jensen and Ruback, 1983 and Jarrell et al., 1988). Thus, informed traders who are
aware of the upcoming M&A announcement should refrain from shorting stocks of the
target ?rms as target ?rms are known to have signi?cant positive gains; but short
more shares of the acquiring ?rm according to the deal-speci?c characteristics.
Therefore, M&A announcements offer a unique window to examine whether short
sellers are informed and the relationship between short sales and future stock prices.
Using daily short-sale transaction data, this study examines the relationship between
short sales and future stock prices of target ?rms and acquiring ?rms around merger
and acquisition (M&A) announcements.
JFEP
1,2
178
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
The rest of the paper is organized as follows: In Section 2, we develop testable
hypotheses related to short-sale activities of target ?rms and acquiring ?rms around
M&Aannouncements; inSection3, we describe the dataandmethodology; inSection4, we
report and analyze empirical results from testing the hypotheses; Section 5 concludes.
2. Hypotheses
A short sale is a sale of stock that one does not actually own, but has borrowed from
someone else. A short seller establishes his short position by selling the borrowed stock
and closes his short position by purchasing the stock and returning it to the lender at a
later time. A short seller pro?ts when the stock price falls. While the maximum gain
from selling short is achieved if the stock price falls to zero, the loss from selling short
can be unlimited if the stock price rises. Owing this high risk and potential to
manipulate stock prices, the SEC substantially restricts short selling (see Dechow et al.,
2001 and Finnerty, 2005 for discussion). For example, Rule 105 governs short-sale
activity immediately prior to public security offerings with the aim of safeguarding the
integrity of the capital raising process. No such rule has thus far been put on other
corporate events such as mergers and acquisitions around which informed traders tend
to be prevalent and manipulative short sales are likely. While the SEC realizes this
need, no empirical study on manipulative or informed short-sale activity prior to M&A
announcements has been documented.
Prior studies examine short seller behavior prior to important corporate news
releases. For example, Genard and Nanda (1993) investigate the potential of
manipulative short selling prior to a SEO and predict increased short selling prior to
SEOs. Christophe et al. (2004) ?nd that short sellers are active prior to earnings
announcements. Desai et al. (2006) ?nd similar results prior to earnings restatements.
M&A announcements, which are more unpredictable than other announcements,
such as earnings announcements, in terms of both timing and magnitude, will have
greater information asymmetry, thus, a more substantial impact on stock prices (Chae,
2005). Informed traders, acting strategically, may attempt to maximize their pro?ts
through short selling prior to M&A announcements. Therefore, around M&A
announcements, it is likely that informed short sellers are prevalent in the market. On
one hand, informed short sellers would refrain from taking a short position in a target
?rmstock as they expect the target ?rmstock price will rise. Hence, we expect to observe
a drop in short-sale activity of target ?rm stock prior to M&A announcements. On the
other hand, informed short sellers would short more shares of the acquiring ?rm’s
stock if the expected pro?t is high enough to cover the cost of short selling. Hence, we
expect to see an increase or no change inthe short-sale activity of the acquiring ?rmprior
to M&A announcements. According to the ef?cient market hypothesis (Fama, 1970),
we expect to observe short-sale activity quickly returns to the normal level after
M&A announcements. In this paper, we examine whether short sellers of target ?rms
(acquiring ?rms) cover (establish) their positions before M&A announcements.
Accordingly, we hypothesize the following:
H1. Short-sale activity of target ?rms (acquiring ?rms) decreases (increases or does
not change) before and quickly returns to normal after M&A announcements.
Diamond and Verrecchia (1987) suggest that short sales are mainly driven by informed
traders due to the fact that short selling costs are so high that liquidity traders ?nd it
Short sales
around M&A
179
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
too costly to short. Empirical studies are largely consistent with the notion that short
sellers are a subset of sophisticated investors in the investment community by
providing evidence from various aspects. For example, Dechow et al. (2001) ?nd that
short sellers use ratios of fundamentals (such as earnings and book value) to market
values to identify stocks with lower expected future returns. Efendi et al. (2009) argue
that short sellers are highly sophisticated investors who can see through accounting
manipulation and thus pro?t from their knowledge. Prior studies also investigate
whether or not short sellers are informed traders through observing short seller
behavior prior to important corporate news releases. Christophe et al. (2004) provide
evidence of informed trading in the ?ve days prior to earnings announcements of
NASDAQ-listed ?rms. However, little is known about informed short selling prior to
M&A announcements.
Previous market microstructure literature on M&A announcements documents
substantial evidence that informed trading increases before M&A announcement[3].
Prior to M&A announcements, informed traders will seek to maximize their pro?ts
based on their private information knowing that the information might be revealed or
detected by the market, thus, lose its value. Short sales, which allow investors to trade
on margin, can be a valuable venue for informed traders to maximize their pro?ts
because a margin account can magnify pro?ts[4]. In this paper, we aim to provide
evidence that information-driven short sales are prevalent before M&A
announcements. We test the following hypothesis:
H2. Short-sale activity before M&A announcements is mainly driven by informed
trades.
The relationship between short sales and subsequent stock returns is of great interest in
the short-sale literature. Figlewski (1981) ?nds that stocks with higher short interest
under-perform in subsequent months. Desai et al. (2002) ?nd that heavily shorted ?rms
experience signi?cant negative abnormal returns. Senchack and Starks (1993) report
abnormal returns around monthly short interest announcements are more negative the
higher the unexpected short interest. Similarly, using data from Australian stock
market, where short sales are made public immediately upon occurrence, Aitken et al.
(1998) ?nd an immediate drop in stock price after short-sale executions. Christophe et al.
(2004) ?nd that abnormally high short sales are linked to abnormally low subsequent
stock returns around earnings announcements. M&A announcements are similar to
other corporate announcements in that they result in information asymmetry in the
market, to varying degrees. Therefore, one would expect that changes in short-sale
activity prior to M&Aannouncements would be associated with both the announcement
and post-announcement stock returns. Speci?cally, a decrease in the short-sale activity
of a target ?rm’s stock prior to an M&A announcement will be associated with an
increase in both the announcement and post-announcement target ?rm’s stock return.
Accordingly, an increase or no change in short-sale activity of acquiring ?rm’s stock is
expected to be associated with a decrease or no change in both the announcement and the
post-announcement stock returns. We develop the following hypothesis:
H3. Changes in short-sale activity before M&Aannouncements are associated with
stock returns on the announcement date and those in the post-announcement
period.
JFEP
1,2
180
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Substantial market liquidity is provided through short selling by market professionals, such
as market makers, block positioners, and specialists, [. . .] To the extent that short sales are
affected in the market by securities professionals, such short-sale activities, in effect, add to
the trading supply of stock available to purchasers and reduce the risk that the price paid by
investors is arti?cially high because of a temporary contraction of supply. – SEC Concept
Release: Short Sales[5].
Although the SEC statement assumes an explicit link between short sales and
liquidity, empirical studies on this issue are limited. Jones (2008) studies changes in
liquidity around events that alter the level of short-sale constraints in the US market.
He ?nds that the introduction of the requirement that brokers secure written
authorization before lending a customer’s shares in 1932 had a negative impact on
liquidity, but the requirement that short sales be executed on an uptick in 1938 had a
positive effect on liquidity. Daouk and Charoenrook (2005) examine short sales from
111 countries and ?nd that, when short selling is possible, there is greater liquidity, as
measured by turnover ratio.
In this paper, we reexamine the link between short sales and trading costs,
measured by the BAS, using an event study of M&A announcements. Prior to M&A
announcements, we expect to observe a decrease in trading costs of target ?rms as
short-sale activity increases. After M&A announcements, the trading costs of both
target ?rms and acquiring ?rms are expected to return to a normal level quickly. The
following hypothesis is formed:
H4. Around M&A announcements, short-sale activity changes are associated
with changes in trading costs.
3. Data and methodology
3.1 Data
We obtain the sample of mergers and acquisitions from the securities data company
database. It includes mergers and acquisitions announced from January 1, 2005 to
December 31, 2005. We select the sample by the following criteria:
Acquiring ?rms must:
.
be listed on New York Stock Exchange (NYSE) or NASDAQ;
.
not use the following deal types: leveraged buyout, spinoff, self-tender,
recapitalization, exchange offer, repurchase, or privatization;
.
have a M&A deal value of one million or more;
.
have a closing price of $5 or more four weeks prior to announcement; and
.
not have any missing data during the sample period.
These criteria yield 1,010 acquiring ?rms.
Target ?rms must:
.
be listed on NYSE or NASDAQ;
.
not use the following deal types: leveraged buyout, spinoff, self-tender,
recapitalization, exchange offer, repurchase, or privatization; and
.
not have any missing data during the sample period.
Short sales
around M&A
181
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
These criteria yield 196 target ?rms. The trading price and quote data are obtained
from trade and quote. The short-sale data are from SEC regulation SHO-mandated
data[6]. Our sample comprises 252 trading days in year 2005.
Table I presents summary information for the sample ?rms. Average ?rm size for
the target ?rms is $3,112.2 million; for the acquiring ?rms is $10,390 million. Average
closing price for the target ?rms is $19.87; for the acquiring ?rms is $28.93. In our
sample, the acquiring ?rms are generally bigger and trade at higher prices than the
target ?rms. The acquiring ?rms are also more actively traded, as indicated by the
mean daily shares traded of 1,151,430 for acquiring ?rms and 534,910 for target ?rms.
The mean daily shorted shares for the acquiring ?rms is 248,590, while for target ?rms
is 110,370. The mean short-sale ratio (the average shorted shares to average total
shares traded) for the target ?rms is 17.93 percent, and for the acquiring ?rms is
24.49 percent. The maximum short-sale ratio exceeds 40 percent for target ?rms and
nearly 50 percent for acquiring ?rms. These ?gures imply that acquiring ?rms have
more short sales than target ?rms.
3.2 Calculation of abnormal short sales
We calculate the daily average cumulative excess short sales (ES) and the daily average
ES to examine the short-sale pattern of the sample ?rms. Let S
f,t
equal ln(1 þ daily
shorted sales) for day t for ?rm f and S
f
equal the mean of S
f,t
for days 2 60 to 221.
For ?rm f, ES is:
ES
f ;t
¼ S
f ;t
2S
f
t ¼ 260· · · þ 20 ð1Þ
We calculate the cumulative excess short sales (CES) on day t for ?rm f:
CES
f ;t
¼ ES
f ;t
þ CES
f ;t21
t ¼ 260· · · þ 20 ð2Þ
Mean Median Maximum Minimum
Panel A: target
Firm size ($ million) 3,112.20 397.31 103,904.00 1.02
Average price ($) 19.87 15.71 70.55 0.08
Average daily shorted shares (000’s shares) 110.37 29.26 2,745.99 0.30
Average daily volume (000’s shares) 534.91 165.15 12,569.12 0.49
Average shorted shares to average total shares (%) 17.93 17.64 42.39 0.71
Panel B: acquirer
Firm size ($ million) 10,390.00 1,297.30 369,166.00 5.61
Average price ($) 28.93 25.00 86.09 3.68
Average daily shorted shares (000’s shares) 248.59 75.17 10,238.22 0.14
Average daily volume (000’s shares) 1,151.43 280.88 60,596.71 1.04
Average shorted shares to average total shares (%) 24.49 23.01 49.18 1.48
Notes: This table contains key characteristics of all ?rms. Firm size is a stock’s market capitalization
calculated as the number of shares outstanding multiplied by price per share. Average price is the
average daily closing price. The average daily shorted shares is a stock’s average daily number of
shares shorted over the 252 trading days in year 2005. A stock’s average daily volume is the average
daily number of shares traded over the 252 day period. A stock’s average shorted shares as a
percentage of total shares is the average daily shorted shares/average daily number of share traded.
The sample contains 192 target ?rms and 1,010 acquiring ?rms that are listed on NYSE or NASDAQ
in year 2005
Table I.
Key characteristics
of involving ?rms
JFEP
1,2
182
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
where CES
f,260
¼ ES
f,260
, day t is relative to the day of the M&A announcement, and
day 0 is the announcement day. If there are unusual short-sale patterns, the mean of
CES
f,t
across ?rms will be signi?cantly greater than zero. We use a one-tailed t-test to
test the null hypothesis that the mean CES for day t is greater than zero. To examine if
there is an abnormal increase or decrease in short-sale activity around M&A
announcements, we use a two-tailed t-test to evaluate the null hypothesis that the mean
ES for day t is different from zero.
3.3 Regression model speci?cation
Following Christophe et al. (2004), we develop two regression models. In the ?rst
model, we de?ne abnormal short sales as the percentage difference between the
average daily number of the ?rm’s shares sold short during the ?ve days preceding the
M&A announcement and the average daily number of the ?rm’s shares sold short
during the non-announcement period. In the second model, we de?ne the relative short
sales in the pre-announcement period as the ratio of shorted shares to traded shares for
the stock from day 25 to 21. Implicit in the two regression models is the assumption
that the average daily short sales during the 41 pre-announcement day period, 260 to
221, is a fair representation of the ?rm’s typical daily short sales. More formally, a
stock’s abnormal short sales during the ?ve days prior to the M&A announcement
ABSS(25, 21), is measured as:
ABSSð25; 21Þ ¼ SSð25; 21Þ=AVESSð260; 221Þ 21 ð3Þ
where:
SS (25, 21) The average daily number of shorted shares for the ?ve days
prior to the M&A announcement.
AVESS(260, 221) The average daily number of shorted shares during the
non-announcement period.
To assess the robustness of our estimation, we specify RELSS(25, 21) as the relative
short sales of a ?rm in the pre-announcement period. RELSS(25, 21) is measured
as the ratio of shorted shares to traded shares for the stock over the interval of days
25 to 21. We test the following regression models:
ABSSð25; 21Þ ¼a
0
þ a
1
RETð0; þ1Þ þ a
2
RETð25; 21Þ þ a
3
RETðþ2; þ5Þ
þ a
4
ABVOLð25; 21Þ þ e
ð4Þ
RELSSð25; 21Þ ¼y
0
þ y
1
RETð0; þ1Þ þ y
2
RETð25; 21Þ þ y
3
RETðþ2; þ5Þ
þ y
4
NORMRELSS þ e
ð5Þ
where:
RET(0, þ 1) The event return on the stock calculated from the closing
prices of days 21 to þ 1, RET(25, 21) is the return on the
stock calculated from the closing prices of days 26 to 21.
ABVOL (25, 21) The average daily abnormal volume in the stock over the
interval of day 25 to 21.
Short sales
around M&A
183
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
The independent variable RET(0, þ 1) serves as a proxy for the announcement surprise.
Thus, a negative two-day return means that the market views the announcement as an
unfavorable (negative) surprise and a positive return means the announcement is more
encouraging (positive) than investors expect. Therefore, a negative a
1
means that short
sales regularly rise prior to disappointing M&A announcements and decrease prior to
favorable announcements. The model contains two control variables, RET(25, 21) and
ABVOL(25, 21). RET(25, 21) represents the movement of the stock price during the
?ve days prior to the announcement. This variable controls for the possibility that
changes in stock prices might affect the level of short sales in the days leading up to the
announcement. A pre-announcement increase in stock price, for example, might affect
short sales by enticing some investors to short the now “over-valued” stock. With this
control variable in place, the model will not incorrectly attribute all pre-announcement
short sales to expectations regarding the announcement. The second control variable,
ABVOL(25, 21), accounts for the potential contemporaneous correlation between
abnormal short sales and spikes in volume, and for the possibility that stocks
experiencing sudden increases in volume might be easier to short. Abnormal volume is
measured as the percentage difference between the average volume in the ?ve-day event
interval, and the average daily volume in the 41 days of the non-announcement period.
RET(þ2, þ 5) is the post-announcement return calculated from the closing prices of
day þ 1 to day þ 5. NORMRELSS, normalized relative short sales, is calculated as the
ratio of the shorted shares to traded shares for a stock during the non-announcement
period, day 260 to 221. The error term, e, captures the combined effects of omitted
variables.
3.4 BASs: calculation and regression model
We calculate four BAS measures: quoted spread, percentage quoted spread, effective
spread, and percentage effective spread as follows:
Quoted spread
it
¼ Ask
it
2Bid
it
Percentage quoted spread
it
¼ 2ðAsk
it
2Bid
it
Þ=ðAsk
it
þ Bid
it
Þ
Effective spread
it
¼ jPrice
it
2ðAsk
it
þ Bid
it
Þ=2j
Percentage effective spread
it
¼ 100ðjPrice
it
2ðAsk
it
þ Bid
it
Þ=2jÞ=ððAsk
it
þ Bid
it
Þ=2Þ
Quotes before the opening and after closing are excluded from the sample. To create
second-by-second data, each quote is carried forward until the next one arrives.
Because we retain each quote until it is updated, our spread measures are time
weighted. Quotes with longer lives have more weight in calculating the average daily
spreads. For each ?rm, a series of time-weighted percentage BAS is created by
averaging second-by-second quotations for each day during the 252 days of the sample
period. Then, for each stock, we average the time-weighted BAS for each day. For each
day, we average the time-weighted spreads across ?rms.
To examine the linkage of the short sale changes and trading cost changes around
M&A announcements, we apply multivariate regressions. McInish and Wood (1992)
suggest that spread is determined by price, activity, volatility, and volume. Here, we
add two short-sale measures to the regressions: the daily average short sales (DailySS)
JFEP
1,2
184
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
and the ratio of average short sales to volume (AvgSSVOLRatio). We estimate the
following regressions over the sample period (260, þ 20):
Spread ¼b
0
þ b
1
ð1=PRICEÞ þ b
2
logðVOLUMEÞ þ b
3
TURNOVER
þ b
4
VOLATILITY þ b
5
DailySS þ b
6
Dummy þ e
ð6Þ
and:
Spread ¼c
0
þ c
1
ð1=PRICEÞ þ c
2
logðVOLUMEÞ þ c
3
TURNOVER
þ c
4
VOLATILITY þ c
5
AvgSSVolRatio þ c
6
Dummy þ e
ð7Þ
Spread is the time-weighted spread measures speci?ed above. Price is the mean value
of daily closing prices. Volume is the dollar trading volume. Turnover is the ratio of the
number of shares traded to number of shares outstanding. Volatility is the standard
deviation of daily return calculated from the daily closing prices. DailySS is the
average daily short sale. AvgSSVolRatio is average daily shorted shares to daily
traded shares. Dummy is 1 if before M&A announcement, (260, 21), and 0 if after
announcement, (0, þ 20). The error term, e, captures the combined effects of omitted
variables. Both White’s test (White, 1980) and Breusch-Pagan test are performed to test
the existence of heteroscedasticity. Neither of the two tests is signi?cant at 5 percent
level. We estimate the regression models using ordinary least squares (OLS).
4. Empirical results
4.1 Hypothesis 1
Table II provides the CES and ES for days 260 to þ20 surrounding the M&A
announcements.
For target ?rms, CES becomes positive and statistically signi?cant at 5 percent level
on day 0. The t-statistics for CES are negative and statistically signi?cant at least at
the 10 percent level for each day from day 2 30 to day 2 21. This evidence con?rms
that there is a decrease in short-sale activity prior to the M&A announcement. More
importantly, this evidence shows that a signi?cant decrease in short-sale activity
occurs in a period from day 2 30 to day 2 21, which is not immediately prior to the
M&A announcement. We do not ?nd evidence that short-sale activity decreases
immediately before the M&A announcement day. After the M&A announcement, the
t-statistics for CES are positive and statistically greater than zero at the 1 percent level
from day þ 1 to day þ 20. This evidence shows that there is an increase in short-sale
activity after the M&A announcement. It does not support H1. However, we observe a
turn in short-sale activity after the M&A announcement. The value of CES ?rst
increases from day þ 1 to day þ 13, and then decreases consistently from day þ 14 to
day þ 20. ES becomes signi?cantly positive for each day from day 2 1 to day þ 7 and
is not signi?cant after day þ 7. Therefore, this result implies that short-sale activity
slowly returns to normal after the M&A announcement.
For acquiring ?rms, CES is negative and signi?cantly from day 2 60 to day 0. This
evidence shows that short-sale activity of acquiring ?rms decreases prior to the M&A
announcement. It does not support H1. CES is not signi?cant from day þ 1 to day þ 6,
and then becomes positive and signi?cantly greater than zero for each day from
day þ 7 to day þ 20. The ES of acquiring ?rms is positive and signi?cantly different
from zero for each day from day 0 to day 20. This shows that there are excess short
sales after the M&A announcement day. While the CES of target ?rms shows a
Short sales
around M&A
185
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
CES ES
Day relative Target t-statistics Acquirer t-statistics Target t-statistics Acquirer t-statistics
260 0.285 1.13 21.660 23.76
* * *
20.040 20.33 20.113 22.44
* * *
230 20.817 21.32
*
23.051 24.54
* * *
20.092 20.83 0.042 0.88
229 21.066 21.81
* *
23.012 24.47
* * *
20.248 21.97
* *
0.039 0.94
227 20.956 21.65
* *
22.996 24.45
* * *
0.087 0.76 20.060 21.10
226 20.915 21.65
* *
22.984 24.41
* * *
0.041 0.38 0.012 0.25
224 20.802 21.57
*
22.857 24.19
* * *
0.006 0.05 0.111 2.47
* * *
223 20.870 21.71
* *
22.815 24.11
* * *
20.068 20.51 0.041 0.83
222 20.845 21.73
* *
22.765 24.02
* * *
0.024 0.21 0.050 1.06
221 20.661 21.41
*
22.677 23.85
* * *
0.184 1.45 0.088 1.98
* *
220 20.621 21.27 22.611 23.76
* * *
0.040 0.32 0.066 1.35
219 20.631 21.25 22.532 23.64
* * *
20.010 20.08 0.080 1.83
*
218 20.667 21.25 22.515 23.61
* * *
20.036 20.29 0.016 0.33
217 20.677 21.19 22.394 23.43
* * *
20.010 20.08 0.120 3.17
* * *
216 20.725 21.17 22.321 23.32
* * *
20.048 20.35 0.074 1.78
* *
215 20.645 20.95 22.335 23.33
* * *
0.083 0.56 20.014 20.29
214 20.430 20.60 22.240 23.18
* * *
0.212 1.93
*
0.095 2.30
* *
213 20.599 20.78 22.152 23.05
* * *
20.170 21.06 0.088 1.93
*
212 20.601 20.73 21.979 22.81
* * *
20.002 20.01 0.173 4.23
* * *
211 20.542 20.63 21.910 22.71
* * *
0.059 0.53 0.069 1.58
210 20.460 20.51 21.859 22.63
* * *
0.082 0.64 0.051 1.12
29 20.250 20.26 21.785 22.52
* * *
0.209 1.61 0.074 1.76
* *
28 0.030 0.03 21.735 22.45
* * *
0.281 1.81
*
0.050 0.96
27 0.210 0.19 21.697 22.38
* * *
0.108 1.26 0.039 0.85
26 0.298 0.27 21.596 22.23
* * *
0.088 0.62 0.101 2.25
* *
25 0.496 0.42 21.583 22.19
* *
0.197 1.41 0.013 0.27
24 0.642 0.51 21.571 22.14
* *
0.147 0.96 0.012 0.23
23 1.094 0.83 21.410 21.91
* *
0.451 3.18
* * *
0.161 3.56
* * *
22 1.310 0.94 21.379 21.85
* *
0.217 1.33 0.031 0.64
21 1.654 1.12 21.333 21.77
* *
0.343 2.31
* *
0.045 0.93
0 3.127 2.02
* *
21.007 21.32
*
1.473 7.91
* * *
0.327 6.85
* * *
1 5.202 3.17
* * *
20.511 20.67 2.075 11.64
* * *
0.495 11.08
* * *
2 6.432 3.75
* * *
20.101 20.13 1.231 7.03
* * *
0.410 9.79
* * *
3 7.416 4.12
* * *
0.235 0.30 0.984 6.27
* * *
0.337 7.69
* * *
4 8.132 4.28
* * *
0.546 0.69 0.715 4.18
* * *
0.311 6.77
* * *
5 8.516 4.26
* * *
0.738 0.93 0.384 2.06
* *
0.192 3.86
* * *
6 8.887 4.26
* * *
1.007 1.25 0.371 2.17
* *
0.268 5.72
* * *
7 9.248 4.22
* * *
1.228 1.51
*
0.361 2.11
* *
0.221 4.43
* * *
8 9.420 4.09
* * *
1.452 1.76
* *
0.173 0.93 0.224 4.97
* * *
9 9.608 3.99
* * *
1.604 1.92
* *
20.083 1.15 0.152 2.98
* * *
10 9.813 3.87
* * *
1.799 2.12
* *
0.205 1.11 0.195 4.26
* * *
11 9.904 3.77
* * *
2.020 2.34
* * *
0.091 0.59 0.221 5.35
* * *
12 9.867 3.60
* * *
2.226 2.54
* * *
20.037 20.22 0.207 4.42
* * *
13 10.032 3.51
* * *
2.387 2.68
* * *
0.165 0.97 0.161 3.22
* * *
14 9.805 3.30
* * *
2.595 2.86
* * *
20.228 21.32 0.208 4.50
* * *
15 9.673 3.14
* * *
2.754 2.98
* * *
20.132 20.74 0.159 3.33
* * *
16 9.513 2.98
* * *
2.941 3.14
* * *
20.160 20.92 0.187 3.93
* * *
17 9.492 2.86
* * *
3.174 3.33
* * *
20.022 20.13 0.233 5.42
* * *
18 9.429 2.75
* * *
3.359 3.47
* * *
20.063 20.40 0.185 4.06
* * *
19 9.197 2.60
* * *
3.542 3.60
* * *
20.231 21.38
*
0.182 4.04
* * *
20 9.073 2.48
* * *
3.715 3.71
* * *
20.125 20.75 0.173 3.68
* * *
Notes: Statistically signi?cant at
*
10%,
* *
5%, and
* * *
1% levels; this table presents the daily average CES and
daily average ES of involving ?rms. We let S
f,t
equal Ln(1 þ Daily Shorted Sales) for day t for ?rm f and S
f
equal
the mean of S
f,t
for days 260 to 221. For ?rm f, ES is ES
f,t
¼ S
f,t
- S
f
for t ¼ 260. . . þ 20. Then, for day t for ?rm
f, we calculate the CES: CES
f,t
¼ ES
f,t
þ CES
f,t21
t ¼ 260. . . þ 20, where CES
f,260
¼ ES
f,260
, day t is relative
to the day of the M&A announcement, and day 0 is the announcement day
Table II.
Daily average CES
and daily average ES
JFEP
1,2
186
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
decrease in value from day þ 14 to day þ 20, the CES of acquiring ?rms continues to
increase after the M&A announcement day. Therefore, the short-sale activity of
acquiring ?rms does not quickly return to normal after M&A announcements. This
evidence does not support H1 either. Overall, we ?nd that the empirical evidence
weakly supports H1 about target ?rms but does not support H1 about acquiring ?rms.
4.2 Hypothesis 2
H2 states that short-sale activity before M&A announcements are driven by informed
traders. Since informed short sellers cannot be identi?ed directly, we need to show that
there are changes in the overall level of short-sale activity before M&A
announcements, and the changes in the level of trading before the announcement is
a result of informed trading.
Table II reports the average ES of both target ?rms and acquiring ?rms. Before the
M&A announcements, we observe an increase in short activity of target ?rms on
day 2 14, day 2 8, day 2 3, and day 2 1 and an increase in the short activity of
acquiring ?rms on days 214 to 212, day 2 9, day 2 6, and day 2 3.
Table III shows a two sample t-test of BAS before and after M&A announcements.
For target ?rms, all four spread measures show a signi?cant decrease after the M&A
announcements. For acquiring ?rms, quote spread and percentage quoted spread show
a signi?cant decrease after the announcements. Effective spread increases and
percentage effective spread does not change after announcements. Overall, we ?nd a
narrowing of BAS of target ?rms. How the BAS of acquiring ?rms change depends on
the spread measure used.
In summary, we ?nd ES on some days before merger announcements for both the
target ?rms and the acquiring ?rms. We do not ?nd consistent evidence of excess short
sale across days. We ?nd a narrowing of the BAS for target ?rms. The BAS of
acquiring ?rms show mixed results. Therefore, we cannot conclude that short-sale
activity before M&A announcements are driven by informed traders.
Pre mean Post mean Difference t-statistic
Target
Quoted spread 0.1240
* * *
0.0920 0.0320 16.15
Quoted spread (%) 0.0140
* * *
0.0106 0.0034 6.11
Effective spread 0.0637
* * *
0.0474 0.0163 9.90
Effective spread (%) 0.0081
* * *
0.0060 0.0021 4.40
Acquirer
Quoted spread 0.0956
* * *
0.0849 0.0107 2.93
Quoted spread (%) 0.0044
* * *
0.0041 0.0003 5.02
Effective spread 0.0606
* *
0.0697 20.0091 22.47
Effective spread (%) 0.0029 0.0028 0.0001 0.37
Notes: Statistically signi?cant at
*
10%,
* *
5%, and
* * *
1% levels; pre mean is calculated as the
average BAS before the M&A announcement from day 2 60 to day 2 1. Post mean is calculated as
the mean BAS after the M&A announcement from day 0 to day þ 20. Two-tailed test is used to test the
null hypothesis that pre mean 2 post mean ¼ 0
Table III.
Two sample t-test of BAS
Short sales
around M&A
187
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
4.3 Hypothesis 3
To test H3, we estimate equations (4) and (5). We conjecture that the transaction ratio
may affect short-sale activity around M&A announcement as an M&A transaction
may have a greater impact on smaller ?rms than on larger ?rms. Also the changes in
short activity in response to the M&A announcement may vary among thinly traded
?rms and actively traded ?rms. Therefore, we divide the sample ?rms into quintiles
according to the transaction ratio of the M&A deal value to ?rm value and the average
trading volume, respectively. Table IV reports the results of the regression of the full
sample and of quintiles determined by the transaction ratio. The quintiles range from
the smallest transaction ratio to the largest transaction ratio. Table V reports the
results of quintiles determined by daily average trading volume. The quintiles range
from the least trading volume to the most trading volume.
From Panel A in Table IV, the estimates of the coef?cients of equations (4) and (5)
for target ?rms are not signi?cant except for ABVOL (25, 21) and NORMRELSS.
That is, when target ?rms are tested as a whole sample, we do not observe a signi?cant
relationship between abnormal short sales and the announcement day return or the
post-announcement return. For acquiring ?rms, the estimates of the coef?cients of
equation (5) are not signi?cant except for NORMRELSS. However, results of equation
(4) show that ABSS (25, 21) is negatively related to RET (0, þ 1), and positively
related to RET (þ2, þ 5). This result provides some evidence that the market
generally views the M&A announcement as an unfavorable surprise for acquiring
?rms and that the announcement day return of acquiring ?rms is negative. The
estimated a
3
, the coef?cient of RET (þ2, þ 5), is signi?cantly positive. This implies
that short sellers of acquiring ?rms generally earn a positive post announcement
return. Therefore, using the pooled target sample and the pooled acquirer sample, H3 is
true only for acquiring ?rms, not for target ?rms.
Panel B in Table IV presents the regression results of target ?rms, by transaction
ratio quintile. We ?nd that for Quintile 1, both a
2
, the coef?cient of RET (25, 21), and
a
3
, the coef?cient of RET (þ2, þ 5), are signi?cantly positive; for Quintile 2, only a
3
is
signi?cantly positive; for other quintiles, the estimates of coef?cients are not
statistically signi?cant. These results are contradictory to our conjecture. These results
show that, for target ?rms with smallest M&A transaction ratio, short sellers earn
positive returns both before and after M&A announcements. For target ?rms in
Quintile 2, short sellers only earn positive post announcement returns. This anomaly is
yet to be explained. Using equation (5), we do not observe signi?cant relation between
the ratio of shorted shares to traded shares, and investor returns before, on, or after
M&A announcements.
Acquiring ?rm results by transaction ratio are in Panel C of Table IV. Equation (4)
results show that for Quintile 1, a
2
, the coef?cient of RET (25, 21), is signi?cantly
positive; for Quintile 3, a
3
, the coef?cient of RET (þ2, þ 5), is signi?cantly positive; for
Quintile 4, a
1
, the coef?cient of RET (0, þ 1), is signi?cantly negative. Similar to the
results of target ?rms, but again contradictory to our conjecture, short sellers generally
earn positive pre-announcement returns for ?rms with the smallest transaction ratio in
our sample. For acquiring ?rms with an average transaction ratio, abnormal short
activity is positively related to post-announcement return. For acquiring ?rms with
larger transaction ratios (Quintile 4), short activity is associated with a negative
announcement day return. Using equation (5), we ?nd that y
2
, the coef?cient of RET
JFEP
1,2
188
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Panel A: regression results of full sample
Equation (4)
a
0
a
1
a
2
a
3
a
4
Adjusted R
2
Target 0.172
*
1.232 0.715 20.035 0.899
* * *
0.4360
(n ¼ 192) (1.80) (0.56) (0.63) (20.42) (12.23)
Acquirer 0.075
*
22.480
* * *
0.225 0.846
* * *
1.749
* * *
0.5312
(n ¼ 1,010) (1.71) (22.83) (0.35) (4.55) (32.48)
Equation (5)
y
0
y
1
y
2
y
3
y
4
Adjusted R
2
Target 0.085
* * *
20.045 20.005 20.005 0.642
* * *
0.2285
(n ¼ 192) (3.91) (20.23) (20.05) (20.62) (7.14)
Acquirer 0.083
* * *
20.020 0.038 0.014 0.682
* * *
0.2850
(n ¼ 1,010) (9.57) (20.35) (20.88) (1.13) (19.96)
Panel B: regression results of target ?rms by transaction ratio quintiles
Equation (4)
Target a
0
a
1
a
2
a
3
a
4
Adjusted R
2
Quintile 1 0.250 2.650 5.537
* *
9.537
* *
0.994
* * *
0.6102
(n ¼ 38) (1.03) (0.66) (2.15) (2.19) (6.16)
Quintile 2 0.185 3.895 27.050 1.750
* *
3.095
* * *
0.8888
(n ¼ 38) (1.10) (0.57) (21.58) (2.00) (14.79)
Quintile 3 20.131 0.655 1.023 20.682 1.169
* * *
0.7325
(n ¼ 38) (21.55) (0.40) (0.59) (20.92) (8.63)
Quintile 4 0.277 21.637 5.364 22.394 0.731
* * *
0.5513
(n ¼ 40) (0.91) (0.33) (0.43) (20.22) (5.86)
Quintile 5 0.326 20.236 21.477 0.503 0.681
* * *
0.1303
(n ¼ 40) (1.13) (20.02) (20.32) (0.28) (1.52)
Equation (5)
Target y
0
y
1
y
2
y
3
y
4
Adjusted R
2
Quintile 1 0.084 20.284 20.169 0.602 0.705
* *
0.1000
(n ¼ 38) (1.02) (20.43) (0.44) (0.86) (2.33)
Quintile 2 0.012 20.552 20.160 0.178 1.058
* * *
0.4671
(n ¼ 38) (0.21) (20.59) (20.25) (1.47) (4.88)
Quintile 3 0.074 0.164 0.281 0.054 0.555
* * *
0.2393
(n ¼ 38) (2.42) (20.66) (0.33) (0.23) (2.35)
Quintile 4 0.214
* *
0.081 20.593 0.581 0.164 0.0300
(n ¼ 40) (0.91) (0.33) (0.43) (20.22) (5.86)
Quintile 5 0.144
* * *
0.105 20.270 0.029 0.315 0.1240
(n ¼ 40) (2.89) (0.15) (20.98) (0.26) (1.20)
Panel C: regression results of acquiring ?rms by transaction ratio quintiles
Equation (4)
Acquirer a
0
a
1
a
2
a
3
a
4
Adjusted R
2
Quintile 1 0.019 20.450 1.320
* * *
20.059 0.960
* * *
0.5939
(n ¼ 197) (0.74) (20.76) (2.36) (20.22) (16.86)
Quintile 2 0.014 20.078 20.073 0.058 0.737
* * *
0.4466
(n ¼ 197) (0.54) (20.08) (20.30) (0.31) (12.71)
Quintile 3 0.196 24.868 21.159 4.841
* * *
2.029
* * *
0.6564
(n ¼ 197) (1.10) (20.98) (20.32) (5.77) (14.84)
Quintile 4 0.093 214.276
* * *
20.971 20.012 0.960
* * *
0.4445
(n ¼ 197) (1.49) (29.94) (20.72) (20.09) (8.00)
Quintile 5 0.108
*
0.196 0.616 20.787 1.279
* * *
0.6602
(n ¼ 200) (1.82) (0.26) (0.73) (20.87) (19.72)
(continued)
Table IV.
Results of OLS
regression: abnormal
short sales and abnormal
relative short sales
around M&A
announcements
Short sales
around M&A
189
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
(25, 21), is signi?cantly positive in Quintile 1 and y
3
, the coef?cient of RET (þ2, þ 5)
is signi?cantly positive in Quintile 3. These results are similar to those of equation (4).
Sample ?rms are also divided into quintiles by average trading volume and the
regression results are in Table V. We do not ?nd evidence of differences between
actively traded target ?rms and thinly traded target ?rms. We do ?nd that y
3
, the
coef?cient of RET (þ2, þ 5), is signi?cantly positive for Quintile 3. This indicates that
the ratio of shorted shares to traded shares is positively related to the
post-announcement return for averagely traded target ?rms. For acquiring ?rms, we
?nd that a
3
, the coef?cient of RET (þ2, þ 5), is signi?cantly positive for Quintile 1 and
a
1
, the coef?cient of RET (0, þ 1), is signi?cantly negative for Quintile 3. Also, y
3
, the
coef?cient of RET (þ2, þ 5), is signi?cantly negative and for Quintile 2 and
signi?cantly positive for Quintile 5. These results show that abnormal short-sale
activity before the M&A announcement is positively related to the post-announcement
return for the least traded acquiring ?rms and is negatively related the announcement
day return for averagely traded acquiring ?rms. The ratio of shorted shares to traded
shares tends to be negatively related to the post-announcement return for acquiring
Equation (5)
Acquirer y
0
y
1
y
2
y
3
y
4
Adjusted R
2
Quintile 1 0.063
* * *
20.066 0.250
* *
20.030 0.747
* * *
0.3695
(n ¼ 197) (3.45) (20.54) (2.18) (20.56) (10.56)
Quintile 2 0.096
* * *
0.026 0.016 0.005 0.619
* * *
0.3061
(n ¼ 197) (5.50) (0.12) (0.29) (0.13) (9.44)
Quintile 3 0.073
* * *
20.076 0.021 0.129
* * *
0.744
* * *
0.2684
(n ¼ 197) (3.12) (20.35) (0.13) (3.76) (8.17)
Quintile 4 0.081
* * *
20.002 20.213 20.008 0.701
* * *
0.2854
(n ¼ 197) (3.98) (20.01) (21.59) (20.57) (8.77)
Quintile 5 0.094
* * *
0.015 0.104 20.103 0.639
* * *
0.2259
(n ¼ 200) (4.70) (0.14) (0.91) (20.84) (7.70)
Notes: Statistically signi?cant at
*
10%,
* *
5%, and
* * *
1% levels
ABSSð25;21Þ ¼a
0
þa
1
RETð0;þ1Þþa
2
RETð25;21Þþa
3
RETðþ2;þ5Þþa
4
ABVOLð25;21Þþe ð4Þ
RELSSð25;21Þ ¼y
0
þy
1
RETð0;þ1Þþy
2
RETð25;21Þþy
3
RETðþ2;þ5Þþy
4
NORMRELSSþe ð5Þ
The results of OLS estimation of these equations, as ?tted to the full sample and sub-samples
determined by transaction ratio of the M&A deal value to ?rm value are shown. ABSS(25, 2 1) is the
average daily abnormal short sales for stock in the pre-announcement period, measured as the average
daily short sale in the pre-announcement period divided by the average daily short sale in the non-
announcement period, all 21. RELSS(25, 21) is the ratio of shorted shares to traded shares in the
stock in the pre-announcement period. RET(0, þ 1) is the stock’s two-day percentage return following
the M&A announcement and measured from the closing price on day 2 1 to that on day þ 1.
RET(25, 21) is the stock’s percentage return before the M&A announcement and measured from the
closing price on day 2 6 to that on day 2 1. RET(þ2, þ 5) is the post-announcement return
calculated from the closing prices of day þ 1 to day þ 5. ABVOL(25, 21) is the stock’s abnormal
volume in the pre-announcement period, measured as the average daily volume in the pre-
announcement period divided by the average daily volume in the non-announcement period, all 21.
NORMRELS is the ratio of the shorted shares to trade shares in the non-announcement period. t-
Statistics are in the parentheses below the coef?cients. Quintiles are determined according to the
transaction ratio Table IV.
JFEP
1,2
190
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Panel A: regression results of target ?rms by average daily volume quintiles
Equation (4)
Target a
0
a
1
a
2
a
3
a
4
Adjusted R
2
Quintile 1 0.324 7.577 0.819 5.003 0.675
* * *
0.4734
(n ¼ 38) (1.22) (0.95) (0.32) (0.95) (5.86)
Quintile 2 0.225 1.939 0.268 0.889 1.026
* * *
0.3312
(n ¼ 38) (1.20) (0.38) (0.06) (0.85) (4.57)
Quintile 3 0.321 1.038 1.860 1.192 2.568
* * *
0.6082
(n ¼ 38) (1.25) (1.19) (0.95) (0.67) (7.80)
Quintile 4 0.008 0.830 20.434 20.040 0.863
* * *
0.8356
(n ¼ 38) (0.13) (0.68) (20.26) (21.46) (12.73)
Quintile 5 20.036 20.865 21.007 20.476 1.045
* * *
0.6623
(n ¼ 38) (20.27) (20.27) (20.35) (20.38) (8.92)
Equation (5)
Target y
0
y
1
y
2
y
3
y
4
Adjusted R
2
Quintile 1 0.111
*
0.043 20.109 0.234 0.519
*
0.1236
(n ¼ 38) (1.74) (0.05) (20.39) (0.41) (1.95)
Quintile 2 0.012 20.552 20.160 0.178 1.058
* * *
0.4671
(n ¼ 38) (0.21) (20.59) (20.25) (1.47) (4.88)
Quintile 3 0.005 0.519 0.058 0.203
* *
1.013
* * *
0.5236
(n ¼ 38) (0.14) (1.24) (0.16) (2.20) (6.46)
Quintile 4 0.142
* *
20.358 0.063 0.040 0.538
* *
0.1472
(n ¼ 40) (2.42) (20.66) (0.33) (0.23) (2.35)
Quintile 5 0.071 0.050 20.075 20.006 0.696
* * *
0.4487
(n ¼ 40) (1.96) (0.23) (20.27) (20.98) (4.82)
Panel B: regression results of acquiring ?rms by average daily volume quintiles
Equation (4)
Acquirer a
0
a
1
a
2
a
3
a
4
Adjusted R
2
Quintile 1 0.219 25.950 22.103 6.418
* * *
2.286
* * *
0.7176
(n ¼ 197) (1.40) (20.99) (20.55) (7.34) (17.18)
Quintile 2 0.049 0.175 0.131 21.011 0.591
* * *
0.3148
(n ¼ 197) (1.36) (0.35) (0.17) (21.43) (9.36)
Quintile 3 0.132
* *
212.077
* * *
0.871 20.585 1.252
* * *
0.4248
(n ¼ 197) (2.12) (28.77) (1.02) (20.89) (8.49)
Quintile 4 20.053
*
21.355 20.252 20.021 1.131
* * *
0.8855
(n ¼ 197) (21.70) (21.33) (20.36) (20.31) (39.27)
Quintile 5 0.041
* *
0.237 0.083 20.161 0.849
* * *
0.5784
(n ¼ 200) (2.20) (0.59) (0.49) (21.17) (16.65)
Equation (5)
Acquirer y
0
y
1
y
2
y
3
y
4
Adjusted R
2
Quintile 1 0.053
* * *
20.004 0.121 20.040 0.800
* * *
0.4284
(n ¼ 197) (3.20) (20.05) (0.99) (21.09) (12.21)
Quintile 2 0.052
* * *
20.253 20.013 20.104
* *
0.831
* * *
0.4359
(n ¼ 197) (3.15) (21.09) (20.12) (22.04) (12.47)
Quintile 3 0.123
* * *
0.108 0.085 0.094 0.524
* * *
0.2116
(n ¼ 197) (6.58) (0.81) (0.55) (1.51) (7.31)
Quintile 4 0.102
* * *
20.027 20.015 20.012 0.581
* * *
0.1779
(continued)
Table V.
Results of OLS
regression: abnormal
short sales and abnormal
relative short sales
around M&A
announcements
Short sales
around M&A
191
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
?rms with below average trading volume, but positively related to the
post-announcement return for acquiring ?rms with the highest trading volume.
4.4 Hypothesis 4
To test whether short sale changes are associated with changes in trading costs, we
estimated equations (6) and (7). Results are presented in Tables VI and VII, respectively.
From Table VI, we can see that b
6
, the coef?cient of Dummy, is statistically signi?cant
when spread is measured as quoted spread, percentage quoted spread, and effective
spread. None of the variables are signi?cant, when spread is measured by percentage
effective spread. When quoted spread is used, b
1
, the coef?cient of (1/price), is
signi?cant. When effective spread is used, b
3
, the coef?cient of turnover, and b
4
, the
coef?cient of volatility, are signi?cant. Overall, when testing equation (6) for target
?rms, Dummy explains a large portion of spread variation before and after the M&A
announcement. b
5
, the coef?cient of DailySS is not signi?cant for any spread measure.
Therefore, we conclude that DailySS has little effect on BAS of target ?rms.
For acquiring ?rms, b
3
, the coef?cient of turnover, and b
4
, the coef?cient of volatility,
are signi?cant when using quoted spread, percentage quoted spread, and effective
spread. b
2
, the coef?cient of log (volume), is signi?cant when using quoted spread. b
1
, the
coef?cient of (1/price), is signi?cant when using percentage quoted spread. None of the
variables is signi?cant when using percentage effective spread. b
5
, the coef?cient of
DailySS, and b
6
, the coef?cient of Dummy, are not signi?cant for any spread measure.
Therefore, neither the daily short sale nor dummy variable can explain spread variation
in acquiring ?rms.
From Table VII, we ?nd similar results for target ?rms as equation (6). For
acquiring ?rms, c
1
, the coef?cient of 1/price), is signi?cant only in percentage quoted
(n ¼ 197) (4.73) (20.16) (20.08) (20.78) (6.76)
Quintile 5 0.073
* * *
20.033 0.027 0.186
* * *
0.715
* * *
0.3036
(n ¼ 200) (3.36) (20.14) (0.43) (4.89) (8.46)
Notes: Statistical signi?cant at
*
10%,
* *
5%, and
* * *
1% levels
ABSSð25;21Þ ¼a
0
þa
1
RETð0;þ1Þþa
2
RETð25;21Þþa
3
RETðþ2;þ5Þþa
4
ABVOLð25;21Þþe ð4Þ
RELSSð25;21Þ ¼y
0
þy
1
RETð0;þ1Þþy
2
RETð25;21Þþy
3
RETðþ2;þ5Þþy
4
NORMRELSSþe ð5Þ
The results of OLS estimation of these equations, as ?tted to the sub-samples determined by
average daily volume are shown. ABSS(25, 2 1) is the average daily abnormal short sales for
stock in the pre-announcement period, measured as the average daily short sale in the pre-
announcement period divided by the average daily short sale in the non-announcement period, all
21. RELSS(25, 21) is the ratio of shorted shares to traded shares in the stock in the pre-
announcement period. RET(0, þ 1) is the stock’s two-day percentage return following the M&A
announcement and measured from the closing price on day 2 1 to that on day þ 1. RET(25, 21)
is the stock’s percentage return before the M&A announcement and measured from the closing
price on day 2 6 to that on day 2 1. RET(þ2, þ 5) is the post-announcement return calculated
from the closing prices of day þ 1 to day þ 5. ABVOL(25, 21) is the stock’s abnormal volume in
the pre-announcement period, measured as the average daily volume in the pre-announcement
period divided by the average daily volume in the non-announcement period, all 21.
NORMRELSS is the ratio of the shorted shares to trade shares in the non-announcement period. t-
Statistics are in the parentheses below the coef?cients. Quintiles are determined according to
average daily volume Table V.
JFEP
1,2
192
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
b
0
b
1
b
2
b
3
b
4
b
5
b
6
A
d
j
u
s
t
e
d
R
2
T
a
r
g
e
t
Q
u
o
t
e
d
s
p
r
e
a
d
0
.
2
6
3
*
(
1
.
6
9
)
2
2
.
2
4
9
*
(
2
1
.
8
7
)
2
0
.
0
0
5
(
2
0
.
8
1
)
2
0
.
1
4
7
(
2
0
.
4
2
)
0
.
0
8
3
(
0
.
6
1
)
0
.
0
0
0
(
0
.
6
2
)
0
.
0
3
9
*
*
*
(
4
.
7
3
)
0
.
7
7
7
2
Q
u
o
t
e
d
s
p
r
e
a
d
(
%
)
0
.
0
1
1
(
0
.
1
9
)
2
0
.
1
4
7
(
2
0
.
3
3
)
0
.
0
0
0
(
0
.
0
0
)
2
0
.
0
0
0
(
2
0
.
0
0
)
0
.
0
2
0
(
0
.
3
9
)
2
0
.
0
0
0
(
2
0
.
1
8
)
0
.
0
0
5
*
(
1
.
7
4
)
0
.
3
7
7
5
E
f
f
e
c
t
i
v
e
s
p
r
e
a
d
0
.
1
9
4
(
1
.
3
1
)
2
0
.
0
0
0
(
2
0
.
0
0
)
2
0
.
0
0
7
(
2
1
.
2
8
)
0
.
5
5
9
*
(
1
.
6
8
)
2
0
.
2
6
6
*
*
(
2
2
.
0
6
)
0
.
0
0
0
(
0
.
4
9
)
0
.
0
1
8
*
*
(
2
.
2
8
)
0
.
5
6
7
8
E
f
f
e
c
t
i
v
e
s
p
r
e
a
d
(
%
)
0
.
0
1
6
(
0
.
2
9
)
0
.
0
4
9
(
0
.
1
2
)
2
0
.
0
0
1
(
2
0
.
3
2
)
0
.
0
6
5
(
0
.
5
3
)
2
0
.
0
1
4
(
2
0
.
2
9
)
2
0
.
0
0
0
(
2
0
.
1
7
)
0
.
0
0
3
(
0
.
9
3
)
0
.
1
9
6
8
A
c
q
u
i
r
e
r
Q
u
o
t
e
d
s
p
r
e
a
d
3
.
3
1
3
(
1
.
5
3
)
1
1
.
6
7
0
(
1
.
1
8
)
2
0
.
1
5
0
*
(
2
1
.
6
8
)
7
.
1
4
3
*
*
(
2
.
2
5
)
2
0
.
0
0
3
*
*
(
2
2
.
3
5
)
0
.
0
0
0
(
0
.
4
7
)
0
.
0
0
2
(
0
.
2
8
)
0
.
2
1
6
7
Q
u
o
t
e
d
s
p
r
e
a
d
(
%
)
0
.
0
5
4
(
1
.
2
6
)
0
.
4
6
1
*
*
(
2
.
3
6
)
2
0
.
0
0
3
(
2
1
.
5
5
)
0
.
1
2
1
*
(
1
.
9
2
)
2
0
.
0
0
0
*
*
(
2
2
.
1
6
)
0
.
0
0
0
(
0
.
9
6
)
0
.
0
0
0
(
0
.
6
9
)
0
.
3
4
0
3
E
f
f
e
c
t
i
v
e
s
p
r
e
a
d
2
2
.
7
2
1
(
2
1
.
2
9
)
2
7
.
1
4
6
(
2
0
.
7
5
)
0
.
1
2
5
(
1
.
4
3
)
2
6
.
8
0
5
*
*
(
2
2
.
2
1
)
0
.
0
0
3
*
*
*
(
2
.
8
2
)
2
0
.
0
0
0
(
2
0
.
1
3
)
2
0
.
0
0
2
(
2
0
.
2
1
)
0
.
2
0
2
5
E
f
f
e
c
t
i
v
e
s
p
r
e
a
d
0
.
0
7
1
(
0
.
5
0
)
0
.
7
3
2
(
1
.
1
3
)
2
0
.
0
0
4
(
2
0
.
6
5
)
2
0
.
0
4
3
(
2
0
.
2
1
)
0
.
0
0
0
(
1
.
0
2
)
0
.
0
0
0
(
0
.
5
2
)
2
0
.
0
0
1
(
2
0
.
9
7
)
0
.
0
3
4
8
N
o
t
e
s
:
S
t
a
t
i
s
t
i
c
a
l
s
i
g
n
i
?
c
a
n
t
a
t
*
1
0
%
,
*
*
5
%
,
a
n
d
*
*
*
1
%
l
e
v
e
l
s
S
p
r
e
a
d
¼
b
0
þ
b
1
ð
1
=
P
R
I
C
E
Þ
þ
b
2
l
o
g
ð
V
O
L
U
M
E
Þ
þ
b
3
T
U
R
N
O
V
E
R
þ
b
4
V
O
L
A
T
I
L
I
T
Y
þ
b
5
D
a
i
l
y
S
S
þ
b
6
D
u
m
m
y
þ
e
ð
6
Þ
S
p
r
e
a
d
i
s
t
h
e
t
i
m
e
-
w
e
i
g
h
t
e
d
s
p
r
e
a
d
m
e
a
s
u
r
e
s
.
P
r
i
c
e
i
s
t
h
e
m
e
a
n
v
a
l
u
e
o
f
d
a
i
l
y
c
l
o
s
i
n
g
p
r
i
c
e
s
.
V
o
l
u
m
e
i
s
t
h
e
d
o
l
l
a
r
t
r
a
d
i
n
g
v
o
l
u
m
e
.
T
u
r
n
o
v
e
r
i
s
t
h
e
r
a
t
i
o
o
f
t
h
e
n
u
m
b
e
r
o
f
s
h
a
r
e
s
t
r
a
d
e
d
t
o
n
u
m
b
e
r
o
f
s
h
a
r
e
s
o
u
t
s
t
a
n
d
i
n
g
.
V
o
l
a
t
u
l
i
t
y
i
s
t
h
e
s
t
a
n
d
a
r
d
d
e
v
i
a
t
i
o
n
o
f
d
a
i
l
y
r
e
t
u
r
n
c
a
l
c
u
l
a
t
e
d
f
r
o
m
t
h
e
d
a
i
l
y
c
l
o
s
i
n
g
p
r
i
c
e
s
.
D
a
i
l
y
S
S
i
s
t
h
e
a
v
e
r
a
g
e
d
a
i
l
y
s
h
o
r
t
s
a
l
e
.
D
U
M
M
Y
i
s
1
i
f
b
e
f
o
r
e
M
&
A
a
n
n
o
u
n
c
e
m
e
n
t
,
(
2
6
0
,
2
1
)
,
a
n
d
0
i
f
a
f
t
e
r
a
n
n
o
u
n
c
e
m
e
n
t
,
(
0
,
þ
2
0
)
Table VI.
Results of OLS
regression: spreads and
average daily short sales
around M&A
announcements
Short sales
around M&A
193
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
c
0
c
1
c
2
c
3
c
4
c
5
c
6
A
d
j
u
s
t
e
d
R
2
T
a
r
g
e
t
Q
u
o
t
e
d
s
p
r
e
a
d
0
.
2
4
0
*
(
1
.
6
9
)
2
2
.
2
9
7
*
(
2
1
.
9
3
)
2
0
.
0
0
4
(
2
0
.
6
9
)
2
0
.
0
5
1
(
2
0
.
1
6
)
0
.
1
0
6
(
0
.
8
1
)
2
0
.
0
1
3
(
2
1
.
4
4
)
0
.
0
4
1
*
*
*
(
5
.
0
4
)
0
.
7
8
2
2
Q
u
o
t
e
d
s
p
r
e
a
d
(
%
)
0
.
0
1
8
(
0
.
3
4
)
2
0
.
1
6
5
(
2
0
.
3
7
)
2
0
.
0
0
0
(
2
0
.
1
3
)
2
0
.
0
0
8
(
2
0
.
0
7
)
0
.
0
1
8
(
0
.
3
7
)
2
0
.
0
0
2
(
2
0
.
6
9
)
0
.
0
0
6
*
(
1
.
8
1
)
0
.
3
8
1
2
E
f
f
e
c
t
i
v
e
s
p
r
e
a
d
0
.
1
6
9
(
1
.
2
4
)
0
.
0
0
8
(
0
.
0
1
)
2
0
.
0
0
6
(
2
1
.
2
0
)
0
.
6
2
7
*
(
2
.
0
6
)
2
0
.
2
5
0
*
(
2
1
.
9
9
)
2
0
.
0
0
3
(
2
0
.
3
1
)
0
.
0
1
9
*
*
(
2
.
3
8
)
0
.
5
6
7
0
E
f
f
e
c
t
i
v
e
s
p
r
e
a
d
(
%
)
0
.
0
2
2
(
0
.
4
3
)
0
.
0
3
5
(
0
.
0
8
)
2
0
.
0
0
1
(
2
0
.
4
9
)
0
.
0
5
8
(
0
.
5
1
)
2
0
.
0
1
6
(
2
0
.
3
3
)
2
0
.
0
0
2
(
2
0
.
5
6
)
0
.
0
0
3
(
0
.
9
7
)
0
.
1
9
9
8
A
c
q
u
i
r
e
r
Q
u
o
t
e
d
s
p
r
e
a
d
1
.
2
6
7
(
1
.
1
4
)
1
4
.
2
2
5
(
1
.
5
3
)
2
0
.
0
8
1
*
(
2
1
.
8
6
)
6
.
7
7
6
*
*
(
2
.
2
6
)
2
0
.
0
0
3
*
*
*
(
2
2
.
7
6
)
1
.
1
2
4
*
*
*
(
2
.
9
7
)
0
.
0
1
1
(
1
.
3
2
)
0
.
2
9
8
1
Q
u
o
t
e
d
s
p
r
e
a
d
(
%
)
2
0
.
0
0
7
(
2
0
.
3
0
)
0
.
5
2
4
*
*
*
(
2
.
8
5
)
2
0
.
0
0
1
(
2
0
.
6
6
)
0
.
1
1
5
*
(
1
.
9
5
)
2
0
.
0
0
0
*
*
(
2
2
.
5
9
)
0
.
0
2
4
*
*
*
(
3
.
2
7
)
0
.
0
0
0
*
(
1
.
8
9
)
0
.
4
1
4
6
E
f
f
e
c
t
i
v
e
S
p
r
e
a
d
2
1
.
5
1
7
(
2
1
.
3
8
)
2
9
.
0
4
2
(
2
0
.
9
8
)
0
.
0
8
8
*
*
(
2
.
0
5
)
2
6
.
4
1
9
*
*
(
2
2
.
1
7
)
0
.
0
0
3
*
*
*
(
3
.
1
8
)
2
0
.
9
3
9
*
*
(
2
2
.
5
2
)
2
0
.
0
0
9
(
2
1
.
0
7
)
0
.
2
6
5
3
E
f
f
e
c
t
i
v
e
s
p
r
e
a
d
(
%
)
0
.
0
3
3
0
.
4
3
(
0
.
7
1
3
)
(
1
.
1
1
)
2
0
.
0
0
2
(
2
0
.
6
3
)
2
0
.
0
2
2
(
2
0
.
1
0
)
0
.
0
0
0
(
1
.
1
3
)
2
0
.
0
2
6
(
2
1
.
0
0
)
2
0
.
0
0
1
(
2
1
.
2
0
)
0
.
0
4
4
0
N
o
t
e
s
:
S
t
a
t
i
s
t
i
c
a
l
l
y
s
i
g
n
i
?
c
a
n
t
a
t
*
1
0
%
,
*
*
5
%
,
a
n
d
*
*
*
1
%
l
e
v
e
l
s
;
S
p
r
e
a
d
¼
c
0
þ
c
1
ð
1
=
P
R
I
C
E
Þ
þ
c
2
l
o
g
ð
V
O
L
U
M
E
Þ
þ
c
3
T
U
R
N
O
V
E
R
þ
c
4
V
O
L
A
T
I
L
I
T
Y
þ
c
5
A
v
g
S
S
V
o
l
R
a
t
i
o
þ
c
6
D
u
m
m
y
þ
e
ð
7
Þ
S
p
r
e
a
d
i
s
t
h
e
t
i
m
e
-
w
e
i
g
h
t
e
d
s
p
r
e
a
d
m
e
a
s
u
r
e
s
.
P
r
i
c
e
i
s
t
h
e
m
e
a
n
v
a
l
u
e
o
f
d
a
i
l
y
c
l
o
s
i
n
g
p
r
i
c
e
s
.
V
o
l
u
m
e
i
s
t
h
e
d
o
l
l
a
r
t
r
a
d
i
n
g
v
o
l
u
m
e
.
T
u
r
n
o
v
e
r
i
s
t
h
e
r
a
t
i
o
o
f
t
h
e
n
u
m
b
e
r
o
f
s
h
a
r
e
s
t
r
a
d
e
d
t
o
n
u
m
b
e
r
o
f
s
h
a
r
e
s
o
u
t
s
t
a
n
d
i
n
g
.
V
o
l
a
t
u
l
i
t
y
i
s
t
h
e
s
t
a
n
d
a
r
d
d
e
v
i
a
t
i
o
n
o
f
d
a
i
l
y
r
e
t
u
r
n
c
a
l
c
u
l
a
t
e
d
f
r
o
m
t
h
e
d
a
i
l
y
c
l
o
s
i
n
g
p
r
i
c
e
s
.
A
v
g
S
S
V
o
l
R
a
t
i
o
i
s
a
v
e
r
a
g
e
d
a
i
l
y
s
h
o
r
t
e
d
s
h
a
r
e
s
t
o
d
a
i
l
y
t
r
a
d
e
d
s
h
a
r
e
s
.
D
u
m
m
y
i
s
1
i
f
b
e
f
o
r
e
M
&
A
a
n
n
o
u
n
c
e
m
e
n
t
,
(
2
6
0
,
2
1
)
,
a
n
d
0
i
f
a
f
t
e
r
a
n
n
o
u
n
c
e
m
e
n
t
,
(
0
,
þ
2
0
)
Table VII.
Results of OLS
regression: spreads and
average daily short-sale
ratio around M&A
announcements
JFEP
1,2
194
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
spread. c
2
, the coef?cient of log (volume), is signi?cant in quoted spread and effective
spread. c
3
, the coef?cient of turnover, and c
4
, the coef?cient of volatility, are signi?cant
in quoted spread, percentage quoted spread and effective spread. c
6
, the coef?cient of
Dummy, is signi?cant only in percentage quoted spread. We ?nd that c
5
, the coef?cient
of AvgSSVolRatio, is positive at the 1 percent level for quoted spread and percentage
quoted spread and is negative at the 5 percent level for effective spread. These results
imply that, as the ratio of shorted shares to traded shares increases, the quoted spreads
and percentage quoted spreads of acquiring ?rms increase. However, the effective
spreads of acquiring ?rms decrease as the ratio of shorted shares to traded shares
increases. Therefore, depending on the BAS measures, the empirical evidence support
H4 only for acquiring ?rms.
5. Summary and conclusions
We examine the changes in short-sale transactions of target ?rms and acquiring ?rms
around merger and acquisition (M&A) announcements using daily short-sale
transaction data from the NYSE and NASDAQ. Overall, we conclude the following
results: First, target ?rms experience signi?cant ES from day 2 1 to day þ 7; while
acquiring ?rm experience signi?cant ES from day 0 to day þ 20. Second, for acquiring
?rms, the ?ve day pre-announcement abnormal short sale is negatively related to the
announcement day return and is positively related to the post-announcement return.
We do not observe such a relationship for target ?rms. Third, we cannot conclude that
short sellers are informed traders around M&A announcements. Fourth, for target
?rms, we ?nd that short activity changes are not signi?cantly related to changes in
trading costs. For acquiring ?rms, short activity changes are positively related to
quoted spreads and percentage quoted spreads and are negatively related to effective
spreads. Given the current credit crisis, the SEC adopted a sequence of actions to adjust
the short-sale rule (SEC release no. 34-58166 (July 15, 2008) and 34-58572 (September
17, 2008) and 34-58592 (September 18, 2008)). Our ?ndings suggest that informed short
sellers, if exist, are hard to detect, therefore, any change to the current short-sale
regulation should be careful and speci?c. An overall short sale ban may compromise
market quality rather than effectively preventing abusive short selling.
Further research effort may be devoted to examine the derivative market around
M&A announcements. Since short sale is not the only way to establish short position
and is relatively costly, stock futures and options may be good alternatives for
informed traders to maximize their pro?t.
Notes
1. The SEC adopted regulation SHO to update short-sale regulation in light of numerous
market developments since short-sale regulations were ?rst adopted in 1938. Compliance
with regulation SHO began on January 3, 2005.
2. “Certain market events and trading strategies may make a security more vulnerable to
abusive short-sale activity. Speci?c market events related to an issuer or a security (such as a
pending merger or acquisition) may cause this increased vulnerability. We therefore, request
comment on whether short selling should continue to be regulated or even prohibited during
speci?c market conditions. Are there corporate events (e.g. mergers, acquisitions, or tender
offers) that make a security vulnerable to abusive short selling? Should short selling be
prohibited for a period preceding a signi?cant corporate or market event? If the Rule was
Short sales
around M&A
195
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
eliminated, should restrictions continue to apply preceding a signi?cant corporate or market
event?” excerpt from SEC Release No. 34-42037 (October 28, 1999), 64 FR 57996.
3. Kweon and Pinkerton (1981) and Ascioglu et al. (2002).
4. According to regulation T, the initial margin for short sales is 150 percent. However, the ?rst
100 percent of the requirement can be satis?ed by the proceeds of the short sale, leaving just
50 percent for the investor to maintain in the initial margin (so short selling looks much like
the case of going long). According to Rule 2520, the maintenance margin for short sales is $5
per share or 30 percent of the current market value, whichever is greater.
5. SEC 17 CFR PART 240, Release No. 34-42037; File No. S7-24-99, RIN 3235-AH84 Short Sales.
6. On June 23, 2004, the SEC adopted regulation SHO to establish uniform locate and delivery
requirements, create uniform marking requirements for sales of all equity securities, and to
establish a procedure to temporarily suspend the price-tests for a set of pilot securities
during the period May 2, 2005-April 28, 2006 in order to examine the effectiveness and
necessity of short-sale price-tests. At the same time, the SEC mandated that all self
regulatory organizations make tick-data on short-sales publicly available starting January 2,
2005. On April 20, 2006, the SEC announced that the short-sale Pilot has been extended to
August 6, 2007. The SHO-mandated data includes the ticker, price, volume, time, listing
market, and trader type (exempt or non-exempt from short-sale rules) for all short sales.
References
Aitken, M.J., Frino, A., McCorry, M.S. and Swan, P.L. (1998), “Short sales are almost
instantaneously bad news: evidence from the Australian stock exchange”, Journal of
Finance, Vol. 53, pp. 2205-23.
Ascioglu, A.N., McInish, T.H. and Wood, R.A. (2002), “Merger announcements and trading”,
Journal of Financial Research, Vol. 25, pp. 263-78.
Brent, A., Morse, D. and Stice, E.K. (1990), “Short interest: explanations and tests”, Journal of
Financial and Quantitative Analysis, Vol. 25, pp. 273-89.
Chae, J. (2005), “Trading volume, information asymmetry, and timing information”, Journal of
Finance, Vol. 60, pp. 413-42.
Christophe, S., Ferri, M. and Angel, J. (2004), “Short-selling prior to earnings announcements”,
Journal of Finance, Vol. 59, pp. 1845-75.
Daouk, H. and Charoenrook, A. (2005), “A study of market-wide short selling restrictions”,
working paper, Cornell University, Ithaca, NY.
Dechow, P.M., Hutton, A.P., Meulbroek, L. and Sloan, R.G. (2001), “Short sellers, fundamental
analysis and stock returns”, Journal of ?nancial Economics, Vol. 61, pp. 77-106.
Desai, H., Krishnamurthy, S. and Venkataraman, K. (2006), “Do short sellers target ?rms with
poor earnings quality? Evidence from earnings restatements”, Review of Accounting
Studies, Vol. 11, pp. 71-90.
Desai, H., Ramesh, K., Ramu Thiagarajan, S. and Balachandran, B.V. (2002), “An investigation of
the informational role of short interest in the NASDAQ market”, Journal of Finance,
Vol. 57, pp. 2263-87.
Diamond, D.W. and Verrecchia, R.E. (1987), “Constraints on short-selling and asset price
adjustment to private information”, Journal of Financial Economics, Vol. 18, pp. 277-311.
Dyvbig, P.H. (1984), “Short sales restrictions and kinks on the mean variance frontier”, Journal of
Finance, Vol. 39, pp. 239-44.
JFEP
1,2
196
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Efendi, J., Swanson, E.P. and Chairholder, D. (2009), “Short sellers trading in companies with
accounting irregularities”, working paper, University of Texas, Arlington, TX.
Fama, E. (1970), “Ef?cient capital markets: a review of theory and empirical work”, Journal of
Finance, Vol. 25, pp. 383-417.
Figlewski, S. (1981), “The informational effects of restrictions on short sales: some empirical
evidence”, Journal of Financial and Quantitative Analysis, Vol. 16, pp. 463-76.
Finnerty, J.D. (2005), “Short selling, death spiral convertibles, and the pro?tability of stock
manipulation”, working paper, Fordham University, New York, NY.
Genard, B. and Nanda, V. (1993), “Trading and manipulation around seasoned equity offerings”,
Journal of Finance, Vol. 48, pp. 213-45.
Jarrell, G., Brickley, J. and Netter, J. (1988), “The market for corporate control: the empirical
evidence since 1980”, Journal of Economic Perspective, Vol. 2, pp. 49-68.
Jarrow, R.A. (1980), “Heterogeneous expectations, restrictions on short sales, and equilibrium
asset prices”, Journal of Finance, Vol. 35, pp. 1105-13.
Jensen, M.C. and Ruback, R.S. (1983), “The market for corporate control: the scienti?c evidence”,
Journal of Financial Economics, Vol. 11, pp. 5-55.
Jones, C.M. (2008), “Shorting restrictions: revisiting the 1930s”, working paper, Columbia
Business School, New York, NY.
Kweon, A.J. and Pinkerton, J.M. (1981), “Merger announcements and insider trading activity:
an empirical investigation”, Journal of Finance, Vol. 36, pp. 855-69.
McInish, T.H. and Wood, A. (1992), “An analysis of intraday patterns in bid-ask spreads for
NYSE stocks”, Journal of Finance, Vol. 47, pp. 753-64.
Miller, E.M. (1977), “Risk, uncertainty, and divergence of opinion”, Journal of Finance, Vol. 32,
pp. 1151-68.
Ross, S.A. (1977), “The capital asset pricing model (CAPM), short-sale restrictions and related
issues”, Journal of Finance, Vol. 32, pp. 177-83.
Senchack, A.J. and Starks, L.T. (1993), “Short-sale restrictions and market reaction to
short-interest announcements”, Journal of Financial and Quantitative Analysis, Vol. 28,
pp. 177-94.
White, H. (1980), “A heteroskedasticity-consistent covariance matrix estimator and a direct test
for heteroskedasticity”, Econometrica, Vol. 48, pp. 817-38.
Woolridge, J.R. and Dickson, A. (1994), “Short selling and common stock prices”, Financial
Analysts Journal, Vol. 50, pp. 20-8.
Further reading
Breusch, T.S. and Pagan, A.R. (1979), “A simple test for heteroscedasticity and random
coef?cient variation”, Econometrica, Vol. 47, pp. 1287-94.
Corresponding author
Liuqing Mai can be contacted at: [email protected]
Short sales
around M&A
197
To purchase reprints of this article please e-mail: [email protected]
Or visit our web site for further details: www.emeraldinsight.com/reprints
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_942385255.pdf
The purpose of this paper is to examine changes in short-sale transactions of target firms
and acquiring firms around merger and acquisition (M&A) announcements using daily short-sale
transaction data from the New York stock exchange and NASDAQ. The paper further aims to
investigate the link between short-sale transactions and trading costs.
Journal of Financial Economic Policy
Short sales around M&A announcements
Liuqing Mai Robert van Ness Bonnie van Ness
Article information:
To cite this document:
Liuqing Mai Robert van Ness Bonnie van Ness, (2009),"Short sales around M&A announcements", J ournal
of Financial Economic Policy, Vol. 1 Iss 2 pp. 177 - 197
Permanent link to this document:
http://dx.doi.org/10.1108/17576380911010272
Downloaded on: 24 January 2016, At: 21:35 (PT)
References: this document contains references to 28 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 399 times since 2009*
Users who downloaded this article also downloaded:
J oa˜o A.C. Santos, (2009),"Do markets “discipline” all banks equally?", J ournal of Financial Economic
Policy, Vol. 1 Iss 1 pp. 107-123 http://dx.doi.org/10.1108/17576380910962402
Bruno Funchal, Mateus Clovis, (2009),"Firms' capital structure and the bankruptcy law design", J ournal of
Financial Economic Policy, Vol. 1 Iss 3 pp. 264-275 http://dx.doi.org/10.1108/17576380911041737
Carlos A. Ulibarri, Ionut Florescu, J oel M. Eidsath, (2009),"Regulating noisy short-selling
of troubled firms?", J ournal of Financial Economic Policy, Vol. 1 Iss 3 pp. 227-245 http://
dx.doi.org/10.1108/17576380911041719
Access to this document was granted through an Emerald subscription provided by emerald-srm:115632 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for
Authors service information about how to choose which publication to write for and submission guidelines
are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well as
providing an extensive range of online products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee
on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive
preservation.
*Related content and download information correct at time of download.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Short sales around M&A
announcements
Liuqing Mai
College of Business Administration, University of Missouri,
Saint Louis, Missouri, USA, and
Robert van Ness and Bonnie van Ness
School of Business Administration, University of Mississippi,
Oxford, Mississippi, USA
Abstract
Purpose – The purpose of this paper is to examine changes in short-sale transactions of target ?rms
and acquiring ?rms around merger and acquisition (M&A) announcements using daily short-sale
transaction data from the New York stock exchange and NASDAQ. The paper further aims to
investigate the link between short-sale transactions and trading costs.
Design/methodology/approach – Two abnormal short-sale measures are developed. Two
regression models based on the two short-sale measures are constructed and ordinary least squares
is used to estimate the regressions. Two samples to test bid-ask spreads (BAS) before and after M&A
announcements t-test are used.
Findings – The paper ?nds that target ?rms experience signi?cant excess short sales (ES) from
day 2 1 to day þ 7; while acquiring ?rms experience signi?cant ES from day 0 to day þ 20. For
acquiring ?rms, the ?ve-day pre-announcement abnormal short sale is negatively related to the
announcement day return and is positively related to post-announcement return. Such a relationship
for target ?rms is not observed. For target ?rms, it is found that changes in short activity are not
signi?cantly related to changes in trading cost. For acquiring ?rms, short activity changes are
positively related to quoted spreads and percentage quoted spreads. The short-sale activity changes
are negatively related to effective spreads.
Research limitations/implications – The paper is a ?rst step to understanding whether short
sales affect market liquidity around M&A announcements; therefore restriction is necessary.
Additional research can be done which should extend the current study to include the options market.
Practical implications – From the results, the paper cannot conclude that short sellers are informed
traders around M&A announcements. Therefore restrictions on short sales around M&A
announcements may not be warranted.
Originality/value – The paper ?lls an important blank in the existing literature by examining
short-sale transactions around M&A announcements. Such an investigation is of particular interest to
market regulators as they try to update the short-sale rules.
Keywords Sales, Acquisitions and mergers, Stock exchanges
Paper type Research paper
1. Introduction
In this paper, we examine changes in short-sale transactions of target ?rms and
acquiring ?rms around merger and acquisition (M&A) announcements. Speci?cally,
we test whether the short-sale transactions before M&A announcements are driven
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1757-6385.htm
The authors would like to thank George Christodoulakis (the Editor) and an anonymous referee
for helpful comments and suggestions. All errors are the responsibility of the authors.
Short sales
around M&A
177
Journal of Financial Economic Policy
Vol. 1 No. 2, 2009
pp. 177-197
qEmerald Group Publishing Limited
1757-6385
DOI 10.1108/17576380911010272
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
mainly by informed traders and associated with post-announcement stock price
performance. Further, we investigate the link between short-sale transactions and
trading costs.
Such an investigation is of particular interest to market regulators. The securities and
exchange commission (SEC) notes the need to update its regulations on short sales and
seeks public comments[1]. One of the SEC’s questions has asked is whether speci?c
market events, such as mergers and acquisitions, make a security more vulnerable to
abusive short-sale activity. If so, short selling could be regulated, or even prohibited,
before and during such events[2]. To date, short-sale research has not offered an answer
to this question. Prior studies on short-sales focus on the pricing implications of
restrictions on short sales and the use of short selling to form investment strategies
(Miller, 1977; Ross, 1977; Diamond and Verrecchia, 1987; Jarrow, 1980; Dyvbig, 1984;
Figlewski, 1981). Researchers examine short-sale transactions around earnings
announcements (Woolridge and Dickson, 1994; Christophe et al., 2004), earnings
restatements (Desai et al., 2006), and seasoned equity offerings (SEO) (Genard and
Nanda, 1993). Short-sale transactions around mergers and acquisitions are yet to be
examined. This paper addresses this gap.
Further, this study differs fromprevious studies in that it examines the link between
short sales and trading costs around M&A announcements. One bene?t provided by
short sales is liquidity which can be measured by the bid-ask spreads (BAS) and depth.
However, howshort sales affect liquidity around M&Aannouncements is not clear. This
study analyzes the issue from a market microstructure perspective by examining the
link between short sales and BASs. Understanding how short sales affect market
liquidity is of particular importance to market regulators because of the need to balance
the costs and bene?ts of restrictions.
This study is also of academic interest. Diamond and Verrecchia (1987) argue that
short sellers are sophisticated investors and hypothesize a negative relation between
short selling and future declines in stock price. However, empirical results are mixed.
Senchack and Starks (1993), Aitken et al. (1998), Dechow et al. (2001), and Christophe
et al. (2004) ?nd supporting evidence for this hypothesis. Woolridge and Dickson (1994)
and Brent et al. (1990) do not. We believe that M&A announcements provide a good
environment to study this hypothesis since informed traders typically trade before
M&A announcements (Kweon and Pinkerton, 1981). Given the cost of short selling, the
proportion of informed short sellers around M&A announcements should increase, as
they are motivated by their information advantage. In particular, prior research ?nds
that, around M&A announcements, target ?rms achieve signi?cant positive abnormal
returns; acquiring ?rm returns are mixed depending on the mode of acquisition, the
method of payment, and on average, acquiring ?rms earn zero abnormal returns
(Jensen and Ruback, 1983 and Jarrell et al., 1988). Thus, informed traders who are
aware of the upcoming M&A announcement should refrain from shorting stocks of the
target ?rms as target ?rms are known to have signi?cant positive gains; but short
more shares of the acquiring ?rm according to the deal-speci?c characteristics.
Therefore, M&A announcements offer a unique window to examine whether short
sellers are informed and the relationship between short sales and future stock prices.
Using daily short-sale transaction data, this study examines the relationship between
short sales and future stock prices of target ?rms and acquiring ?rms around merger
and acquisition (M&A) announcements.
JFEP
1,2
178
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
The rest of the paper is organized as follows: In Section 2, we develop testable
hypotheses related to short-sale activities of target ?rms and acquiring ?rms around
M&Aannouncements; inSection3, we describe the dataandmethodology; inSection4, we
report and analyze empirical results from testing the hypotheses; Section 5 concludes.
2. Hypotheses
A short sale is a sale of stock that one does not actually own, but has borrowed from
someone else. A short seller establishes his short position by selling the borrowed stock
and closes his short position by purchasing the stock and returning it to the lender at a
later time. A short seller pro?ts when the stock price falls. While the maximum gain
from selling short is achieved if the stock price falls to zero, the loss from selling short
can be unlimited if the stock price rises. Owing this high risk and potential to
manipulate stock prices, the SEC substantially restricts short selling (see Dechow et al.,
2001 and Finnerty, 2005 for discussion). For example, Rule 105 governs short-sale
activity immediately prior to public security offerings with the aim of safeguarding the
integrity of the capital raising process. No such rule has thus far been put on other
corporate events such as mergers and acquisitions around which informed traders tend
to be prevalent and manipulative short sales are likely. While the SEC realizes this
need, no empirical study on manipulative or informed short-sale activity prior to M&A
announcements has been documented.
Prior studies examine short seller behavior prior to important corporate news
releases. For example, Genard and Nanda (1993) investigate the potential of
manipulative short selling prior to a SEO and predict increased short selling prior to
SEOs. Christophe et al. (2004) ?nd that short sellers are active prior to earnings
announcements. Desai et al. (2006) ?nd similar results prior to earnings restatements.
M&A announcements, which are more unpredictable than other announcements,
such as earnings announcements, in terms of both timing and magnitude, will have
greater information asymmetry, thus, a more substantial impact on stock prices (Chae,
2005). Informed traders, acting strategically, may attempt to maximize their pro?ts
through short selling prior to M&A announcements. Therefore, around M&A
announcements, it is likely that informed short sellers are prevalent in the market. On
one hand, informed short sellers would refrain from taking a short position in a target
?rmstock as they expect the target ?rmstock price will rise. Hence, we expect to observe
a drop in short-sale activity of target ?rm stock prior to M&A announcements. On the
other hand, informed short sellers would short more shares of the acquiring ?rm’s
stock if the expected pro?t is high enough to cover the cost of short selling. Hence, we
expect to see an increase or no change inthe short-sale activity of the acquiring ?rmprior
to M&A announcements. According to the ef?cient market hypothesis (Fama, 1970),
we expect to observe short-sale activity quickly returns to the normal level after
M&A announcements. In this paper, we examine whether short sellers of target ?rms
(acquiring ?rms) cover (establish) their positions before M&A announcements.
Accordingly, we hypothesize the following:
H1. Short-sale activity of target ?rms (acquiring ?rms) decreases (increases or does
not change) before and quickly returns to normal after M&A announcements.
Diamond and Verrecchia (1987) suggest that short sales are mainly driven by informed
traders due to the fact that short selling costs are so high that liquidity traders ?nd it
Short sales
around M&A
179
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
too costly to short. Empirical studies are largely consistent with the notion that short
sellers are a subset of sophisticated investors in the investment community by
providing evidence from various aspects. For example, Dechow et al. (2001) ?nd that
short sellers use ratios of fundamentals (such as earnings and book value) to market
values to identify stocks with lower expected future returns. Efendi et al. (2009) argue
that short sellers are highly sophisticated investors who can see through accounting
manipulation and thus pro?t from their knowledge. Prior studies also investigate
whether or not short sellers are informed traders through observing short seller
behavior prior to important corporate news releases. Christophe et al. (2004) provide
evidence of informed trading in the ?ve days prior to earnings announcements of
NASDAQ-listed ?rms. However, little is known about informed short selling prior to
M&A announcements.
Previous market microstructure literature on M&A announcements documents
substantial evidence that informed trading increases before M&A announcement[3].
Prior to M&A announcements, informed traders will seek to maximize their pro?ts
based on their private information knowing that the information might be revealed or
detected by the market, thus, lose its value. Short sales, which allow investors to trade
on margin, can be a valuable venue for informed traders to maximize their pro?ts
because a margin account can magnify pro?ts[4]. In this paper, we aim to provide
evidence that information-driven short sales are prevalent before M&A
announcements. We test the following hypothesis:
H2. Short-sale activity before M&A announcements is mainly driven by informed
trades.
The relationship between short sales and subsequent stock returns is of great interest in
the short-sale literature. Figlewski (1981) ?nds that stocks with higher short interest
under-perform in subsequent months. Desai et al. (2002) ?nd that heavily shorted ?rms
experience signi?cant negative abnormal returns. Senchack and Starks (1993) report
abnormal returns around monthly short interest announcements are more negative the
higher the unexpected short interest. Similarly, using data from Australian stock
market, where short sales are made public immediately upon occurrence, Aitken et al.
(1998) ?nd an immediate drop in stock price after short-sale executions. Christophe et al.
(2004) ?nd that abnormally high short sales are linked to abnormally low subsequent
stock returns around earnings announcements. M&A announcements are similar to
other corporate announcements in that they result in information asymmetry in the
market, to varying degrees. Therefore, one would expect that changes in short-sale
activity prior to M&Aannouncements would be associated with both the announcement
and post-announcement stock returns. Speci?cally, a decrease in the short-sale activity
of a target ?rm’s stock prior to an M&A announcement will be associated with an
increase in both the announcement and post-announcement target ?rm’s stock return.
Accordingly, an increase or no change in short-sale activity of acquiring ?rm’s stock is
expected to be associated with a decrease or no change in both the announcement and the
post-announcement stock returns. We develop the following hypothesis:
H3. Changes in short-sale activity before M&Aannouncements are associated with
stock returns on the announcement date and those in the post-announcement
period.
JFEP
1,2
180
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Substantial market liquidity is provided through short selling by market professionals, such
as market makers, block positioners, and specialists, [. . .] To the extent that short sales are
affected in the market by securities professionals, such short-sale activities, in effect, add to
the trading supply of stock available to purchasers and reduce the risk that the price paid by
investors is arti?cially high because of a temporary contraction of supply. – SEC Concept
Release: Short Sales[5].
Although the SEC statement assumes an explicit link between short sales and
liquidity, empirical studies on this issue are limited. Jones (2008) studies changes in
liquidity around events that alter the level of short-sale constraints in the US market.
He ?nds that the introduction of the requirement that brokers secure written
authorization before lending a customer’s shares in 1932 had a negative impact on
liquidity, but the requirement that short sales be executed on an uptick in 1938 had a
positive effect on liquidity. Daouk and Charoenrook (2005) examine short sales from
111 countries and ?nd that, when short selling is possible, there is greater liquidity, as
measured by turnover ratio.
In this paper, we reexamine the link between short sales and trading costs,
measured by the BAS, using an event study of M&A announcements. Prior to M&A
announcements, we expect to observe a decrease in trading costs of target ?rms as
short-sale activity increases. After M&A announcements, the trading costs of both
target ?rms and acquiring ?rms are expected to return to a normal level quickly. The
following hypothesis is formed:
H4. Around M&A announcements, short-sale activity changes are associated
with changes in trading costs.
3. Data and methodology
3.1 Data
We obtain the sample of mergers and acquisitions from the securities data company
database. It includes mergers and acquisitions announced from January 1, 2005 to
December 31, 2005. We select the sample by the following criteria:
Acquiring ?rms must:
.
be listed on New York Stock Exchange (NYSE) or NASDAQ;
.
not use the following deal types: leveraged buyout, spinoff, self-tender,
recapitalization, exchange offer, repurchase, or privatization;
.
have a M&A deal value of one million or more;
.
have a closing price of $5 or more four weeks prior to announcement; and
.
not have any missing data during the sample period.
These criteria yield 1,010 acquiring ?rms.
Target ?rms must:
.
be listed on NYSE or NASDAQ;
.
not use the following deal types: leveraged buyout, spinoff, self-tender,
recapitalization, exchange offer, repurchase, or privatization; and
.
not have any missing data during the sample period.
Short sales
around M&A
181
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
These criteria yield 196 target ?rms. The trading price and quote data are obtained
from trade and quote. The short-sale data are from SEC regulation SHO-mandated
data[6]. Our sample comprises 252 trading days in year 2005.
Table I presents summary information for the sample ?rms. Average ?rm size for
the target ?rms is $3,112.2 million; for the acquiring ?rms is $10,390 million. Average
closing price for the target ?rms is $19.87; for the acquiring ?rms is $28.93. In our
sample, the acquiring ?rms are generally bigger and trade at higher prices than the
target ?rms. The acquiring ?rms are also more actively traded, as indicated by the
mean daily shares traded of 1,151,430 for acquiring ?rms and 534,910 for target ?rms.
The mean daily shorted shares for the acquiring ?rms is 248,590, while for target ?rms
is 110,370. The mean short-sale ratio (the average shorted shares to average total
shares traded) for the target ?rms is 17.93 percent, and for the acquiring ?rms is
24.49 percent. The maximum short-sale ratio exceeds 40 percent for target ?rms and
nearly 50 percent for acquiring ?rms. These ?gures imply that acquiring ?rms have
more short sales than target ?rms.
3.2 Calculation of abnormal short sales
We calculate the daily average cumulative excess short sales (ES) and the daily average
ES to examine the short-sale pattern of the sample ?rms. Let S
f,t
equal ln(1 þ daily
shorted sales) for day t for ?rm f and S
f
equal the mean of S
f,t
for days 2 60 to 221.
For ?rm f, ES is:
ES
f ;t
¼ S
f ;t
2S
f
t ¼ 260· · · þ 20 ð1Þ
We calculate the cumulative excess short sales (CES) on day t for ?rm f:
CES
f ;t
¼ ES
f ;t
þ CES
f ;t21
t ¼ 260· · · þ 20 ð2Þ
Mean Median Maximum Minimum
Panel A: target
Firm size ($ million) 3,112.20 397.31 103,904.00 1.02
Average price ($) 19.87 15.71 70.55 0.08
Average daily shorted shares (000’s shares) 110.37 29.26 2,745.99 0.30
Average daily volume (000’s shares) 534.91 165.15 12,569.12 0.49
Average shorted shares to average total shares (%) 17.93 17.64 42.39 0.71
Panel B: acquirer
Firm size ($ million) 10,390.00 1,297.30 369,166.00 5.61
Average price ($) 28.93 25.00 86.09 3.68
Average daily shorted shares (000’s shares) 248.59 75.17 10,238.22 0.14
Average daily volume (000’s shares) 1,151.43 280.88 60,596.71 1.04
Average shorted shares to average total shares (%) 24.49 23.01 49.18 1.48
Notes: This table contains key characteristics of all ?rms. Firm size is a stock’s market capitalization
calculated as the number of shares outstanding multiplied by price per share. Average price is the
average daily closing price. The average daily shorted shares is a stock’s average daily number of
shares shorted over the 252 trading days in year 2005. A stock’s average daily volume is the average
daily number of shares traded over the 252 day period. A stock’s average shorted shares as a
percentage of total shares is the average daily shorted shares/average daily number of share traded.
The sample contains 192 target ?rms and 1,010 acquiring ?rms that are listed on NYSE or NASDAQ
in year 2005
Table I.
Key characteristics
of involving ?rms
JFEP
1,2
182
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
where CES
f,260
¼ ES
f,260
, day t is relative to the day of the M&A announcement, and
day 0 is the announcement day. If there are unusual short-sale patterns, the mean of
CES
f,t
across ?rms will be signi?cantly greater than zero. We use a one-tailed t-test to
test the null hypothesis that the mean CES for day t is greater than zero. To examine if
there is an abnormal increase or decrease in short-sale activity around M&A
announcements, we use a two-tailed t-test to evaluate the null hypothesis that the mean
ES for day t is different from zero.
3.3 Regression model speci?cation
Following Christophe et al. (2004), we develop two regression models. In the ?rst
model, we de?ne abnormal short sales as the percentage difference between the
average daily number of the ?rm’s shares sold short during the ?ve days preceding the
M&A announcement and the average daily number of the ?rm’s shares sold short
during the non-announcement period. In the second model, we de?ne the relative short
sales in the pre-announcement period as the ratio of shorted shares to traded shares for
the stock from day 25 to 21. Implicit in the two regression models is the assumption
that the average daily short sales during the 41 pre-announcement day period, 260 to
221, is a fair representation of the ?rm’s typical daily short sales. More formally, a
stock’s abnormal short sales during the ?ve days prior to the M&A announcement
ABSS(25, 21), is measured as:
ABSSð25; 21Þ ¼ SSð25; 21Þ=AVESSð260; 221Þ 21 ð3Þ
where:
SS (25, 21) The average daily number of shorted shares for the ?ve days
prior to the M&A announcement.
AVESS(260, 221) The average daily number of shorted shares during the
non-announcement period.
To assess the robustness of our estimation, we specify RELSS(25, 21) as the relative
short sales of a ?rm in the pre-announcement period. RELSS(25, 21) is measured
as the ratio of shorted shares to traded shares for the stock over the interval of days
25 to 21. We test the following regression models:
ABSSð25; 21Þ ¼a
0
þ a
1
RETð0; þ1Þ þ a
2
RETð25; 21Þ þ a
3
RETðþ2; þ5Þ
þ a
4
ABVOLð25; 21Þ þ e
ð4Þ
RELSSð25; 21Þ ¼y
0
þ y
1
RETð0; þ1Þ þ y
2
RETð25; 21Þ þ y
3
RETðþ2; þ5Þ
þ y
4
NORMRELSS þ e
ð5Þ
where:
RET(0, þ 1) The event return on the stock calculated from the closing
prices of days 21 to þ 1, RET(25, 21) is the return on the
stock calculated from the closing prices of days 26 to 21.
ABVOL (25, 21) The average daily abnormal volume in the stock over the
interval of day 25 to 21.
Short sales
around M&A
183
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
The independent variable RET(0, þ 1) serves as a proxy for the announcement surprise.
Thus, a negative two-day return means that the market views the announcement as an
unfavorable (negative) surprise and a positive return means the announcement is more
encouraging (positive) than investors expect. Therefore, a negative a
1
means that short
sales regularly rise prior to disappointing M&A announcements and decrease prior to
favorable announcements. The model contains two control variables, RET(25, 21) and
ABVOL(25, 21). RET(25, 21) represents the movement of the stock price during the
?ve days prior to the announcement. This variable controls for the possibility that
changes in stock prices might affect the level of short sales in the days leading up to the
announcement. A pre-announcement increase in stock price, for example, might affect
short sales by enticing some investors to short the now “over-valued” stock. With this
control variable in place, the model will not incorrectly attribute all pre-announcement
short sales to expectations regarding the announcement. The second control variable,
ABVOL(25, 21), accounts for the potential contemporaneous correlation between
abnormal short sales and spikes in volume, and for the possibility that stocks
experiencing sudden increases in volume might be easier to short. Abnormal volume is
measured as the percentage difference between the average volume in the ?ve-day event
interval, and the average daily volume in the 41 days of the non-announcement period.
RET(þ2, þ 5) is the post-announcement return calculated from the closing prices of
day þ 1 to day þ 5. NORMRELSS, normalized relative short sales, is calculated as the
ratio of the shorted shares to traded shares for a stock during the non-announcement
period, day 260 to 221. The error term, e, captures the combined effects of omitted
variables.
3.4 BASs: calculation and regression model
We calculate four BAS measures: quoted spread, percentage quoted spread, effective
spread, and percentage effective spread as follows:
Quoted spread
it
¼ Ask
it
2Bid
it
Percentage quoted spread
it
¼ 2ðAsk
it
2Bid
it
Þ=ðAsk
it
þ Bid
it
Þ
Effective spread
it
¼ jPrice
it
2ðAsk
it
þ Bid
it
Þ=2j
Percentage effective spread
it
¼ 100ðjPrice
it
2ðAsk
it
þ Bid
it
Þ=2jÞ=ððAsk
it
þ Bid
it
Þ=2Þ
Quotes before the opening and after closing are excluded from the sample. To create
second-by-second data, each quote is carried forward until the next one arrives.
Because we retain each quote until it is updated, our spread measures are time
weighted. Quotes with longer lives have more weight in calculating the average daily
spreads. For each ?rm, a series of time-weighted percentage BAS is created by
averaging second-by-second quotations for each day during the 252 days of the sample
period. Then, for each stock, we average the time-weighted BAS for each day. For each
day, we average the time-weighted spreads across ?rms.
To examine the linkage of the short sale changes and trading cost changes around
M&A announcements, we apply multivariate regressions. McInish and Wood (1992)
suggest that spread is determined by price, activity, volatility, and volume. Here, we
add two short-sale measures to the regressions: the daily average short sales (DailySS)
JFEP
1,2
184
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
and the ratio of average short sales to volume (AvgSSVOLRatio). We estimate the
following regressions over the sample period (260, þ 20):
Spread ¼b
0
þ b
1
ð1=PRICEÞ þ b
2
logðVOLUMEÞ þ b
3
TURNOVER
þ b
4
VOLATILITY þ b
5
DailySS þ b
6
Dummy þ e
ð6Þ
and:
Spread ¼c
0
þ c
1
ð1=PRICEÞ þ c
2
logðVOLUMEÞ þ c
3
TURNOVER
þ c
4
VOLATILITY þ c
5
AvgSSVolRatio þ c
6
Dummy þ e
ð7Þ
Spread is the time-weighted spread measures speci?ed above. Price is the mean value
of daily closing prices. Volume is the dollar trading volume. Turnover is the ratio of the
number of shares traded to number of shares outstanding. Volatility is the standard
deviation of daily return calculated from the daily closing prices. DailySS is the
average daily short sale. AvgSSVolRatio is average daily shorted shares to daily
traded shares. Dummy is 1 if before M&A announcement, (260, 21), and 0 if after
announcement, (0, þ 20). The error term, e, captures the combined effects of omitted
variables. Both White’s test (White, 1980) and Breusch-Pagan test are performed to test
the existence of heteroscedasticity. Neither of the two tests is signi?cant at 5 percent
level. We estimate the regression models using ordinary least squares (OLS).
4. Empirical results
4.1 Hypothesis 1
Table II provides the CES and ES for days 260 to þ20 surrounding the M&A
announcements.
For target ?rms, CES becomes positive and statistically signi?cant at 5 percent level
on day 0. The t-statistics for CES are negative and statistically signi?cant at least at
the 10 percent level for each day from day 2 30 to day 2 21. This evidence con?rms
that there is a decrease in short-sale activity prior to the M&A announcement. More
importantly, this evidence shows that a signi?cant decrease in short-sale activity
occurs in a period from day 2 30 to day 2 21, which is not immediately prior to the
M&A announcement. We do not ?nd evidence that short-sale activity decreases
immediately before the M&A announcement day. After the M&A announcement, the
t-statistics for CES are positive and statistically greater than zero at the 1 percent level
from day þ 1 to day þ 20. This evidence shows that there is an increase in short-sale
activity after the M&A announcement. It does not support H1. However, we observe a
turn in short-sale activity after the M&A announcement. The value of CES ?rst
increases from day þ 1 to day þ 13, and then decreases consistently from day þ 14 to
day þ 20. ES becomes signi?cantly positive for each day from day 2 1 to day þ 7 and
is not signi?cant after day þ 7. Therefore, this result implies that short-sale activity
slowly returns to normal after the M&A announcement.
For acquiring ?rms, CES is negative and signi?cantly from day 2 60 to day 0. This
evidence shows that short-sale activity of acquiring ?rms decreases prior to the M&A
announcement. It does not support H1. CES is not signi?cant from day þ 1 to day þ 6,
and then becomes positive and signi?cantly greater than zero for each day from
day þ 7 to day þ 20. The ES of acquiring ?rms is positive and signi?cantly different
from zero for each day from day 0 to day 20. This shows that there are excess short
sales after the M&A announcement day. While the CES of target ?rms shows a
Short sales
around M&A
185
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
CES ES
Day relative Target t-statistics Acquirer t-statistics Target t-statistics Acquirer t-statistics
260 0.285 1.13 21.660 23.76
* * *
20.040 20.33 20.113 22.44
* * *
230 20.817 21.32
*
23.051 24.54
* * *
20.092 20.83 0.042 0.88
229 21.066 21.81
* *
23.012 24.47
* * *
20.248 21.97
* *
0.039 0.94
227 20.956 21.65
* *
22.996 24.45
* * *
0.087 0.76 20.060 21.10
226 20.915 21.65
* *
22.984 24.41
* * *
0.041 0.38 0.012 0.25
224 20.802 21.57
*
22.857 24.19
* * *
0.006 0.05 0.111 2.47
* * *
223 20.870 21.71
* *
22.815 24.11
* * *
20.068 20.51 0.041 0.83
222 20.845 21.73
* *
22.765 24.02
* * *
0.024 0.21 0.050 1.06
221 20.661 21.41
*
22.677 23.85
* * *
0.184 1.45 0.088 1.98
* *
220 20.621 21.27 22.611 23.76
* * *
0.040 0.32 0.066 1.35
219 20.631 21.25 22.532 23.64
* * *
20.010 20.08 0.080 1.83
*
218 20.667 21.25 22.515 23.61
* * *
20.036 20.29 0.016 0.33
217 20.677 21.19 22.394 23.43
* * *
20.010 20.08 0.120 3.17
* * *
216 20.725 21.17 22.321 23.32
* * *
20.048 20.35 0.074 1.78
* *
215 20.645 20.95 22.335 23.33
* * *
0.083 0.56 20.014 20.29
214 20.430 20.60 22.240 23.18
* * *
0.212 1.93
*
0.095 2.30
* *
213 20.599 20.78 22.152 23.05
* * *
20.170 21.06 0.088 1.93
*
212 20.601 20.73 21.979 22.81
* * *
20.002 20.01 0.173 4.23
* * *
211 20.542 20.63 21.910 22.71
* * *
0.059 0.53 0.069 1.58
210 20.460 20.51 21.859 22.63
* * *
0.082 0.64 0.051 1.12
29 20.250 20.26 21.785 22.52
* * *
0.209 1.61 0.074 1.76
* *
28 0.030 0.03 21.735 22.45
* * *
0.281 1.81
*
0.050 0.96
27 0.210 0.19 21.697 22.38
* * *
0.108 1.26 0.039 0.85
26 0.298 0.27 21.596 22.23
* * *
0.088 0.62 0.101 2.25
* *
25 0.496 0.42 21.583 22.19
* *
0.197 1.41 0.013 0.27
24 0.642 0.51 21.571 22.14
* *
0.147 0.96 0.012 0.23
23 1.094 0.83 21.410 21.91
* *
0.451 3.18
* * *
0.161 3.56
* * *
22 1.310 0.94 21.379 21.85
* *
0.217 1.33 0.031 0.64
21 1.654 1.12 21.333 21.77
* *
0.343 2.31
* *
0.045 0.93
0 3.127 2.02
* *
21.007 21.32
*
1.473 7.91
* * *
0.327 6.85
* * *
1 5.202 3.17
* * *
20.511 20.67 2.075 11.64
* * *
0.495 11.08
* * *
2 6.432 3.75
* * *
20.101 20.13 1.231 7.03
* * *
0.410 9.79
* * *
3 7.416 4.12
* * *
0.235 0.30 0.984 6.27
* * *
0.337 7.69
* * *
4 8.132 4.28
* * *
0.546 0.69 0.715 4.18
* * *
0.311 6.77
* * *
5 8.516 4.26
* * *
0.738 0.93 0.384 2.06
* *
0.192 3.86
* * *
6 8.887 4.26
* * *
1.007 1.25 0.371 2.17
* *
0.268 5.72
* * *
7 9.248 4.22
* * *
1.228 1.51
*
0.361 2.11
* *
0.221 4.43
* * *
8 9.420 4.09
* * *
1.452 1.76
* *
0.173 0.93 0.224 4.97
* * *
9 9.608 3.99
* * *
1.604 1.92
* *
20.083 1.15 0.152 2.98
* * *
10 9.813 3.87
* * *
1.799 2.12
* *
0.205 1.11 0.195 4.26
* * *
11 9.904 3.77
* * *
2.020 2.34
* * *
0.091 0.59 0.221 5.35
* * *
12 9.867 3.60
* * *
2.226 2.54
* * *
20.037 20.22 0.207 4.42
* * *
13 10.032 3.51
* * *
2.387 2.68
* * *
0.165 0.97 0.161 3.22
* * *
14 9.805 3.30
* * *
2.595 2.86
* * *
20.228 21.32 0.208 4.50
* * *
15 9.673 3.14
* * *
2.754 2.98
* * *
20.132 20.74 0.159 3.33
* * *
16 9.513 2.98
* * *
2.941 3.14
* * *
20.160 20.92 0.187 3.93
* * *
17 9.492 2.86
* * *
3.174 3.33
* * *
20.022 20.13 0.233 5.42
* * *
18 9.429 2.75
* * *
3.359 3.47
* * *
20.063 20.40 0.185 4.06
* * *
19 9.197 2.60
* * *
3.542 3.60
* * *
20.231 21.38
*
0.182 4.04
* * *
20 9.073 2.48
* * *
3.715 3.71
* * *
20.125 20.75 0.173 3.68
* * *
Notes: Statistically signi?cant at
*
10%,
* *
5%, and
* * *
1% levels; this table presents the daily average CES and
daily average ES of involving ?rms. We let S
f,t
equal Ln(1 þ Daily Shorted Sales) for day t for ?rm f and S
f
equal
the mean of S
f,t
for days 260 to 221. For ?rm f, ES is ES
f,t
¼ S
f,t
- S
f
for t ¼ 260. . . þ 20. Then, for day t for ?rm
f, we calculate the CES: CES
f,t
¼ ES
f,t
þ CES
f,t21
t ¼ 260. . . þ 20, where CES
f,260
¼ ES
f,260
, day t is relative
to the day of the M&A announcement, and day 0 is the announcement day
Table II.
Daily average CES
and daily average ES
JFEP
1,2
186
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
decrease in value from day þ 14 to day þ 20, the CES of acquiring ?rms continues to
increase after the M&A announcement day. Therefore, the short-sale activity of
acquiring ?rms does not quickly return to normal after M&A announcements. This
evidence does not support H1 either. Overall, we ?nd that the empirical evidence
weakly supports H1 about target ?rms but does not support H1 about acquiring ?rms.
4.2 Hypothesis 2
H2 states that short-sale activity before M&A announcements are driven by informed
traders. Since informed short sellers cannot be identi?ed directly, we need to show that
there are changes in the overall level of short-sale activity before M&A
announcements, and the changes in the level of trading before the announcement is
a result of informed trading.
Table II reports the average ES of both target ?rms and acquiring ?rms. Before the
M&A announcements, we observe an increase in short activity of target ?rms on
day 2 14, day 2 8, day 2 3, and day 2 1 and an increase in the short activity of
acquiring ?rms on days 214 to 212, day 2 9, day 2 6, and day 2 3.
Table III shows a two sample t-test of BAS before and after M&A announcements.
For target ?rms, all four spread measures show a signi?cant decrease after the M&A
announcements. For acquiring ?rms, quote spread and percentage quoted spread show
a signi?cant decrease after the announcements. Effective spread increases and
percentage effective spread does not change after announcements. Overall, we ?nd a
narrowing of BAS of target ?rms. How the BAS of acquiring ?rms change depends on
the spread measure used.
In summary, we ?nd ES on some days before merger announcements for both the
target ?rms and the acquiring ?rms. We do not ?nd consistent evidence of excess short
sale across days. We ?nd a narrowing of the BAS for target ?rms. The BAS of
acquiring ?rms show mixed results. Therefore, we cannot conclude that short-sale
activity before M&A announcements are driven by informed traders.
Pre mean Post mean Difference t-statistic
Target
Quoted spread 0.1240
* * *
0.0920 0.0320 16.15
Quoted spread (%) 0.0140
* * *
0.0106 0.0034 6.11
Effective spread 0.0637
* * *
0.0474 0.0163 9.90
Effective spread (%) 0.0081
* * *
0.0060 0.0021 4.40
Acquirer
Quoted spread 0.0956
* * *
0.0849 0.0107 2.93
Quoted spread (%) 0.0044
* * *
0.0041 0.0003 5.02
Effective spread 0.0606
* *
0.0697 20.0091 22.47
Effective spread (%) 0.0029 0.0028 0.0001 0.37
Notes: Statistically signi?cant at
*
10%,
* *
5%, and
* * *
1% levels; pre mean is calculated as the
average BAS before the M&A announcement from day 2 60 to day 2 1. Post mean is calculated as
the mean BAS after the M&A announcement from day 0 to day þ 20. Two-tailed test is used to test the
null hypothesis that pre mean 2 post mean ¼ 0
Table III.
Two sample t-test of BAS
Short sales
around M&A
187
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
4.3 Hypothesis 3
To test H3, we estimate equations (4) and (5). We conjecture that the transaction ratio
may affect short-sale activity around M&A announcement as an M&A transaction
may have a greater impact on smaller ?rms than on larger ?rms. Also the changes in
short activity in response to the M&A announcement may vary among thinly traded
?rms and actively traded ?rms. Therefore, we divide the sample ?rms into quintiles
according to the transaction ratio of the M&A deal value to ?rm value and the average
trading volume, respectively. Table IV reports the results of the regression of the full
sample and of quintiles determined by the transaction ratio. The quintiles range from
the smallest transaction ratio to the largest transaction ratio. Table V reports the
results of quintiles determined by daily average trading volume. The quintiles range
from the least trading volume to the most trading volume.
From Panel A in Table IV, the estimates of the coef?cients of equations (4) and (5)
for target ?rms are not signi?cant except for ABVOL (25, 21) and NORMRELSS.
That is, when target ?rms are tested as a whole sample, we do not observe a signi?cant
relationship between abnormal short sales and the announcement day return or the
post-announcement return. For acquiring ?rms, the estimates of the coef?cients of
equation (5) are not signi?cant except for NORMRELSS. However, results of equation
(4) show that ABSS (25, 21) is negatively related to RET (0, þ 1), and positively
related to RET (þ2, þ 5). This result provides some evidence that the market
generally views the M&A announcement as an unfavorable surprise for acquiring
?rms and that the announcement day return of acquiring ?rms is negative. The
estimated a
3
, the coef?cient of RET (þ2, þ 5), is signi?cantly positive. This implies
that short sellers of acquiring ?rms generally earn a positive post announcement
return. Therefore, using the pooled target sample and the pooled acquirer sample, H3 is
true only for acquiring ?rms, not for target ?rms.
Panel B in Table IV presents the regression results of target ?rms, by transaction
ratio quintile. We ?nd that for Quintile 1, both a
2
, the coef?cient of RET (25, 21), and
a
3
, the coef?cient of RET (þ2, þ 5), are signi?cantly positive; for Quintile 2, only a
3
is
signi?cantly positive; for other quintiles, the estimates of coef?cients are not
statistically signi?cant. These results are contradictory to our conjecture. These results
show that, for target ?rms with smallest M&A transaction ratio, short sellers earn
positive returns both before and after M&A announcements. For target ?rms in
Quintile 2, short sellers only earn positive post announcement returns. This anomaly is
yet to be explained. Using equation (5), we do not observe signi?cant relation between
the ratio of shorted shares to traded shares, and investor returns before, on, or after
M&A announcements.
Acquiring ?rm results by transaction ratio are in Panel C of Table IV. Equation (4)
results show that for Quintile 1, a
2
, the coef?cient of RET (25, 21), is signi?cantly
positive; for Quintile 3, a
3
, the coef?cient of RET (þ2, þ 5), is signi?cantly positive; for
Quintile 4, a
1
, the coef?cient of RET (0, þ 1), is signi?cantly negative. Similar to the
results of target ?rms, but again contradictory to our conjecture, short sellers generally
earn positive pre-announcement returns for ?rms with the smallest transaction ratio in
our sample. For acquiring ?rms with an average transaction ratio, abnormal short
activity is positively related to post-announcement return. For acquiring ?rms with
larger transaction ratios (Quintile 4), short activity is associated with a negative
announcement day return. Using equation (5), we ?nd that y
2
, the coef?cient of RET
JFEP
1,2
188
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Panel A: regression results of full sample
Equation (4)
a
0
a
1
a
2
a
3
a
4
Adjusted R
2
Target 0.172
*
1.232 0.715 20.035 0.899
* * *
0.4360
(n ¼ 192) (1.80) (0.56) (0.63) (20.42) (12.23)
Acquirer 0.075
*
22.480
* * *
0.225 0.846
* * *
1.749
* * *
0.5312
(n ¼ 1,010) (1.71) (22.83) (0.35) (4.55) (32.48)
Equation (5)
y
0
y
1
y
2
y
3
y
4
Adjusted R
2
Target 0.085
* * *
20.045 20.005 20.005 0.642
* * *
0.2285
(n ¼ 192) (3.91) (20.23) (20.05) (20.62) (7.14)
Acquirer 0.083
* * *
20.020 0.038 0.014 0.682
* * *
0.2850
(n ¼ 1,010) (9.57) (20.35) (20.88) (1.13) (19.96)
Panel B: regression results of target ?rms by transaction ratio quintiles
Equation (4)
Target a
0
a
1
a
2
a
3
a
4
Adjusted R
2
Quintile 1 0.250 2.650 5.537
* *
9.537
* *
0.994
* * *
0.6102
(n ¼ 38) (1.03) (0.66) (2.15) (2.19) (6.16)
Quintile 2 0.185 3.895 27.050 1.750
* *
3.095
* * *
0.8888
(n ¼ 38) (1.10) (0.57) (21.58) (2.00) (14.79)
Quintile 3 20.131 0.655 1.023 20.682 1.169
* * *
0.7325
(n ¼ 38) (21.55) (0.40) (0.59) (20.92) (8.63)
Quintile 4 0.277 21.637 5.364 22.394 0.731
* * *
0.5513
(n ¼ 40) (0.91) (0.33) (0.43) (20.22) (5.86)
Quintile 5 0.326 20.236 21.477 0.503 0.681
* * *
0.1303
(n ¼ 40) (1.13) (20.02) (20.32) (0.28) (1.52)
Equation (5)
Target y
0
y
1
y
2
y
3
y
4
Adjusted R
2
Quintile 1 0.084 20.284 20.169 0.602 0.705
* *
0.1000
(n ¼ 38) (1.02) (20.43) (0.44) (0.86) (2.33)
Quintile 2 0.012 20.552 20.160 0.178 1.058
* * *
0.4671
(n ¼ 38) (0.21) (20.59) (20.25) (1.47) (4.88)
Quintile 3 0.074 0.164 0.281 0.054 0.555
* * *
0.2393
(n ¼ 38) (2.42) (20.66) (0.33) (0.23) (2.35)
Quintile 4 0.214
* *
0.081 20.593 0.581 0.164 0.0300
(n ¼ 40) (0.91) (0.33) (0.43) (20.22) (5.86)
Quintile 5 0.144
* * *
0.105 20.270 0.029 0.315 0.1240
(n ¼ 40) (2.89) (0.15) (20.98) (0.26) (1.20)
Panel C: regression results of acquiring ?rms by transaction ratio quintiles
Equation (4)
Acquirer a
0
a
1
a
2
a
3
a
4
Adjusted R
2
Quintile 1 0.019 20.450 1.320
* * *
20.059 0.960
* * *
0.5939
(n ¼ 197) (0.74) (20.76) (2.36) (20.22) (16.86)
Quintile 2 0.014 20.078 20.073 0.058 0.737
* * *
0.4466
(n ¼ 197) (0.54) (20.08) (20.30) (0.31) (12.71)
Quintile 3 0.196 24.868 21.159 4.841
* * *
2.029
* * *
0.6564
(n ¼ 197) (1.10) (20.98) (20.32) (5.77) (14.84)
Quintile 4 0.093 214.276
* * *
20.971 20.012 0.960
* * *
0.4445
(n ¼ 197) (1.49) (29.94) (20.72) (20.09) (8.00)
Quintile 5 0.108
*
0.196 0.616 20.787 1.279
* * *
0.6602
(n ¼ 200) (1.82) (0.26) (0.73) (20.87) (19.72)
(continued)
Table IV.
Results of OLS
regression: abnormal
short sales and abnormal
relative short sales
around M&A
announcements
Short sales
around M&A
189
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
(25, 21), is signi?cantly positive in Quintile 1 and y
3
, the coef?cient of RET (þ2, þ 5)
is signi?cantly positive in Quintile 3. These results are similar to those of equation (4).
Sample ?rms are also divided into quintiles by average trading volume and the
regression results are in Table V. We do not ?nd evidence of differences between
actively traded target ?rms and thinly traded target ?rms. We do ?nd that y
3
, the
coef?cient of RET (þ2, þ 5), is signi?cantly positive for Quintile 3. This indicates that
the ratio of shorted shares to traded shares is positively related to the
post-announcement return for averagely traded target ?rms. For acquiring ?rms, we
?nd that a
3
, the coef?cient of RET (þ2, þ 5), is signi?cantly positive for Quintile 1 and
a
1
, the coef?cient of RET (0, þ 1), is signi?cantly negative for Quintile 3. Also, y
3
, the
coef?cient of RET (þ2, þ 5), is signi?cantly negative and for Quintile 2 and
signi?cantly positive for Quintile 5. These results show that abnormal short-sale
activity before the M&A announcement is positively related to the post-announcement
return for the least traded acquiring ?rms and is negatively related the announcement
day return for averagely traded acquiring ?rms. The ratio of shorted shares to traded
shares tends to be negatively related to the post-announcement return for acquiring
Equation (5)
Acquirer y
0
y
1
y
2
y
3
y
4
Adjusted R
2
Quintile 1 0.063
* * *
20.066 0.250
* *
20.030 0.747
* * *
0.3695
(n ¼ 197) (3.45) (20.54) (2.18) (20.56) (10.56)
Quintile 2 0.096
* * *
0.026 0.016 0.005 0.619
* * *
0.3061
(n ¼ 197) (5.50) (0.12) (0.29) (0.13) (9.44)
Quintile 3 0.073
* * *
20.076 0.021 0.129
* * *
0.744
* * *
0.2684
(n ¼ 197) (3.12) (20.35) (0.13) (3.76) (8.17)
Quintile 4 0.081
* * *
20.002 20.213 20.008 0.701
* * *
0.2854
(n ¼ 197) (3.98) (20.01) (21.59) (20.57) (8.77)
Quintile 5 0.094
* * *
0.015 0.104 20.103 0.639
* * *
0.2259
(n ¼ 200) (4.70) (0.14) (0.91) (20.84) (7.70)
Notes: Statistically signi?cant at
*
10%,
* *
5%, and
* * *
1% levels
ABSSð25;21Þ ¼a
0
þa
1
RETð0;þ1Þþa
2
RETð25;21Þþa
3
RETðþ2;þ5Þþa
4
ABVOLð25;21Þþe ð4Þ
RELSSð25;21Þ ¼y
0
þy
1
RETð0;þ1Þþy
2
RETð25;21Þþy
3
RETðþ2;þ5Þþy
4
NORMRELSSþe ð5Þ
The results of OLS estimation of these equations, as ?tted to the full sample and sub-samples
determined by transaction ratio of the M&A deal value to ?rm value are shown. ABSS(25, 2 1) is the
average daily abnormal short sales for stock in the pre-announcement period, measured as the average
daily short sale in the pre-announcement period divided by the average daily short sale in the non-
announcement period, all 21. RELSS(25, 21) is the ratio of shorted shares to traded shares in the
stock in the pre-announcement period. RET(0, þ 1) is the stock’s two-day percentage return following
the M&A announcement and measured from the closing price on day 2 1 to that on day þ 1.
RET(25, 21) is the stock’s percentage return before the M&A announcement and measured from the
closing price on day 2 6 to that on day 2 1. RET(þ2, þ 5) is the post-announcement return
calculated from the closing prices of day þ 1 to day þ 5. ABVOL(25, 21) is the stock’s abnormal
volume in the pre-announcement period, measured as the average daily volume in the pre-
announcement period divided by the average daily volume in the non-announcement period, all 21.
NORMRELS is the ratio of the shorted shares to trade shares in the non-announcement period. t-
Statistics are in the parentheses below the coef?cients. Quintiles are determined according to the
transaction ratio Table IV.
JFEP
1,2
190
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Panel A: regression results of target ?rms by average daily volume quintiles
Equation (4)
Target a
0
a
1
a
2
a
3
a
4
Adjusted R
2
Quintile 1 0.324 7.577 0.819 5.003 0.675
* * *
0.4734
(n ¼ 38) (1.22) (0.95) (0.32) (0.95) (5.86)
Quintile 2 0.225 1.939 0.268 0.889 1.026
* * *
0.3312
(n ¼ 38) (1.20) (0.38) (0.06) (0.85) (4.57)
Quintile 3 0.321 1.038 1.860 1.192 2.568
* * *
0.6082
(n ¼ 38) (1.25) (1.19) (0.95) (0.67) (7.80)
Quintile 4 0.008 0.830 20.434 20.040 0.863
* * *
0.8356
(n ¼ 38) (0.13) (0.68) (20.26) (21.46) (12.73)
Quintile 5 20.036 20.865 21.007 20.476 1.045
* * *
0.6623
(n ¼ 38) (20.27) (20.27) (20.35) (20.38) (8.92)
Equation (5)
Target y
0
y
1
y
2
y
3
y
4
Adjusted R
2
Quintile 1 0.111
*
0.043 20.109 0.234 0.519
*
0.1236
(n ¼ 38) (1.74) (0.05) (20.39) (0.41) (1.95)
Quintile 2 0.012 20.552 20.160 0.178 1.058
* * *
0.4671
(n ¼ 38) (0.21) (20.59) (20.25) (1.47) (4.88)
Quintile 3 0.005 0.519 0.058 0.203
* *
1.013
* * *
0.5236
(n ¼ 38) (0.14) (1.24) (0.16) (2.20) (6.46)
Quintile 4 0.142
* *
20.358 0.063 0.040 0.538
* *
0.1472
(n ¼ 40) (2.42) (20.66) (0.33) (0.23) (2.35)
Quintile 5 0.071 0.050 20.075 20.006 0.696
* * *
0.4487
(n ¼ 40) (1.96) (0.23) (20.27) (20.98) (4.82)
Panel B: regression results of acquiring ?rms by average daily volume quintiles
Equation (4)
Acquirer a
0
a
1
a
2
a
3
a
4
Adjusted R
2
Quintile 1 0.219 25.950 22.103 6.418
* * *
2.286
* * *
0.7176
(n ¼ 197) (1.40) (20.99) (20.55) (7.34) (17.18)
Quintile 2 0.049 0.175 0.131 21.011 0.591
* * *
0.3148
(n ¼ 197) (1.36) (0.35) (0.17) (21.43) (9.36)
Quintile 3 0.132
* *
212.077
* * *
0.871 20.585 1.252
* * *
0.4248
(n ¼ 197) (2.12) (28.77) (1.02) (20.89) (8.49)
Quintile 4 20.053
*
21.355 20.252 20.021 1.131
* * *
0.8855
(n ¼ 197) (21.70) (21.33) (20.36) (20.31) (39.27)
Quintile 5 0.041
* *
0.237 0.083 20.161 0.849
* * *
0.5784
(n ¼ 200) (2.20) (0.59) (0.49) (21.17) (16.65)
Equation (5)
Acquirer y
0
y
1
y
2
y
3
y
4
Adjusted R
2
Quintile 1 0.053
* * *
20.004 0.121 20.040 0.800
* * *
0.4284
(n ¼ 197) (3.20) (20.05) (0.99) (21.09) (12.21)
Quintile 2 0.052
* * *
20.253 20.013 20.104
* *
0.831
* * *
0.4359
(n ¼ 197) (3.15) (21.09) (20.12) (22.04) (12.47)
Quintile 3 0.123
* * *
0.108 0.085 0.094 0.524
* * *
0.2116
(n ¼ 197) (6.58) (0.81) (0.55) (1.51) (7.31)
Quintile 4 0.102
* * *
20.027 20.015 20.012 0.581
* * *
0.1779
(continued)
Table V.
Results of OLS
regression: abnormal
short sales and abnormal
relative short sales
around M&A
announcements
Short sales
around M&A
191
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
?rms with below average trading volume, but positively related to the
post-announcement return for acquiring ?rms with the highest trading volume.
4.4 Hypothesis 4
To test whether short sale changes are associated with changes in trading costs, we
estimated equations (6) and (7). Results are presented in Tables VI and VII, respectively.
From Table VI, we can see that b
6
, the coef?cient of Dummy, is statistically signi?cant
when spread is measured as quoted spread, percentage quoted spread, and effective
spread. None of the variables are signi?cant, when spread is measured by percentage
effective spread. When quoted spread is used, b
1
, the coef?cient of (1/price), is
signi?cant. When effective spread is used, b
3
, the coef?cient of turnover, and b
4
, the
coef?cient of volatility, are signi?cant. Overall, when testing equation (6) for target
?rms, Dummy explains a large portion of spread variation before and after the M&A
announcement. b
5
, the coef?cient of DailySS is not signi?cant for any spread measure.
Therefore, we conclude that DailySS has little effect on BAS of target ?rms.
For acquiring ?rms, b
3
, the coef?cient of turnover, and b
4
, the coef?cient of volatility,
are signi?cant when using quoted spread, percentage quoted spread, and effective
spread. b
2
, the coef?cient of log (volume), is signi?cant when using quoted spread. b
1
, the
coef?cient of (1/price), is signi?cant when using percentage quoted spread. None of the
variables is signi?cant when using percentage effective spread. b
5
, the coef?cient of
DailySS, and b
6
, the coef?cient of Dummy, are not signi?cant for any spread measure.
Therefore, neither the daily short sale nor dummy variable can explain spread variation
in acquiring ?rms.
From Table VII, we ?nd similar results for target ?rms as equation (6). For
acquiring ?rms, c
1
, the coef?cient of 1/price), is signi?cant only in percentage quoted
(n ¼ 197) (4.73) (20.16) (20.08) (20.78) (6.76)
Quintile 5 0.073
* * *
20.033 0.027 0.186
* * *
0.715
* * *
0.3036
(n ¼ 200) (3.36) (20.14) (0.43) (4.89) (8.46)
Notes: Statistical signi?cant at
*
10%,
* *
5%, and
* * *
1% levels
ABSSð25;21Þ ¼a
0
þa
1
RETð0;þ1Þþa
2
RETð25;21Þþa
3
RETðþ2;þ5Þþa
4
ABVOLð25;21Þþe ð4Þ
RELSSð25;21Þ ¼y
0
þy
1
RETð0;þ1Þþy
2
RETð25;21Þþy
3
RETðþ2;þ5Þþy
4
NORMRELSSþe ð5Þ
The results of OLS estimation of these equations, as ?tted to the sub-samples determined by
average daily volume are shown. ABSS(25, 2 1) is the average daily abnormal short sales for
stock in the pre-announcement period, measured as the average daily short sale in the pre-
announcement period divided by the average daily short sale in the non-announcement period, all
21. RELSS(25, 21) is the ratio of shorted shares to traded shares in the stock in the pre-
announcement period. RET(0, þ 1) is the stock’s two-day percentage return following the M&A
announcement and measured from the closing price on day 2 1 to that on day þ 1. RET(25, 21)
is the stock’s percentage return before the M&A announcement and measured from the closing
price on day 2 6 to that on day 2 1. RET(þ2, þ 5) is the post-announcement return calculated
from the closing prices of day þ 1 to day þ 5. ABVOL(25, 21) is the stock’s abnormal volume in
the pre-announcement period, measured as the average daily volume in the pre-announcement
period divided by the average daily volume in the non-announcement period, all 21.
NORMRELSS is the ratio of the shorted shares to trade shares in the non-announcement period. t-
Statistics are in the parentheses below the coef?cients. Quintiles are determined according to
average daily volume Table V.
JFEP
1,2
192
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
b
0
b
1
b
2
b
3
b
4
b
5
b
6
A
d
j
u
s
t
e
d
R
2
T
a
r
g
e
t
Q
u
o
t
e
d
s
p
r
e
a
d
0
.
2
6
3
*
(
1
.
6
9
)
2
2
.
2
4
9
*
(
2
1
.
8
7
)
2
0
.
0
0
5
(
2
0
.
8
1
)
2
0
.
1
4
7
(
2
0
.
4
2
)
0
.
0
8
3
(
0
.
6
1
)
0
.
0
0
0
(
0
.
6
2
)
0
.
0
3
9
*
*
*
(
4
.
7
3
)
0
.
7
7
7
2
Q
u
o
t
e
d
s
p
r
e
a
d
(
%
)
0
.
0
1
1
(
0
.
1
9
)
2
0
.
1
4
7
(
2
0
.
3
3
)
0
.
0
0
0
(
0
.
0
0
)
2
0
.
0
0
0
(
2
0
.
0
0
)
0
.
0
2
0
(
0
.
3
9
)
2
0
.
0
0
0
(
2
0
.
1
8
)
0
.
0
0
5
*
(
1
.
7
4
)
0
.
3
7
7
5
E
f
f
e
c
t
i
v
e
s
p
r
e
a
d
0
.
1
9
4
(
1
.
3
1
)
2
0
.
0
0
0
(
2
0
.
0
0
)
2
0
.
0
0
7
(
2
1
.
2
8
)
0
.
5
5
9
*
(
1
.
6
8
)
2
0
.
2
6
6
*
*
(
2
2
.
0
6
)
0
.
0
0
0
(
0
.
4
9
)
0
.
0
1
8
*
*
(
2
.
2
8
)
0
.
5
6
7
8
E
f
f
e
c
t
i
v
e
s
p
r
e
a
d
(
%
)
0
.
0
1
6
(
0
.
2
9
)
0
.
0
4
9
(
0
.
1
2
)
2
0
.
0
0
1
(
2
0
.
3
2
)
0
.
0
6
5
(
0
.
5
3
)
2
0
.
0
1
4
(
2
0
.
2
9
)
2
0
.
0
0
0
(
2
0
.
1
7
)
0
.
0
0
3
(
0
.
9
3
)
0
.
1
9
6
8
A
c
q
u
i
r
e
r
Q
u
o
t
e
d
s
p
r
e
a
d
3
.
3
1
3
(
1
.
5
3
)
1
1
.
6
7
0
(
1
.
1
8
)
2
0
.
1
5
0
*
(
2
1
.
6
8
)
7
.
1
4
3
*
*
(
2
.
2
5
)
2
0
.
0
0
3
*
*
(
2
2
.
3
5
)
0
.
0
0
0
(
0
.
4
7
)
0
.
0
0
2
(
0
.
2
8
)
0
.
2
1
6
7
Q
u
o
t
e
d
s
p
r
e
a
d
(
%
)
0
.
0
5
4
(
1
.
2
6
)
0
.
4
6
1
*
*
(
2
.
3
6
)
2
0
.
0
0
3
(
2
1
.
5
5
)
0
.
1
2
1
*
(
1
.
9
2
)
2
0
.
0
0
0
*
*
(
2
2
.
1
6
)
0
.
0
0
0
(
0
.
9
6
)
0
.
0
0
0
(
0
.
6
9
)
0
.
3
4
0
3
E
f
f
e
c
t
i
v
e
s
p
r
e
a
d
2
2
.
7
2
1
(
2
1
.
2
9
)
2
7
.
1
4
6
(
2
0
.
7
5
)
0
.
1
2
5
(
1
.
4
3
)
2
6
.
8
0
5
*
*
(
2
2
.
2
1
)
0
.
0
0
3
*
*
*
(
2
.
8
2
)
2
0
.
0
0
0
(
2
0
.
1
3
)
2
0
.
0
0
2
(
2
0
.
2
1
)
0
.
2
0
2
5
E
f
f
e
c
t
i
v
e
s
p
r
e
a
d
0
.
0
7
1
(
0
.
5
0
)
0
.
7
3
2
(
1
.
1
3
)
2
0
.
0
0
4
(
2
0
.
6
5
)
2
0
.
0
4
3
(
2
0
.
2
1
)
0
.
0
0
0
(
1
.
0
2
)
0
.
0
0
0
(
0
.
5
2
)
2
0
.
0
0
1
(
2
0
.
9
7
)
0
.
0
3
4
8
N
o
t
e
s
:
S
t
a
t
i
s
t
i
c
a
l
s
i
g
n
i
?
c
a
n
t
a
t
*
1
0
%
,
*
*
5
%
,
a
n
d
*
*
*
1
%
l
e
v
e
l
s
S
p
r
e
a
d
¼
b
0
þ
b
1
ð
1
=
P
R
I
C
E
Þ
þ
b
2
l
o
g
ð
V
O
L
U
M
E
Þ
þ
b
3
T
U
R
N
O
V
E
R
þ
b
4
V
O
L
A
T
I
L
I
T
Y
þ
b
5
D
a
i
l
y
S
S
þ
b
6
D
u
m
m
y
þ
e
ð
6
Þ
S
p
r
e
a
d
i
s
t
h
e
t
i
m
e
-
w
e
i
g
h
t
e
d
s
p
r
e
a
d
m
e
a
s
u
r
e
s
.
P
r
i
c
e
i
s
t
h
e
m
e
a
n
v
a
l
u
e
o
f
d
a
i
l
y
c
l
o
s
i
n
g
p
r
i
c
e
s
.
V
o
l
u
m
e
i
s
t
h
e
d
o
l
l
a
r
t
r
a
d
i
n
g
v
o
l
u
m
e
.
T
u
r
n
o
v
e
r
i
s
t
h
e
r
a
t
i
o
o
f
t
h
e
n
u
m
b
e
r
o
f
s
h
a
r
e
s
t
r
a
d
e
d
t
o
n
u
m
b
e
r
o
f
s
h
a
r
e
s
o
u
t
s
t
a
n
d
i
n
g
.
V
o
l
a
t
u
l
i
t
y
i
s
t
h
e
s
t
a
n
d
a
r
d
d
e
v
i
a
t
i
o
n
o
f
d
a
i
l
y
r
e
t
u
r
n
c
a
l
c
u
l
a
t
e
d
f
r
o
m
t
h
e
d
a
i
l
y
c
l
o
s
i
n
g
p
r
i
c
e
s
.
D
a
i
l
y
S
S
i
s
t
h
e
a
v
e
r
a
g
e
d
a
i
l
y
s
h
o
r
t
s
a
l
e
.
D
U
M
M
Y
i
s
1
i
f
b
e
f
o
r
e
M
&
A
a
n
n
o
u
n
c
e
m
e
n
t
,
(
2
6
0
,
2
1
)
,
a
n
d
0
i
f
a
f
t
e
r
a
n
n
o
u
n
c
e
m
e
n
t
,
(
0
,
þ
2
0
)
Table VI.
Results of OLS
regression: spreads and
average daily short sales
around M&A
announcements
Short sales
around M&A
193
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
c
0
c
1
c
2
c
3
c
4
c
5
c
6
A
d
j
u
s
t
e
d
R
2
T
a
r
g
e
t
Q
u
o
t
e
d
s
p
r
e
a
d
0
.
2
4
0
*
(
1
.
6
9
)
2
2
.
2
9
7
*
(
2
1
.
9
3
)
2
0
.
0
0
4
(
2
0
.
6
9
)
2
0
.
0
5
1
(
2
0
.
1
6
)
0
.
1
0
6
(
0
.
8
1
)
2
0
.
0
1
3
(
2
1
.
4
4
)
0
.
0
4
1
*
*
*
(
5
.
0
4
)
0
.
7
8
2
2
Q
u
o
t
e
d
s
p
r
e
a
d
(
%
)
0
.
0
1
8
(
0
.
3
4
)
2
0
.
1
6
5
(
2
0
.
3
7
)
2
0
.
0
0
0
(
2
0
.
1
3
)
2
0
.
0
0
8
(
2
0
.
0
7
)
0
.
0
1
8
(
0
.
3
7
)
2
0
.
0
0
2
(
2
0
.
6
9
)
0
.
0
0
6
*
(
1
.
8
1
)
0
.
3
8
1
2
E
f
f
e
c
t
i
v
e
s
p
r
e
a
d
0
.
1
6
9
(
1
.
2
4
)
0
.
0
0
8
(
0
.
0
1
)
2
0
.
0
0
6
(
2
1
.
2
0
)
0
.
6
2
7
*
(
2
.
0
6
)
2
0
.
2
5
0
*
(
2
1
.
9
9
)
2
0
.
0
0
3
(
2
0
.
3
1
)
0
.
0
1
9
*
*
(
2
.
3
8
)
0
.
5
6
7
0
E
f
f
e
c
t
i
v
e
s
p
r
e
a
d
(
%
)
0
.
0
2
2
(
0
.
4
3
)
0
.
0
3
5
(
0
.
0
8
)
2
0
.
0
0
1
(
2
0
.
4
9
)
0
.
0
5
8
(
0
.
5
1
)
2
0
.
0
1
6
(
2
0
.
3
3
)
2
0
.
0
0
2
(
2
0
.
5
6
)
0
.
0
0
3
(
0
.
9
7
)
0
.
1
9
9
8
A
c
q
u
i
r
e
r
Q
u
o
t
e
d
s
p
r
e
a
d
1
.
2
6
7
(
1
.
1
4
)
1
4
.
2
2
5
(
1
.
5
3
)
2
0
.
0
8
1
*
(
2
1
.
8
6
)
6
.
7
7
6
*
*
(
2
.
2
6
)
2
0
.
0
0
3
*
*
*
(
2
2
.
7
6
)
1
.
1
2
4
*
*
*
(
2
.
9
7
)
0
.
0
1
1
(
1
.
3
2
)
0
.
2
9
8
1
Q
u
o
t
e
d
s
p
r
e
a
d
(
%
)
2
0
.
0
0
7
(
2
0
.
3
0
)
0
.
5
2
4
*
*
*
(
2
.
8
5
)
2
0
.
0
0
1
(
2
0
.
6
6
)
0
.
1
1
5
*
(
1
.
9
5
)
2
0
.
0
0
0
*
*
(
2
2
.
5
9
)
0
.
0
2
4
*
*
*
(
3
.
2
7
)
0
.
0
0
0
*
(
1
.
8
9
)
0
.
4
1
4
6
E
f
f
e
c
t
i
v
e
S
p
r
e
a
d
2
1
.
5
1
7
(
2
1
.
3
8
)
2
9
.
0
4
2
(
2
0
.
9
8
)
0
.
0
8
8
*
*
(
2
.
0
5
)
2
6
.
4
1
9
*
*
(
2
2
.
1
7
)
0
.
0
0
3
*
*
*
(
3
.
1
8
)
2
0
.
9
3
9
*
*
(
2
2
.
5
2
)
2
0
.
0
0
9
(
2
1
.
0
7
)
0
.
2
6
5
3
E
f
f
e
c
t
i
v
e
s
p
r
e
a
d
(
%
)
0
.
0
3
3
0
.
4
3
(
0
.
7
1
3
)
(
1
.
1
1
)
2
0
.
0
0
2
(
2
0
.
6
3
)
2
0
.
0
2
2
(
2
0
.
1
0
)
0
.
0
0
0
(
1
.
1
3
)
2
0
.
0
2
6
(
2
1
.
0
0
)
2
0
.
0
0
1
(
2
1
.
2
0
)
0
.
0
4
4
0
N
o
t
e
s
:
S
t
a
t
i
s
t
i
c
a
l
l
y
s
i
g
n
i
?
c
a
n
t
a
t
*
1
0
%
,
*
*
5
%
,
a
n
d
*
*
*
1
%
l
e
v
e
l
s
;
S
p
r
e
a
d
¼
c
0
þ
c
1
ð
1
=
P
R
I
C
E
Þ
þ
c
2
l
o
g
ð
V
O
L
U
M
E
Þ
þ
c
3
T
U
R
N
O
V
E
R
þ
c
4
V
O
L
A
T
I
L
I
T
Y
þ
c
5
A
v
g
S
S
V
o
l
R
a
t
i
o
þ
c
6
D
u
m
m
y
þ
e
ð
7
Þ
S
p
r
e
a
d
i
s
t
h
e
t
i
m
e
-
w
e
i
g
h
t
e
d
s
p
r
e
a
d
m
e
a
s
u
r
e
s
.
P
r
i
c
e
i
s
t
h
e
m
e
a
n
v
a
l
u
e
o
f
d
a
i
l
y
c
l
o
s
i
n
g
p
r
i
c
e
s
.
V
o
l
u
m
e
i
s
t
h
e
d
o
l
l
a
r
t
r
a
d
i
n
g
v
o
l
u
m
e
.
T
u
r
n
o
v
e
r
i
s
t
h
e
r
a
t
i
o
o
f
t
h
e
n
u
m
b
e
r
o
f
s
h
a
r
e
s
t
r
a
d
e
d
t
o
n
u
m
b
e
r
o
f
s
h
a
r
e
s
o
u
t
s
t
a
n
d
i
n
g
.
V
o
l
a
t
u
l
i
t
y
i
s
t
h
e
s
t
a
n
d
a
r
d
d
e
v
i
a
t
i
o
n
o
f
d
a
i
l
y
r
e
t
u
r
n
c
a
l
c
u
l
a
t
e
d
f
r
o
m
t
h
e
d
a
i
l
y
c
l
o
s
i
n
g
p
r
i
c
e
s
.
A
v
g
S
S
V
o
l
R
a
t
i
o
i
s
a
v
e
r
a
g
e
d
a
i
l
y
s
h
o
r
t
e
d
s
h
a
r
e
s
t
o
d
a
i
l
y
t
r
a
d
e
d
s
h
a
r
e
s
.
D
u
m
m
y
i
s
1
i
f
b
e
f
o
r
e
M
&
A
a
n
n
o
u
n
c
e
m
e
n
t
,
(
2
6
0
,
2
1
)
,
a
n
d
0
i
f
a
f
t
e
r
a
n
n
o
u
n
c
e
m
e
n
t
,
(
0
,
þ
2
0
)
Table VII.
Results of OLS
regression: spreads and
average daily short-sale
ratio around M&A
announcements
JFEP
1,2
194
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
spread. c
2
, the coef?cient of log (volume), is signi?cant in quoted spread and effective
spread. c
3
, the coef?cient of turnover, and c
4
, the coef?cient of volatility, are signi?cant
in quoted spread, percentage quoted spread and effective spread. c
6
, the coef?cient of
Dummy, is signi?cant only in percentage quoted spread. We ?nd that c
5
, the coef?cient
of AvgSSVolRatio, is positive at the 1 percent level for quoted spread and percentage
quoted spread and is negative at the 5 percent level for effective spread. These results
imply that, as the ratio of shorted shares to traded shares increases, the quoted spreads
and percentage quoted spreads of acquiring ?rms increase. However, the effective
spreads of acquiring ?rms decrease as the ratio of shorted shares to traded shares
increases. Therefore, depending on the BAS measures, the empirical evidence support
H4 only for acquiring ?rms.
5. Summary and conclusions
We examine the changes in short-sale transactions of target ?rms and acquiring ?rms
around merger and acquisition (M&A) announcements using daily short-sale
transaction data from the NYSE and NASDAQ. Overall, we conclude the following
results: First, target ?rms experience signi?cant ES from day 2 1 to day þ 7; while
acquiring ?rm experience signi?cant ES from day 0 to day þ 20. Second, for acquiring
?rms, the ?ve day pre-announcement abnormal short sale is negatively related to the
announcement day return and is positively related to the post-announcement return.
We do not observe such a relationship for target ?rms. Third, we cannot conclude that
short sellers are informed traders around M&A announcements. Fourth, for target
?rms, we ?nd that short activity changes are not signi?cantly related to changes in
trading costs. For acquiring ?rms, short activity changes are positively related to
quoted spreads and percentage quoted spreads and are negatively related to effective
spreads. Given the current credit crisis, the SEC adopted a sequence of actions to adjust
the short-sale rule (SEC release no. 34-58166 (July 15, 2008) and 34-58572 (September
17, 2008) and 34-58592 (September 18, 2008)). Our ?ndings suggest that informed short
sellers, if exist, are hard to detect, therefore, any change to the current short-sale
regulation should be careful and speci?c. An overall short sale ban may compromise
market quality rather than effectively preventing abusive short selling.
Further research effort may be devoted to examine the derivative market around
M&A announcements. Since short sale is not the only way to establish short position
and is relatively costly, stock futures and options may be good alternatives for
informed traders to maximize their pro?t.
Notes
1. The SEC adopted regulation SHO to update short-sale regulation in light of numerous
market developments since short-sale regulations were ?rst adopted in 1938. Compliance
with regulation SHO began on January 3, 2005.
2. “Certain market events and trading strategies may make a security more vulnerable to
abusive short-sale activity. Speci?c market events related to an issuer or a security (such as a
pending merger or acquisition) may cause this increased vulnerability. We therefore, request
comment on whether short selling should continue to be regulated or even prohibited during
speci?c market conditions. Are there corporate events (e.g. mergers, acquisitions, or tender
offers) that make a security vulnerable to abusive short selling? Should short selling be
prohibited for a period preceding a signi?cant corporate or market event? If the Rule was
Short sales
around M&A
195
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
eliminated, should restrictions continue to apply preceding a signi?cant corporate or market
event?” excerpt from SEC Release No. 34-42037 (October 28, 1999), 64 FR 57996.
3. Kweon and Pinkerton (1981) and Ascioglu et al. (2002).
4. According to regulation T, the initial margin for short sales is 150 percent. However, the ?rst
100 percent of the requirement can be satis?ed by the proceeds of the short sale, leaving just
50 percent for the investor to maintain in the initial margin (so short selling looks much like
the case of going long). According to Rule 2520, the maintenance margin for short sales is $5
per share or 30 percent of the current market value, whichever is greater.
5. SEC 17 CFR PART 240, Release No. 34-42037; File No. S7-24-99, RIN 3235-AH84 Short Sales.
6. On June 23, 2004, the SEC adopted regulation SHO to establish uniform locate and delivery
requirements, create uniform marking requirements for sales of all equity securities, and to
establish a procedure to temporarily suspend the price-tests for a set of pilot securities
during the period May 2, 2005-April 28, 2006 in order to examine the effectiveness and
necessity of short-sale price-tests. At the same time, the SEC mandated that all self
regulatory organizations make tick-data on short-sales publicly available starting January 2,
2005. On April 20, 2006, the SEC announced that the short-sale Pilot has been extended to
August 6, 2007. The SHO-mandated data includes the ticker, price, volume, time, listing
market, and trader type (exempt or non-exempt from short-sale rules) for all short sales.
References
Aitken, M.J., Frino, A., McCorry, M.S. and Swan, P.L. (1998), “Short sales are almost
instantaneously bad news: evidence from the Australian stock exchange”, Journal of
Finance, Vol. 53, pp. 2205-23.
Ascioglu, A.N., McInish, T.H. and Wood, R.A. (2002), “Merger announcements and trading”,
Journal of Financial Research, Vol. 25, pp. 263-78.
Brent, A., Morse, D. and Stice, E.K. (1990), “Short interest: explanations and tests”, Journal of
Financial and Quantitative Analysis, Vol. 25, pp. 273-89.
Chae, J. (2005), “Trading volume, information asymmetry, and timing information”, Journal of
Finance, Vol. 60, pp. 413-42.
Christophe, S., Ferri, M. and Angel, J. (2004), “Short-selling prior to earnings announcements”,
Journal of Finance, Vol. 59, pp. 1845-75.
Daouk, H. and Charoenrook, A. (2005), “A study of market-wide short selling restrictions”,
working paper, Cornell University, Ithaca, NY.
Dechow, P.M., Hutton, A.P., Meulbroek, L. and Sloan, R.G. (2001), “Short sellers, fundamental
analysis and stock returns”, Journal of ?nancial Economics, Vol. 61, pp. 77-106.
Desai, H., Krishnamurthy, S. and Venkataraman, K. (2006), “Do short sellers target ?rms with
poor earnings quality? Evidence from earnings restatements”, Review of Accounting
Studies, Vol. 11, pp. 71-90.
Desai, H., Ramesh, K., Ramu Thiagarajan, S. and Balachandran, B.V. (2002), “An investigation of
the informational role of short interest in the NASDAQ market”, Journal of Finance,
Vol. 57, pp. 2263-87.
Diamond, D.W. and Verrecchia, R.E. (1987), “Constraints on short-selling and asset price
adjustment to private information”, Journal of Financial Economics, Vol. 18, pp. 277-311.
Dyvbig, P.H. (1984), “Short sales restrictions and kinks on the mean variance frontier”, Journal of
Finance, Vol. 39, pp. 239-44.
JFEP
1,2
196
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Efendi, J., Swanson, E.P. and Chairholder, D. (2009), “Short sellers trading in companies with
accounting irregularities”, working paper, University of Texas, Arlington, TX.
Fama, E. (1970), “Ef?cient capital markets: a review of theory and empirical work”, Journal of
Finance, Vol. 25, pp. 383-417.
Figlewski, S. (1981), “The informational effects of restrictions on short sales: some empirical
evidence”, Journal of Financial and Quantitative Analysis, Vol. 16, pp. 463-76.
Finnerty, J.D. (2005), “Short selling, death spiral convertibles, and the pro?tability of stock
manipulation”, working paper, Fordham University, New York, NY.
Genard, B. and Nanda, V. (1993), “Trading and manipulation around seasoned equity offerings”,
Journal of Finance, Vol. 48, pp. 213-45.
Jarrell, G., Brickley, J. and Netter, J. (1988), “The market for corporate control: the empirical
evidence since 1980”, Journal of Economic Perspective, Vol. 2, pp. 49-68.
Jarrow, R.A. (1980), “Heterogeneous expectations, restrictions on short sales, and equilibrium
asset prices”, Journal of Finance, Vol. 35, pp. 1105-13.
Jensen, M.C. and Ruback, R.S. (1983), “The market for corporate control: the scienti?c evidence”,
Journal of Financial Economics, Vol. 11, pp. 5-55.
Jones, C.M. (2008), “Shorting restrictions: revisiting the 1930s”, working paper, Columbia
Business School, New York, NY.
Kweon, A.J. and Pinkerton, J.M. (1981), “Merger announcements and insider trading activity:
an empirical investigation”, Journal of Finance, Vol. 36, pp. 855-69.
McInish, T.H. and Wood, A. (1992), “An analysis of intraday patterns in bid-ask spreads for
NYSE stocks”, Journal of Finance, Vol. 47, pp. 753-64.
Miller, E.M. (1977), “Risk, uncertainty, and divergence of opinion”, Journal of Finance, Vol. 32,
pp. 1151-68.
Ross, S.A. (1977), “The capital asset pricing model (CAPM), short-sale restrictions and related
issues”, Journal of Finance, Vol. 32, pp. 177-83.
Senchack, A.J. and Starks, L.T. (1993), “Short-sale restrictions and market reaction to
short-interest announcements”, Journal of Financial and Quantitative Analysis, Vol. 28,
pp. 177-94.
White, H. (1980), “A heteroskedasticity-consistent covariance matrix estimator and a direct test
for heteroskedasticity”, Econometrica, Vol. 48, pp. 817-38.
Woolridge, J.R. and Dickson, A. (1994), “Short selling and common stock prices”, Financial
Analysts Journal, Vol. 50, pp. 20-8.
Further reading
Breusch, T.S. and Pagan, A.R. (1979), “A simple test for heteroscedasticity and random
coef?cient variation”, Econometrica, Vol. 47, pp. 1287-94.
Corresponding author
Liuqing Mai can be contacted at: [email protected]
Short sales
around M&A
197
To purchase reprints of this article please e-mail: [email protected]
Or visit our web site for further details: www.emeraldinsight.com/reprints
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
3
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_942385255.pdf