Directory UMM :Data Elmu:jurnal:J-a:Journal Of Banking And Finance:Vol24.Issue8.2000:
Journal of Banking & Finance 24 (2000) 1307±1321
www.elsevier.com/locate/econbase
Information-signaling and competitive eects of
foreign acquisitions in the US
Aigbe Akhigbe a, Anna D. Martin
a
b,*
Frederick W. Moyer Chair in Finance, College of Business Administration, The University
of Akron, Akron, OH 44325-4805, USA
b
School of Business, Fair®eld University, Fair®eld, CT 06430, USA
Received 14 September 1998; accepted 17 June 1999
Abstract
The stock price eects for the domestic competitors of foreign acquisition targets in
the US are found to be signi®cantly positive. These results imply that signals of favorable industry conditions conveyed through cross-border acquisitions dominate any
perceived changes in competitive balance. Consistent with the information-signaling
hypothesis, the stock price eects are more favorable for relatively small competitors,
for rivals with stock returns that are highly correlated with the targetÕs stock returns,
when the targets experience favorable stock price eects, in technologically-intense industries, for rivals with poor prior performance, and with related acquisitions. Consistent with the competitive hypothesis, the stock price eects are less favorable with
high degrees of ®nancial leverage and when the acquirers already have a presence in the
US. Ó 2000 Elsevier Science B.V. All rights reserved.
JEL classi®cation: F23; G34
Keywords: Foreign acquisitions; Cross-border acquisitions; Intra-industry eects; Rivals
*
Corresponding author. Tel.: +203-254-4000 ext. 2881; fax: +203-254-4105.
E-mail address: amartin@fair1.fair®eld.edu (A.D. Martin).
0378-4266/00/$ - see front matter Ó 2000 Elsevier Science B.V. All rights reserved.
PII: S 0 3 7 8 - 4 2 6 6 ( 9 9 ) 0 0 0 7 4 - 6
1308
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
1. Introduction
The stock price eects of cross-border acquisitions have received much attention in recent years (e.g., Harris and Ravenscraft, 1991; Kang, 1993; Pettway et al., 1993; Servaes and Zenner, 1994; Eun et al., 1996). These studies
focus on the stock price response of the target and/or acquirers and show that
shareholders of US targets gain from foreign acquisitions while the shareholders of the foreign acquirers do not consistently gain or lose.
To date, however, there has been no assessment of the stock price eects for
the competitors of foreign acquisition targets. Consider the agreement by
Daimler-Benz to acquire Chrysler in a stock transaction valued at about US$35
billion on May 7, 1998. This combination was expected to enhance ChryslerÕs
competitiveness by giving it access to German engineering prowess, and potentially reducing the cash¯ows to Ford and General Motors. Upon announcement,
the market price of Chrysler rose sharply, and the market prices of Ford and
General Motors fell at the news of the proposed acquisition. Thus, an announced
acquisition of a US ®rm by a foreign ®rm could be expected to aect the equity
values of the target as well as its local rivals. This paper estimates the impact of
foreign acquisitions on the US target ®rmÕs industry rivals and investigates
various factors that may explain why those eects dier across acquisitions.
Two competing hypotheses are proposed to describe the stock price response of the targetÕs rivals. Under the information-signaling hypothesis, foreign acquisitions would indicate that further acquisitions are imminent
(Knickerbocker, 1974; Flowers, 1976) and domestic competitors experience a
positive stock price reaction as their probability of becoming a takeover target
increases. In contrast, under the competitive hypothesis, ®rms that make crossborder investments possess sucient competitive advantages to more than
oset the inherent disadvantages (Caves, 1971) and domestic competitors experience a negative stock price reaction as the entrance of the foreign acquirers
into the industry increases the degree of competition they face.
This study shows that the domestic competitors experience signi®cantly
positive valuation eects upon announcement of foreign acquisitions of US
®rms. Overall, these results provide evidence that signals of favorable industry
conditions conveyed through cross-border acquisitions dominate any perceived
changes in competitive balance. The evidence in this study is important in light
of the direct foreign investment (DFI) theories that argue that DFI may either
bene®t or adversely aect competitors.
2. Theory and testable hypotheses
Although a single theory to explain DFI has not been developed, the industrial organization view helps to understand the motivation for investing
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
1309
abroad (e.g., Hymer, 1960; Vernon, 1966; Caves, 1971; Flowers, 1976;
Dunning, 1988). These related theories emphasize that market imperfections
and the pursuit of leveraging or preserving competitive advantages are the
primary motivating factors in cross-border investments. Imperfections may
arise naturally from the operation of the ®rm, may be arti®cially induced by
government policies, or may occur in the ®nancial markets (Scholes and
Wolfson, 1990; Froot and Stein, 1991; Harris and Ravenscraft, 1991;
Servaes and Zenner, 1994). Examples of some advantages that may be exploited are economies of scale and scope, managerial talent, and technological expertise.
2.1. Information-signaling hypothesis
This hypothesis suggests that acquisitions convey information about the
possibility of further takeover activity within the industry, which should bene®t
the target ®rmÕs competitors. If foreign acquisitions are utilized by aggressive
®rms to expand geographically, oligopolistic rival reactions are likely to occur
(e.g., Knickerbocker, 1974; Flowers, 1976). Foreign ®rms are more likely to
aggressively pursue expansion eorts in the US when DFI is relatively appealing. 1 Thus, announcements of foreign acquisitions in the US may indicate
that US assets are relatively appealing and that further acquisitions in the US
are probable. Thus, it can be hypothesized that the stock price eects for the
domestic competitors of US targets of direct foreign investment are positive.
Several factors are examined to determine whether there are cross-sectional
dierences in the information eects.
Impact of the relative size of the rivals. Based on the work of Atiase (1985), it
can be argued that the information-signaling eects are inversely related to the
relative size of the rival (e.g., Slovin et al., 1991). Incremental information
should be conveyed about relatively small rivals at the time of the announcement, if relatively small ®rms are less closely followed by the market. In addition, to the extent that larger rivals are more dicult to acquire, the
likelihood of further acquisitions involving relatively large rivals is lower. 2
1
Direct foreign investment in the US may be relatively appealing for a variety of reasons.
First, US assets may be valuable to foreign ®rms due to market imperfections particular to
them in developing, exploiting, or preserving competitive advantages. For example, foreign
®rms may discover it less costly to coordinate their activities or more eective to identify local
product requirements with direct operations in the US. Secondly, foreign ®rms may ®nd US tax
laws relatively advantageous or may desire to avoid taris and quotas (Scholes and Wolfson,
1990; Servaes and Zenner, 1994). Thirdly, US assets appear less expensive when foreign
currencies are strong relative to the US dollar (Froot and Stein, 1991; Harris and Ravenscraft,
1991).
2
We thank an anonymous reviewer for strengthening this hypothesis.
1310
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
Thus, it is hypothesized that the information eects are greater (smaller) for
relatively small (large) rivals.
Impact of similar cash ¯ows. The degree of similarity between the cash
¯ows of the rivals and the cash ¯ows of the targets may help dierentiate
the rival eects. The announcement of the acquisition indicates that the
operating structure of the target is desirable. 3 If an oligopolistic rival reaction is anticipated, rivals with comparable operating structures should be
candidates for future takeovers and experience more favorable information
eects. 4
Impact of other factors. Incremental information about the future prospects
for an industry is revealed in the CARs that accrue to the target and should
also have value for the industry rivals. 5 Information eects may be greater for
rivals in industries that are more attractive to foreign ®rms, thus related acquisitions may signal industry consolidation and indicate further acquisitions
within the industry.
2.2. Competitive hypothesis
This hypothesis is based on the idea that acquisitions may adversely
aect the future performance of the targetÕs rivals. DFI theories imply that
®rms must possess substantial competitive advantages to more than oset
their inherent disadvantages from operating abroad, such as managing
geographically dispersed operations and conducting business in an unfamiliar culture and environment (e.g., Caves, 1971). 6 To the extent that
foreign acquirers overcome their inherent disadvantages with convincing
3
Lang and Stulz (1992) present a similar argument where the information eect is expected to be
larger when rivals have similar cash ¯ows.
4
If the current acquisition is an oligopolistic reaction to a previous acquisition, rivals with
similar cash ¯ows may be viewed as the ``losing'' targets, even if the rivals had not been formally or
publicly under consideration. In this case, rivals with similar cash ¯ows would experience less
favorable information eects.
5
Hertzel (1991) implies that the larger the impact of announcements on the targets the greater
the information signal.
6
Foreign ®rms competing in the US may possess a variety of advantages over domestic ®rms.
First, the nature of ®nancial and managerial decisions made by foreign managers may dier from
domestic managers (e.g. Doyle et al., 1992). Secondly, foreign ®rms may more eectively
capitalize on basic research. Eun et al. (1996) provide some evidence that cross-border
acquisitions of US ®rms are synergistic due to research and development capabilities.
Additionally, the products of foreign ®rms are sometimes perceived to have superior quality
(e.g., Shapiro, 1996).
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
1311
advantages, resulting in a net advantage, the stock price eects for the
target ®rmÕs rivals are hypothesized to be negative. 7 Several factors are
investigated to determine whether there are cross-sectional dierences in the
competitive eects.
Impact of ®nancial leverage. Financial leverage may limit a ®rmÕs ability to
make investments to respond to competitive challenges (Stulz, 1990). Therefore, it is hypothesized that rivals with higher degrees of ®nancial leverage incur
more negative competitive eects due to their limited ability to expeditiously
compete.
Impact of the degree of competition. Another factor that is explored is the
degree of competition within the industry. It can be argued that rivals in highly
competitive industries experience greater competitive eects. Since rivals in
competitive industries are already aggressively competing, the additional
challenge from a foreign rival with presumably net competitive advantages may
have a more detrimental impact. However, a detrimental impact on the rivals
may also occur in less competitive industries, but for a dierent reason. Industries with a low degree of competition tend to participate in collusive behavior (Sudharshan, 1995). In this environment, the foreign competitors may
not be as likely to join the coordination eorts, thus may adversely impact the
rivals. 8
Impact of the rivals' prior performance. Rivals that have had poor performance prior to the acquisition are not likely to be positioned to eectively
compete with a new entrant. Thus, it can be hypothesized that underperforming rivals are more adversely aected by the acquisition. Alternatively,
underperforming rivals may be expected to improve their performance due to
the threat of an acquisition or an actual acquisition. Upon announcement, this
information-signal may favorably impact the returns of poor performing rivals.
Impact of other factors. Related acquisitions are more likely to achieve
economies of scale or scope (e.g., Caves, 1971). Foreign acquirers with prior
US experience may not suer as greatly from the inherent disadvantages from
operating abroad and instead are able to focus on their competitive strengths,
and the real or perceived competitive threat may be greater depending upon the
acquirerÕs country of origin.
7
Hennart and Park (1993) argue that acquisitions of existing US operations, as opposed to
establishing new US subsidiaries, are used by Japanese ®rms that do not have strong competitive
advantages. If acquisitions are generally used by foreign ®rms that do not have strong competitive
advantages, the competitive eects described here may not be elicited.
8
If the rival ®rms in less competitive industries are expected to focus on improving their
performance as a result of new competition, the rival eects would be positive.
1312
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
3. Sample characteristics and valuation methodology
3.1. Sample characteristics
To be included in the sample of foreign acquisitions made in the US, the
following selection and screening techniques are used:
1. Foreign acquisitions of US ®rms are listed in Mergers and Acquisitions with
completion dates between 1985 and 1996.
2. The bidder gains at least 50% control of the target.
3. The target has stock price data available on NYSE, ASE, or NASDAQ
CRSP ®les at the time of the acquisition announcement.
4. Acquisition announcement dates are available in Predicast's F & S. The announcements are screened for confounding events. Some announcement
dates are in 1984 for acquisitions completed in 1985.
5. The announcements are made in daily publications such as the Wall Street
Journal, New York Times, Financial Times, etc.
The ®nal sample consists of 165 acquisition announcements over 1984±1996.
A wide range of industries are represented in the sample with 119 distinct fourdigit SIC codes included. Table 1 summarizes the sample distribution by event
year and country. Note that 94 (57%) of the 165 acquisitions occurred in the
late 1980s (1986±1989). British ®rms conducted the most acquisitions (40%) in
this sample.
Table 2 provides further summary statistics on the targets and their industry
rivals. These statistics are provided for three dierent de®nitions of industry
rivals, US ®rms that share the same four-digit, three-digit, and two-digit SIC
code with the target at the time of the acquisition announcement. The SIC
codes are provided in the CRSP ®les. 9 The mean market value of the targets is
$1.1 billion, while the mean market value of the rivals ranges from $1.4 billion
(two-digit SIC code de®nition) to $2.0 billion (four-digit SIC code de®nition).
The mean, median, minimum, and maximum number of rivals per event are
also provided. As expected, the mean, median, and maximum number of rivals
increase as the de®nition of ÔrivalÕ is broadened.
3.2. Valuation methodology
The valuation eects of foreign acquisitions in the US are estimated for
equally-weighted portfolios of the industry rivals for each announcement. A
simple market model is used to predict daily share price returns, with a 200day estimation period from 220 days before the announcement to 20 days
before the announcement for each portfolio of rivals. Daily returns for the
9
We ®nd qualitatively similar results when the SIC codes from Compustat are used.
Country
Number of acquisitions
1984±
1996
England
Canada
France
Japan
Germany
Switzerland
Netherlands
Otherb
Total
a
1984
1985
1986
1987
1988
1989
1990
66
18
15
13
12
10
8
23
1
2
2
0
0
1
1
1
2
0
1
0
0
0
0
1
12
5
2
0
1
1
2
4
13
0
0
0
0
1
0
2
165
8
4
27
16
1991
1992
1993
1994
10
1
2
3
2
2
0
4
6
6
1
7
1
1
1
4
24
27
1995
1996
5
0
4
1
2
0
0
1
3
0
1
0
0
0
0
1
1
0
0
1
0
0
0
0
3
0
0
1
1
0
0
1
5
0
0
0
2
2
0
0
3
2
1
0
2
2
2
3
2
2
1
0
1
0
2
1
13
5
2
6
9
15
9
The sample of 165 foreign acquisitions of US ®rms is derived from Mergers and Acquisitions with completion dates between 1985 and 1996, where the
bidder gains at least 50% control of the target, the target has stock price data available, and announcement dates are available in Predicast's F & S from
daily publications.
b
Includes acquirers from Argentina, Australia, Belgium, Ireland, Israel, Italy, Mexico, South Korea, Sweden, and Taiwan.
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
Table 1
The sample distribution of 165 foreign acquisitions of US targets by year and by countrya
1313
1314
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
Table 2
Summary statistics of 165 US targets acquired by foreign ®rms and their corresponding industry
rivals, de®ned by CRSP four-digit, three-digit and two-digit SIC codes
Mean market value of target ($million)
Mean market value of rivals ($million)
Mean number of rivals per acquisition
Median number of rivals per acquisition
Minimum number of rivals per acquisition
Maximum number of rivals per acquisition
Four-digit
SIC
Three-digit
SIC
Two-digit
SIC
1101.78
1954.16
15.8
7
1
124
1101.78
1923.80
27.1
18
1
159
1101.78
1388.81
99.3
78
1
638
rivals are taken from the CRSP ®les to create an equally-weighted rival
portfolio. The CRSP equally-weighted index is used as a proxy for the
market.
Daily abnormal returns are then calculated over an examination period of
21 days, from 10 days pre-announcement to 10 days post-announcement.
Based on the hypotheses in this study, positive abnormal returns support the
information-signaling argument and negative abnormal returns support the
competitive argument. z-Statistics are calculated using the methodology of
Mikkelson and Partch (1988).
4. Results
4.1. Valuation eects
Table 3 reports the valuation eects for US targets acquired by foreign ®rms
and their corresponding rival portfolios. Panel A reports the valuation eects
for the 165 targets. On average, the 2-day event period cumulative AR (CAR)
is 23.39% and signi®cant at the 0.001 level. This result is consistent with the
aforementioned studies that document favorable valuation eects for targets
upon the announcements of acquisitions.
The valuation eects of the rival portfolios are provided in Panels B, C,
and D of Table 3. 10 These panels report the valuation eects for the rival
portfolios using three dierent de®nitions for ÔrivalÕ. When the rivals are
de®ned as those ®rms that share the same four-digit SIC code (Panel B), the
2-day event period mean CAR is 0.50% and signi®cant at the 0.001 level. Panel
C (Panel D) reports the valuation eects as 0.48% (0.23%) when the rivals are
de®ned as those ®rms that share the same three-digit (two-digit) SIC code.
10
Excluding losing bidders from the rival portfolios does not signi®cantly change the rival
portfolio eects.
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
1315
Table 3
Valuation eects for US targets acquired by foreign ®rms and their corresponding industry rival
portfolios, de®ned by CRSP four-digit, three-digit and two-digit SIC codesa
Examination period
CAR
z-Statistics
% Pos
12.11d
80.95d
0.19
63
91
47
Panel B: industry rivals of US targets, de®ned by four-digit SIC
Pre-event period ()11,)2)
0.0006
)0.71
Event period ()1,0)
0.0050
3.61d
Post-event period (+1,+10)
)0.0009
0.13
46
56
47
Panel C: industry rivals of US targets, de®ned by three-digit SIC
Pre-event period ()11,)2)
0.0038
0.78
Event period ()1,0)
0.0048
3.53d
Post-event period (+1,+10)
)0.0010
0.39
47
58
51
Panel D: industry rivals of US targets, de®ned by two-digit SIC
Pre-event period ()11,)2)
0.0013
0.95
Event period ()1,0)
0.0023
3.55d
Post-event period (+1,+10)
0.0014
0.44
51
59
49
Panel A: US targets acquired by foreign ®rms
Pre-event period ()11,)2)
0.0698
Event period ()1,0)
0.2339
Post-event period (+1,+10)
)0.0020
a
The reported CARs are cumulative abnormal returns over three dierent examination periods. d
denotes signi®cance at the 0.001 level. The average abnormal return and z-statistics can dier in
sign because the former assigns uniform weights to each observation and the latter assigns nonuniform weights (see Mikkelson and Partch, 1988).
Thus, it appears the rival stock price eects are somewhat robust to the de®nition of ÔrivalÕ. Although, the eects are reduced as a more broad de®nition of
rival is used.
In this study, industry rivals will be primarily de®ned as those ®rms with the
same four-digit SIC code as the target (Hertzel, 1991; Slovin et al., 1991; 1992;
Szewczyk, 1992). Industry rivals de®ned in this manner are available for 120 of
the 165 acquisitions. In those cases where there are no rival ®rms with the same
four-digit SIC code at the time of the announcement, the rivals are identi®ed
using ®rms with the same three-digit SIC code. 11 Using this approach, the 2day event period mean CAR is 0.44% and signi®cant at the 0.001 level. 12
The results in Table 3 indicate that the information-signaling eects dominate the competitive eects. This evidence is interesting in light of the DFI
theories that suggest cross-border acquisitions could either bene®t or adversely
aect competitors.
11
This scheme is also applied to generating the cross-sectional variables.
The distribution of the 2-day CARs does not reveal a majority of ®rms with large, positive
ARs. Of the 165 rival portfolios, 95 accumulate positive CARs, 70 accumulate negative CARs, and
86 accumulate between plus and minus 1%.
12
1316
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
4.2. Cross-sectional regression analysis of rival stock price eects
Dierences in the valuation eects across the domestic competitors are explored using regression analysis. Based on the information-signaling hypothesis, we expect a more positive rival stock price response for relatively small
rivals, for rivals with similar cash ¯ows, when the targets experience positive
stock price eects, and for rivals in industries that are more attractive to foreign ®rms. Based on the competitive hypothesis, we expect a more negative
rival stock price response for highly leveraged rivals, for rivals operating in
industries with a high or low degree of competition, for poorly performing
rivals, and when the acquirer already has an established presence in the US.
Relative size (RELSIZE). This variable takes on the value of one for relatively small rivals, and zero otherwise. 13 Market values are calculated as the
number of common shares outstanding multiplied by the share price 14
Correlation in stock returns (CORR). As a proxy for the similarity in cash
¯ows, the correlation between the stock returns of the target and the rival
portfolio is calculated over the 200-day period (tÿ220 to tÿ20 ) prior to the acquisition announcement.
Target abnormal return (TARGAR). The 2-day cumulative abnormal return
for each target is calculated in the same manner as the rival portfolio.
Financial leverage (FLEV). This industry-speci®c variable is de®ned as the
median ratio of total long-term debt to total equity of the rivals. 15
Degree of competition (COMP). The Her®ndahl index is used as a proxy for
the degree of competition. It is de®ned as the squared sum of the fractions of
industry market capitalization by the rival ®rms. 16 An industry is considered
less competitive with a greater Her®ndahl index.
Performance (PERFORM). The average monthly abnormal return for each
rival is estimated over the 12 month period prior to the acquisition announcement. The median AR of the rivals is used to indicate the prior performance of the industry rivals.
13
A continuous variable, measured as the ratio of the market value of the target to the median
market value of the rivals, is also considered. However, after removing outliers, the results are
essentially the same as those reported.
14
The information eects may not be driven by the size of the rivals, but instead it may be driven
by the size of the targets. Szewczyk (1992) implies that acquisitions of large targets convey more
information than acquisitions of small targets.
15
These data are taken from Compustat. In order to generate the ®nancial leverage variable for
the industry rivals, a two-digit SIC code de®nition is used in seven cases, a three-digit SIC code
de®nition is used in 61 cases, and a four-digit SIC code de®nition is used in 97 cases.
16
The results are qualitatively the same when COMP is de®ned as the squared sum of the
fractions of industry sales.
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
1317
Currency strength (XSTRENGTH). The relative strength of the bidderÕs
currency is used as a control variable. This variable is estimated following
Harris and Ravenscraft (1991), Cebenoyan et al. (1992). Positive (negative)
values of XSTRENGTH occur when the bidderÕs currency is weak (strong)
relative to the US dollar. 17 The exchange rates are in units of foreign currency
per US dollar and are available in the Wall Street Journal.
Dummy variables. The rival eects are also examined for cross-sectional
variation by industry, 18 the relatedness of the acquisition at the two-digit SIC
code level, and the acquirerÕs prior presence in the US 19 using dummy variables. 20
Table 4 displays the results of the regression analysis. To control for heteroskedasticity, the variables have been standardized by the standard error of
the model used to estimate the abnormal returns. The variance in¯ation factors
(VIFs) are reported to assess the extent of multicollinearity. The VIFs range
from 1.16 to 1.93, indicating multicollinearity is not substantially in¯uencing
the coecients. 21
The cross-sectional results provide evidence supporting both the information-signaling and competitive hypotheses. More speci®cally, RELSIZE is
found to be positive and signi®cant. As hypothesized, the information eects
are greater (smaller) for relatively small (large) rivals.
17
In measuring XSTRENGTH, the exchange rate of the year of the announced acquisition is
used if the announcement occurred in the second half of the year. The exchange rate of the year
prior to the announcement is used if the announcement occurred in the ®rst half of the year (Harris
and Ravenscraft, 1991).
18
Each industry is classi®ed into one of three groups (technology-intensive, other manufacturing-oriented, or service-oriented) following Cebenoyan et al. (1992). DUMTECH has a value of 1
for technologically intense industries (two-digit SIC codes 28 (chemicals), 33 and 34 (metals), 35
(machinery), 36 (electrical), and 38 (scienti®c instruments)) and 0 otherwise; DUMIND has a value
of 1 for manufacturing-oriented but not technologically intense industries (two-digit SIC codes 10
(mining), 13 (oil/gas), 14 (non-metallic minerals), 20 and 39 (manufacturing), 22 (textile mills), 26
(paper), 27 (printing/publishing), 29 (petroleum/coal), 30 (rubber/plastics), 32 (stone/clay/glass),
and 37 (transportation equipment)), and 0 otherwise). The eects of the service-oriented industries
(two-digit SIC codes 47, 73, 80, 82, and 87 (services), 48 (communication), 50 and 51 (wholesale),
53, 54, and 54 (retail), 60, 62, 63, and 64 (®nance/insurance/real estate), 67 (holding cos.), 70
(hotels), and 78 (motion pictures)) are captured by the intercept.
19
Prior presence is established by reviewing the Directory of Corporate Aliations and Standard
and Poor's Register for US operating locations at the time of the acquisition.
20
When the acquisitions are grouped by the country of the acquirer, no speci®c country reveals
signi®cant mean CARs. When the acquisitions are grouped by year, the acquisitions in 1994 reveal
a signi®cant (p < 0.05) mean CAR of 1.18%. Since there is little evidence of country and year eects,
these dummy variables are excluded.
21
Some studies use VIF > 5 as an indication of severe multicollinearity (see Marquardt and Snee,
1975). Furthermore, univariate regressions show the magnitude and signi®cance of the coecients
are stable.
1318
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
Table 4
Cross-sectional factors of the rival eects of foreign acquisitions in the USa
Variables
Coecients
Intercept
RELSIZE
CORR
TARGAR
DUMTECH
DUMIND
FINLEV
COMP
PERFORM
RELATED
PRIORP
XSTRENGTH
)0.0193
0.0148
0.0663
0.0381
0.0165
0.0056
0.0203
)0.0187
)0.0502
0.0278
)0.0214
)0.0132
Sample size
F-value
Adj. R2
t-Statistics
)1.41
1.92a
3.58d
2.84c
2.09b
0.53
0.76
)1.48
)4.32d
3.43d
)2.56c
)0.52
VIF
0.00
1.16
1.36
1.28
1.50
1.93
1.37
1.28
1.29
1.49
1.50
1.51
165
12.28d
0.4321
a
The dependent variable is the rival portfolio two-day cumulative AR. a, b, c, and d denote signi®cance at the 0.10, 0.05, 0.01, and 0.001 levels, respectively. RELSIZE 1 if target market value>median market value of rivals, 0 otherwise; CORR median correlation between target and
rivals stock returns over tÿ220 and tÿ20 ; TARGAR target two-day CAR; DUMTECH 1 if target
industry is technologically-intense, 0 otherwise; DUMIND 1 if target industry is manufacturing
but not technologically-intense, 0 otherwise; FINLEV median debt/equity of rivals;
COMP Her®ndahl index; PERFORM median of the rivals average AR over 12 months prior to
the acquisition announcement; RELATED 1 if target has same two-digit SIC code as acquirer, 0
otherwise; PRIORP 1 if acquirer has prior presence in the US, 0 otherwise; XSTRENGTH
dierence of the average exchange rate of the dollar (domestic currency per US dollar) over 1984±
1996 and the exchange rate in the year of the announcement divided by the average exchange rate
of the dollar. VIF statistics are reported to demonstrate that multicollinearity is not substantially
in¯uencing the coecients.
The valuation eects are greater for rival portfolios when their stock returns
are highly correlated with the returns of the target (CORR). To the extent that
an oligopolistic rival reaction is anticipated, rivals with similar cash ¯ows may
be candidates for future takeovers.
The stock price eects of the rivals are directly related to the stock price
eects of the targets (TARGAR). This intra-industry eect reveals that the
target response is also valuable for the rivals.
Rival portfolios experience positive and signi®cant eects in technologicallyintense (DUMTECH) industries. Thus, it would appear that technologicallyintense industries bene®t from the possibility of future takeover activity.
Rivals with poor performance prior to the acquisition (PERFORM) are
favorably aected by the acquisitions. This information eect reveals the expectation of future performance improvements either due to the threat of
takeover or an actual takeover.
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
1319
RELATED is found to be positive and signi®cant which is consistent with
the information eect. Related acquisitions can signal consolidation in the
industry and indicate that further acquisitions are imminent.
Consistent with the competitive hypothesis, PRIORP is signi®cant and
negative. Hence, foreign acquirers that have already established a US presence
evidently adversely aect their US rivals. This relationship may occur if the
inherent disadvantages from operating abroad that Caves (1971) discusses have
been alleviated through their experience, allowing the foreign acquirer to now
focus on their competitive advantages.
In the regression results, FLEV is not shown to be statistically signi®cant. 22
However, when only those ®rms with the greatest degrees of ®nancial leverage
are examined, a signi®cant and negative relationship is detected. Thus, there is
some support for the competitive eects of ®nancial leverage.
5. Summary
This study provides evidence that, on average, the stock price response of
the rivals of US targets of foreign acquisitions is positive and signi®cant. These
results indicate that the information-signaling eects resulting from crossborder acquisitions outweigh the competitive eects. These empirical results
are informative since the DFI theories argue that industry rivals may either
bene®t or be adversely aected by direct foreign investment.
The cross-sectional analysis ®nds evidence supporting both the information-signaling and competitive hypotheses. Consistent with the informationsignaling hypothesis, the stock price response of target rival ®rms is more
favorable for relatively small rivals, for rivals with stock returns that are
strongly correlated with the stock returns of the target, when the targets
experience favorable valuation eects, in technologically-intense industries,
for rivals with poor prior performance, and with related acquisitions. Consistent with the competitive hypothesis, the stock price response of target
rival ®rms is less positive when the industry has a high degree of ®nancial
leverage and when the acquirers already have an established presence in the
US.
22
COMP is also not statistically signi®cant. Although, not detecting a strong relationship
between the degree of industry competition and the rival eects was anticipated due to osetting
forces.
1320
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
References
Atiase, R., 1985. Predisclosure information, ®rm capitalization, and security price behavior around
earnings announcements. Journal of Accounting Research 23, 21±36.
Caves, R.E., 1971. International corporations: The industrial economics of foreign investment.
Economica 38, 1±27.
Cebenoyan, A.S., Papaioannou, G.J., Travlos, N.G., 1992. Foreign takeover activity in the US and
wealth eects for target ®rm shareholders. Financial Management (Autumn), 58±68.
Doyle, P., Saunders, J., Wong, V., 1992. Competition in global markets: A case study of American
and Japanese competition in the British market. Journal of International Business Studies, 419±
442.
Dunning, J.H., 1988. The eclectic paradigm of international products: A restatement and some
possible extensions. Journal of International Business Studies, 1±31.
Eun, C.S., Kolodny, R., Scheraga, C., 1996. Cross-border acquisitions and shareholder wealth:
Tests of the synergy and internationalization hypotheses. Journal of Banking and Finance 20,
1559±1582.
Flowers, E.B., 1976. Oligopolistic reactions in European and Canadian direct investment in the
United States. Journal of International Business Studies, 43±55.
Froot, K.A., Stein, J.C., 1991. Exchange rates and foreign direct investment: An imperfect capital
markets approach. Quarterly Journal of Economics 106 (November), 1191±1217.
Harris, R.S., Ravenscraft, D., 1991. The role of acquisitions in foreign direct investment: Evidence
from the US stock market. Journal of Finance 46, 825±844.
Hennart, J., Park, Y., 1993. Green®eld vs. acquisition: The strategy of Japanese investors in the
United States. Management Science (September), 1054±1070.
Hertzel, M.G., 1991. The eects of stock repurchases on rival ®rms. Journal of Finance 46, 707±
716.
Hymer, S.H., 1960. International operations of national ®rms ± a study of direct foreign
investment. Dissertation. Massachusetts Institute of Technology, Cambridge, MA.
Kang, J., 1993. The international market for corporate control: Mergers and acquisitions of US
®rms by Japanese ®rms. Journal of Financial Economics, 34 (December), 345±371.
Knickerbocker, F., 1974. Oligopolistic Reaction and Multinational Enterprises. Harvard University Press, Cambridge, MA.
Lang, L.H.P., Stulz, R.M., 1992. Contagion and competitive intra-industry eects of bankruptcy
announcements. Journal of Financial Economics 12, 45±60.
Marquardt, D., Snee, R., 1975. Ridge regression in practice. American Statistician 29, 3±19.
Mikkelson, W.H., Partch, M., 1988. Withdrawn security oerings. Journal of Financial and
Quantitative Analysis (June), 119±134.
Pettway, R.H., Sicherman, N.W., Spiess, D.K., 1993. Japanese foreign direct investment: Wealth
eects from purchases and sales of US assets. Financial Management (Winter), 82±95.
Scholes, M.S., Wolfson, M.A., 1990. The eects of changes in tax laws on corporate reorganization
activity. Journal of Business (January), s141±s164.
Servaes, H., Zenner, M., 1994. Taxes and the returns to foreign acquisitions in the United States.
Financial Management (Winter), 42±56.
Shapiro, A.C., 1996. Multinational Financial Management. Prentice-Hall, Upper Saddle River, NJ.
Slovin, M.B., Sushka, M.E., Bendeck, Y.M., 1991. The intra-industry eects of going-private
transactions. Journal of Finance 46, 1537±1549.
Slovin, M.B., Sushka, M.E., Polonchek, J.A., 1992. Informational externalities of seasoned equity
issues. Journal of Financial Economics 32, 87±101.
Stulz, R.M., 1990. Managerial discretion and ®nancing policies. Journal of Financial Economics
26, 3±27.
Sudharshan, D., 1995. Marketing Strategy. Prentice-Hall, Englewood Clis, NJ.
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
1321
Szewczyk, S.H., 1992. The intra-industry transfer of information inferred from announcements of
corporate security oerings. Journal of Finance 47, 1935±1945.
Vernon, R., 1966. International investment and international trade in the product cycle. Quarterly
Journal of Economics 81 (May), 190±207.
www.elsevier.com/locate/econbase
Information-signaling and competitive eects of
foreign acquisitions in the US
Aigbe Akhigbe a, Anna D. Martin
a
b,*
Frederick W. Moyer Chair in Finance, College of Business Administration, The University
of Akron, Akron, OH 44325-4805, USA
b
School of Business, Fair®eld University, Fair®eld, CT 06430, USA
Received 14 September 1998; accepted 17 June 1999
Abstract
The stock price eects for the domestic competitors of foreign acquisition targets in
the US are found to be signi®cantly positive. These results imply that signals of favorable industry conditions conveyed through cross-border acquisitions dominate any
perceived changes in competitive balance. Consistent with the information-signaling
hypothesis, the stock price eects are more favorable for relatively small competitors,
for rivals with stock returns that are highly correlated with the targetÕs stock returns,
when the targets experience favorable stock price eects, in technologically-intense industries, for rivals with poor prior performance, and with related acquisitions. Consistent with the competitive hypothesis, the stock price eects are less favorable with
high degrees of ®nancial leverage and when the acquirers already have a presence in the
US. Ó 2000 Elsevier Science B.V. All rights reserved.
JEL classi®cation: F23; G34
Keywords: Foreign acquisitions; Cross-border acquisitions; Intra-industry eects; Rivals
*
Corresponding author. Tel.: +203-254-4000 ext. 2881; fax: +203-254-4105.
E-mail address: amartin@fair1.fair®eld.edu (A.D. Martin).
0378-4266/00/$ - see front matter Ó 2000 Elsevier Science B.V. All rights reserved.
PII: S 0 3 7 8 - 4 2 6 6 ( 9 9 ) 0 0 0 7 4 - 6
1308
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
1. Introduction
The stock price eects of cross-border acquisitions have received much attention in recent years (e.g., Harris and Ravenscraft, 1991; Kang, 1993; Pettway et al., 1993; Servaes and Zenner, 1994; Eun et al., 1996). These studies
focus on the stock price response of the target and/or acquirers and show that
shareholders of US targets gain from foreign acquisitions while the shareholders of the foreign acquirers do not consistently gain or lose.
To date, however, there has been no assessment of the stock price eects for
the competitors of foreign acquisition targets. Consider the agreement by
Daimler-Benz to acquire Chrysler in a stock transaction valued at about US$35
billion on May 7, 1998. This combination was expected to enhance ChryslerÕs
competitiveness by giving it access to German engineering prowess, and potentially reducing the cash¯ows to Ford and General Motors. Upon announcement,
the market price of Chrysler rose sharply, and the market prices of Ford and
General Motors fell at the news of the proposed acquisition. Thus, an announced
acquisition of a US ®rm by a foreign ®rm could be expected to aect the equity
values of the target as well as its local rivals. This paper estimates the impact of
foreign acquisitions on the US target ®rmÕs industry rivals and investigates
various factors that may explain why those eects dier across acquisitions.
Two competing hypotheses are proposed to describe the stock price response of the targetÕs rivals. Under the information-signaling hypothesis, foreign acquisitions would indicate that further acquisitions are imminent
(Knickerbocker, 1974; Flowers, 1976) and domestic competitors experience a
positive stock price reaction as their probability of becoming a takeover target
increases. In contrast, under the competitive hypothesis, ®rms that make crossborder investments possess sucient competitive advantages to more than
oset the inherent disadvantages (Caves, 1971) and domestic competitors experience a negative stock price reaction as the entrance of the foreign acquirers
into the industry increases the degree of competition they face.
This study shows that the domestic competitors experience signi®cantly
positive valuation eects upon announcement of foreign acquisitions of US
®rms. Overall, these results provide evidence that signals of favorable industry
conditions conveyed through cross-border acquisitions dominate any perceived
changes in competitive balance. The evidence in this study is important in light
of the direct foreign investment (DFI) theories that argue that DFI may either
bene®t or adversely aect competitors.
2. Theory and testable hypotheses
Although a single theory to explain DFI has not been developed, the industrial organization view helps to understand the motivation for investing
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
1309
abroad (e.g., Hymer, 1960; Vernon, 1966; Caves, 1971; Flowers, 1976;
Dunning, 1988). These related theories emphasize that market imperfections
and the pursuit of leveraging or preserving competitive advantages are the
primary motivating factors in cross-border investments. Imperfections may
arise naturally from the operation of the ®rm, may be arti®cially induced by
government policies, or may occur in the ®nancial markets (Scholes and
Wolfson, 1990; Froot and Stein, 1991; Harris and Ravenscraft, 1991;
Servaes and Zenner, 1994). Examples of some advantages that may be exploited are economies of scale and scope, managerial talent, and technological expertise.
2.1. Information-signaling hypothesis
This hypothesis suggests that acquisitions convey information about the
possibility of further takeover activity within the industry, which should bene®t
the target ®rmÕs competitors. If foreign acquisitions are utilized by aggressive
®rms to expand geographically, oligopolistic rival reactions are likely to occur
(e.g., Knickerbocker, 1974; Flowers, 1976). Foreign ®rms are more likely to
aggressively pursue expansion eorts in the US when DFI is relatively appealing. 1 Thus, announcements of foreign acquisitions in the US may indicate
that US assets are relatively appealing and that further acquisitions in the US
are probable. Thus, it can be hypothesized that the stock price eects for the
domestic competitors of US targets of direct foreign investment are positive.
Several factors are examined to determine whether there are cross-sectional
dierences in the information eects.
Impact of the relative size of the rivals. Based on the work of Atiase (1985), it
can be argued that the information-signaling eects are inversely related to the
relative size of the rival (e.g., Slovin et al., 1991). Incremental information
should be conveyed about relatively small rivals at the time of the announcement, if relatively small ®rms are less closely followed by the market. In addition, to the extent that larger rivals are more dicult to acquire, the
likelihood of further acquisitions involving relatively large rivals is lower. 2
1
Direct foreign investment in the US may be relatively appealing for a variety of reasons.
First, US assets may be valuable to foreign ®rms due to market imperfections particular to
them in developing, exploiting, or preserving competitive advantages. For example, foreign
®rms may discover it less costly to coordinate their activities or more eective to identify local
product requirements with direct operations in the US. Secondly, foreign ®rms may ®nd US tax
laws relatively advantageous or may desire to avoid taris and quotas (Scholes and Wolfson,
1990; Servaes and Zenner, 1994). Thirdly, US assets appear less expensive when foreign
currencies are strong relative to the US dollar (Froot and Stein, 1991; Harris and Ravenscraft,
1991).
2
We thank an anonymous reviewer for strengthening this hypothesis.
1310
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
Thus, it is hypothesized that the information eects are greater (smaller) for
relatively small (large) rivals.
Impact of similar cash ¯ows. The degree of similarity between the cash
¯ows of the rivals and the cash ¯ows of the targets may help dierentiate
the rival eects. The announcement of the acquisition indicates that the
operating structure of the target is desirable. 3 If an oligopolistic rival reaction is anticipated, rivals with comparable operating structures should be
candidates for future takeovers and experience more favorable information
eects. 4
Impact of other factors. Incremental information about the future prospects
for an industry is revealed in the CARs that accrue to the target and should
also have value for the industry rivals. 5 Information eects may be greater for
rivals in industries that are more attractive to foreign ®rms, thus related acquisitions may signal industry consolidation and indicate further acquisitions
within the industry.
2.2. Competitive hypothesis
This hypothesis is based on the idea that acquisitions may adversely
aect the future performance of the targetÕs rivals. DFI theories imply that
®rms must possess substantial competitive advantages to more than oset
their inherent disadvantages from operating abroad, such as managing
geographically dispersed operations and conducting business in an unfamiliar culture and environment (e.g., Caves, 1971). 6 To the extent that
foreign acquirers overcome their inherent disadvantages with convincing
3
Lang and Stulz (1992) present a similar argument where the information eect is expected to be
larger when rivals have similar cash ¯ows.
4
If the current acquisition is an oligopolistic reaction to a previous acquisition, rivals with
similar cash ¯ows may be viewed as the ``losing'' targets, even if the rivals had not been formally or
publicly under consideration. In this case, rivals with similar cash ¯ows would experience less
favorable information eects.
5
Hertzel (1991) implies that the larger the impact of announcements on the targets the greater
the information signal.
6
Foreign ®rms competing in the US may possess a variety of advantages over domestic ®rms.
First, the nature of ®nancial and managerial decisions made by foreign managers may dier from
domestic managers (e.g. Doyle et al., 1992). Secondly, foreign ®rms may more eectively
capitalize on basic research. Eun et al. (1996) provide some evidence that cross-border
acquisitions of US ®rms are synergistic due to research and development capabilities.
Additionally, the products of foreign ®rms are sometimes perceived to have superior quality
(e.g., Shapiro, 1996).
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
1311
advantages, resulting in a net advantage, the stock price eects for the
target ®rmÕs rivals are hypothesized to be negative. 7 Several factors are
investigated to determine whether there are cross-sectional dierences in the
competitive eects.
Impact of ®nancial leverage. Financial leverage may limit a ®rmÕs ability to
make investments to respond to competitive challenges (Stulz, 1990). Therefore, it is hypothesized that rivals with higher degrees of ®nancial leverage incur
more negative competitive eects due to their limited ability to expeditiously
compete.
Impact of the degree of competition. Another factor that is explored is the
degree of competition within the industry. It can be argued that rivals in highly
competitive industries experience greater competitive eects. Since rivals in
competitive industries are already aggressively competing, the additional
challenge from a foreign rival with presumably net competitive advantages may
have a more detrimental impact. However, a detrimental impact on the rivals
may also occur in less competitive industries, but for a dierent reason. Industries with a low degree of competition tend to participate in collusive behavior (Sudharshan, 1995). In this environment, the foreign competitors may
not be as likely to join the coordination eorts, thus may adversely impact the
rivals. 8
Impact of the rivals' prior performance. Rivals that have had poor performance prior to the acquisition are not likely to be positioned to eectively
compete with a new entrant. Thus, it can be hypothesized that underperforming rivals are more adversely aected by the acquisition. Alternatively,
underperforming rivals may be expected to improve their performance due to
the threat of an acquisition or an actual acquisition. Upon announcement, this
information-signal may favorably impact the returns of poor performing rivals.
Impact of other factors. Related acquisitions are more likely to achieve
economies of scale or scope (e.g., Caves, 1971). Foreign acquirers with prior
US experience may not suer as greatly from the inherent disadvantages from
operating abroad and instead are able to focus on their competitive strengths,
and the real or perceived competitive threat may be greater depending upon the
acquirerÕs country of origin.
7
Hennart and Park (1993) argue that acquisitions of existing US operations, as opposed to
establishing new US subsidiaries, are used by Japanese ®rms that do not have strong competitive
advantages. If acquisitions are generally used by foreign ®rms that do not have strong competitive
advantages, the competitive eects described here may not be elicited.
8
If the rival ®rms in less competitive industries are expected to focus on improving their
performance as a result of new competition, the rival eects would be positive.
1312
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
3. Sample characteristics and valuation methodology
3.1. Sample characteristics
To be included in the sample of foreign acquisitions made in the US, the
following selection and screening techniques are used:
1. Foreign acquisitions of US ®rms are listed in Mergers and Acquisitions with
completion dates between 1985 and 1996.
2. The bidder gains at least 50% control of the target.
3. The target has stock price data available on NYSE, ASE, or NASDAQ
CRSP ®les at the time of the acquisition announcement.
4. Acquisition announcement dates are available in Predicast's F & S. The announcements are screened for confounding events. Some announcement
dates are in 1984 for acquisitions completed in 1985.
5. The announcements are made in daily publications such as the Wall Street
Journal, New York Times, Financial Times, etc.
The ®nal sample consists of 165 acquisition announcements over 1984±1996.
A wide range of industries are represented in the sample with 119 distinct fourdigit SIC codes included. Table 1 summarizes the sample distribution by event
year and country. Note that 94 (57%) of the 165 acquisitions occurred in the
late 1980s (1986±1989). British ®rms conducted the most acquisitions (40%) in
this sample.
Table 2 provides further summary statistics on the targets and their industry
rivals. These statistics are provided for three dierent de®nitions of industry
rivals, US ®rms that share the same four-digit, three-digit, and two-digit SIC
code with the target at the time of the acquisition announcement. The SIC
codes are provided in the CRSP ®les. 9 The mean market value of the targets is
$1.1 billion, while the mean market value of the rivals ranges from $1.4 billion
(two-digit SIC code de®nition) to $2.0 billion (four-digit SIC code de®nition).
The mean, median, minimum, and maximum number of rivals per event are
also provided. As expected, the mean, median, and maximum number of rivals
increase as the de®nition of ÔrivalÕ is broadened.
3.2. Valuation methodology
The valuation eects of foreign acquisitions in the US are estimated for
equally-weighted portfolios of the industry rivals for each announcement. A
simple market model is used to predict daily share price returns, with a 200day estimation period from 220 days before the announcement to 20 days
before the announcement for each portfolio of rivals. Daily returns for the
9
We ®nd qualitatively similar results when the SIC codes from Compustat are used.
Country
Number of acquisitions
1984±
1996
England
Canada
France
Japan
Germany
Switzerland
Netherlands
Otherb
Total
a
1984
1985
1986
1987
1988
1989
1990
66
18
15
13
12
10
8
23
1
2
2
0
0
1
1
1
2
0
1
0
0
0
0
1
12
5
2
0
1
1
2
4
13
0
0
0
0
1
0
2
165
8
4
27
16
1991
1992
1993
1994
10
1
2
3
2
2
0
4
6
6
1
7
1
1
1
4
24
27
1995
1996
5
0
4
1
2
0
0
1
3
0
1
0
0
0
0
1
1
0
0
1
0
0
0
0
3
0
0
1
1
0
0
1
5
0
0
0
2
2
0
0
3
2
1
0
2
2
2
3
2
2
1
0
1
0
2
1
13
5
2
6
9
15
9
The sample of 165 foreign acquisitions of US ®rms is derived from Mergers and Acquisitions with completion dates between 1985 and 1996, where the
bidder gains at least 50% control of the target, the target has stock price data available, and announcement dates are available in Predicast's F & S from
daily publications.
b
Includes acquirers from Argentina, Australia, Belgium, Ireland, Israel, Italy, Mexico, South Korea, Sweden, and Taiwan.
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
Table 1
The sample distribution of 165 foreign acquisitions of US targets by year and by countrya
1313
1314
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
Table 2
Summary statistics of 165 US targets acquired by foreign ®rms and their corresponding industry
rivals, de®ned by CRSP four-digit, three-digit and two-digit SIC codes
Mean market value of target ($million)
Mean market value of rivals ($million)
Mean number of rivals per acquisition
Median number of rivals per acquisition
Minimum number of rivals per acquisition
Maximum number of rivals per acquisition
Four-digit
SIC
Three-digit
SIC
Two-digit
SIC
1101.78
1954.16
15.8
7
1
124
1101.78
1923.80
27.1
18
1
159
1101.78
1388.81
99.3
78
1
638
rivals are taken from the CRSP ®les to create an equally-weighted rival
portfolio. The CRSP equally-weighted index is used as a proxy for the
market.
Daily abnormal returns are then calculated over an examination period of
21 days, from 10 days pre-announcement to 10 days post-announcement.
Based on the hypotheses in this study, positive abnormal returns support the
information-signaling argument and negative abnormal returns support the
competitive argument. z-Statistics are calculated using the methodology of
Mikkelson and Partch (1988).
4. Results
4.1. Valuation eects
Table 3 reports the valuation eects for US targets acquired by foreign ®rms
and their corresponding rival portfolios. Panel A reports the valuation eects
for the 165 targets. On average, the 2-day event period cumulative AR (CAR)
is 23.39% and signi®cant at the 0.001 level. This result is consistent with the
aforementioned studies that document favorable valuation eects for targets
upon the announcements of acquisitions.
The valuation eects of the rival portfolios are provided in Panels B, C,
and D of Table 3. 10 These panels report the valuation eects for the rival
portfolios using three dierent de®nitions for ÔrivalÕ. When the rivals are
de®ned as those ®rms that share the same four-digit SIC code (Panel B), the
2-day event period mean CAR is 0.50% and signi®cant at the 0.001 level. Panel
C (Panel D) reports the valuation eects as 0.48% (0.23%) when the rivals are
de®ned as those ®rms that share the same three-digit (two-digit) SIC code.
10
Excluding losing bidders from the rival portfolios does not signi®cantly change the rival
portfolio eects.
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
1315
Table 3
Valuation eects for US targets acquired by foreign ®rms and their corresponding industry rival
portfolios, de®ned by CRSP four-digit, three-digit and two-digit SIC codesa
Examination period
CAR
z-Statistics
% Pos
12.11d
80.95d
0.19
63
91
47
Panel B: industry rivals of US targets, de®ned by four-digit SIC
Pre-event period ()11,)2)
0.0006
)0.71
Event period ()1,0)
0.0050
3.61d
Post-event period (+1,+10)
)0.0009
0.13
46
56
47
Panel C: industry rivals of US targets, de®ned by three-digit SIC
Pre-event period ()11,)2)
0.0038
0.78
Event period ()1,0)
0.0048
3.53d
Post-event period (+1,+10)
)0.0010
0.39
47
58
51
Panel D: industry rivals of US targets, de®ned by two-digit SIC
Pre-event period ()11,)2)
0.0013
0.95
Event period ()1,0)
0.0023
3.55d
Post-event period (+1,+10)
0.0014
0.44
51
59
49
Panel A: US targets acquired by foreign ®rms
Pre-event period ()11,)2)
0.0698
Event period ()1,0)
0.2339
Post-event period (+1,+10)
)0.0020
a
The reported CARs are cumulative abnormal returns over three dierent examination periods. d
denotes signi®cance at the 0.001 level. The average abnormal return and z-statistics can dier in
sign because the former assigns uniform weights to each observation and the latter assigns nonuniform weights (see Mikkelson and Partch, 1988).
Thus, it appears the rival stock price eects are somewhat robust to the de®nition of ÔrivalÕ. Although, the eects are reduced as a more broad de®nition of
rival is used.
In this study, industry rivals will be primarily de®ned as those ®rms with the
same four-digit SIC code as the target (Hertzel, 1991; Slovin et al., 1991; 1992;
Szewczyk, 1992). Industry rivals de®ned in this manner are available for 120 of
the 165 acquisitions. In those cases where there are no rival ®rms with the same
four-digit SIC code at the time of the announcement, the rivals are identi®ed
using ®rms with the same three-digit SIC code. 11 Using this approach, the 2day event period mean CAR is 0.44% and signi®cant at the 0.001 level. 12
The results in Table 3 indicate that the information-signaling eects dominate the competitive eects. This evidence is interesting in light of the DFI
theories that suggest cross-border acquisitions could either bene®t or adversely
aect competitors.
11
This scheme is also applied to generating the cross-sectional variables.
The distribution of the 2-day CARs does not reveal a majority of ®rms with large, positive
ARs. Of the 165 rival portfolios, 95 accumulate positive CARs, 70 accumulate negative CARs, and
86 accumulate between plus and minus 1%.
12
1316
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
4.2. Cross-sectional regression analysis of rival stock price eects
Dierences in the valuation eects across the domestic competitors are explored using regression analysis. Based on the information-signaling hypothesis, we expect a more positive rival stock price response for relatively small
rivals, for rivals with similar cash ¯ows, when the targets experience positive
stock price eects, and for rivals in industries that are more attractive to foreign ®rms. Based on the competitive hypothesis, we expect a more negative
rival stock price response for highly leveraged rivals, for rivals operating in
industries with a high or low degree of competition, for poorly performing
rivals, and when the acquirer already has an established presence in the US.
Relative size (RELSIZE). This variable takes on the value of one for relatively small rivals, and zero otherwise. 13 Market values are calculated as the
number of common shares outstanding multiplied by the share price 14
Correlation in stock returns (CORR). As a proxy for the similarity in cash
¯ows, the correlation between the stock returns of the target and the rival
portfolio is calculated over the 200-day period (tÿ220 to tÿ20 ) prior to the acquisition announcement.
Target abnormal return (TARGAR). The 2-day cumulative abnormal return
for each target is calculated in the same manner as the rival portfolio.
Financial leverage (FLEV). This industry-speci®c variable is de®ned as the
median ratio of total long-term debt to total equity of the rivals. 15
Degree of competition (COMP). The Her®ndahl index is used as a proxy for
the degree of competition. It is de®ned as the squared sum of the fractions of
industry market capitalization by the rival ®rms. 16 An industry is considered
less competitive with a greater Her®ndahl index.
Performance (PERFORM). The average monthly abnormal return for each
rival is estimated over the 12 month period prior to the acquisition announcement. The median AR of the rivals is used to indicate the prior performance of the industry rivals.
13
A continuous variable, measured as the ratio of the market value of the target to the median
market value of the rivals, is also considered. However, after removing outliers, the results are
essentially the same as those reported.
14
The information eects may not be driven by the size of the rivals, but instead it may be driven
by the size of the targets. Szewczyk (1992) implies that acquisitions of large targets convey more
information than acquisitions of small targets.
15
These data are taken from Compustat. In order to generate the ®nancial leverage variable for
the industry rivals, a two-digit SIC code de®nition is used in seven cases, a three-digit SIC code
de®nition is used in 61 cases, and a four-digit SIC code de®nition is used in 97 cases.
16
The results are qualitatively the same when COMP is de®ned as the squared sum of the
fractions of industry sales.
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
1317
Currency strength (XSTRENGTH). The relative strength of the bidderÕs
currency is used as a control variable. This variable is estimated following
Harris and Ravenscraft (1991), Cebenoyan et al. (1992). Positive (negative)
values of XSTRENGTH occur when the bidderÕs currency is weak (strong)
relative to the US dollar. 17 The exchange rates are in units of foreign currency
per US dollar and are available in the Wall Street Journal.
Dummy variables. The rival eects are also examined for cross-sectional
variation by industry, 18 the relatedness of the acquisition at the two-digit SIC
code level, and the acquirerÕs prior presence in the US 19 using dummy variables. 20
Table 4 displays the results of the regression analysis. To control for heteroskedasticity, the variables have been standardized by the standard error of
the model used to estimate the abnormal returns. The variance in¯ation factors
(VIFs) are reported to assess the extent of multicollinearity. The VIFs range
from 1.16 to 1.93, indicating multicollinearity is not substantially in¯uencing
the coecients. 21
The cross-sectional results provide evidence supporting both the information-signaling and competitive hypotheses. More speci®cally, RELSIZE is
found to be positive and signi®cant. As hypothesized, the information eects
are greater (smaller) for relatively small (large) rivals.
17
In measuring XSTRENGTH, the exchange rate of the year of the announced acquisition is
used if the announcement occurred in the second half of the year. The exchange rate of the year
prior to the announcement is used if the announcement occurred in the ®rst half of the year (Harris
and Ravenscraft, 1991).
18
Each industry is classi®ed into one of three groups (technology-intensive, other manufacturing-oriented, or service-oriented) following Cebenoyan et al. (1992). DUMTECH has a value of 1
for technologically intense industries (two-digit SIC codes 28 (chemicals), 33 and 34 (metals), 35
(machinery), 36 (electrical), and 38 (scienti®c instruments)) and 0 otherwise; DUMIND has a value
of 1 for manufacturing-oriented but not technologically intense industries (two-digit SIC codes 10
(mining), 13 (oil/gas), 14 (non-metallic minerals), 20 and 39 (manufacturing), 22 (textile mills), 26
(paper), 27 (printing/publishing), 29 (petroleum/coal), 30 (rubber/plastics), 32 (stone/clay/glass),
and 37 (transportation equipment)), and 0 otherwise). The eects of the service-oriented industries
(two-digit SIC codes 47, 73, 80, 82, and 87 (services), 48 (communication), 50 and 51 (wholesale),
53, 54, and 54 (retail), 60, 62, 63, and 64 (®nance/insurance/real estate), 67 (holding cos.), 70
(hotels), and 78 (motion pictures)) are captured by the intercept.
19
Prior presence is established by reviewing the Directory of Corporate Aliations and Standard
and Poor's Register for US operating locations at the time of the acquisition.
20
When the acquisitions are grouped by the country of the acquirer, no speci®c country reveals
signi®cant mean CARs. When the acquisitions are grouped by year, the acquisitions in 1994 reveal
a signi®cant (p < 0.05) mean CAR of 1.18%. Since there is little evidence of country and year eects,
these dummy variables are excluded.
21
Some studies use VIF > 5 as an indication of severe multicollinearity (see Marquardt and Snee,
1975). Furthermore, univariate regressions show the magnitude and signi®cance of the coecients
are stable.
1318
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
Table 4
Cross-sectional factors of the rival eects of foreign acquisitions in the USa
Variables
Coecients
Intercept
RELSIZE
CORR
TARGAR
DUMTECH
DUMIND
FINLEV
COMP
PERFORM
RELATED
PRIORP
XSTRENGTH
)0.0193
0.0148
0.0663
0.0381
0.0165
0.0056
0.0203
)0.0187
)0.0502
0.0278
)0.0214
)0.0132
Sample size
F-value
Adj. R2
t-Statistics
)1.41
1.92a
3.58d
2.84c
2.09b
0.53
0.76
)1.48
)4.32d
3.43d
)2.56c
)0.52
VIF
0.00
1.16
1.36
1.28
1.50
1.93
1.37
1.28
1.29
1.49
1.50
1.51
165
12.28d
0.4321
a
The dependent variable is the rival portfolio two-day cumulative AR. a, b, c, and d denote signi®cance at the 0.10, 0.05, 0.01, and 0.001 levels, respectively. RELSIZE 1 if target market value>median market value of rivals, 0 otherwise; CORR median correlation between target and
rivals stock returns over tÿ220 and tÿ20 ; TARGAR target two-day CAR; DUMTECH 1 if target
industry is technologically-intense, 0 otherwise; DUMIND 1 if target industry is manufacturing
but not technologically-intense, 0 otherwise; FINLEV median debt/equity of rivals;
COMP Her®ndahl index; PERFORM median of the rivals average AR over 12 months prior to
the acquisition announcement; RELATED 1 if target has same two-digit SIC code as acquirer, 0
otherwise; PRIORP 1 if acquirer has prior presence in the US, 0 otherwise; XSTRENGTH
dierence of the average exchange rate of the dollar (domestic currency per US dollar) over 1984±
1996 and the exchange rate in the year of the announcement divided by the average exchange rate
of the dollar. VIF statistics are reported to demonstrate that multicollinearity is not substantially
in¯uencing the coecients.
The valuation eects are greater for rival portfolios when their stock returns
are highly correlated with the returns of the target (CORR). To the extent that
an oligopolistic rival reaction is anticipated, rivals with similar cash ¯ows may
be candidates for future takeovers.
The stock price eects of the rivals are directly related to the stock price
eects of the targets (TARGAR). This intra-industry eect reveals that the
target response is also valuable for the rivals.
Rival portfolios experience positive and signi®cant eects in technologicallyintense (DUMTECH) industries. Thus, it would appear that technologicallyintense industries bene®t from the possibility of future takeover activity.
Rivals with poor performance prior to the acquisition (PERFORM) are
favorably aected by the acquisitions. This information eect reveals the expectation of future performance improvements either due to the threat of
takeover or an actual takeover.
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
1319
RELATED is found to be positive and signi®cant which is consistent with
the information eect. Related acquisitions can signal consolidation in the
industry and indicate that further acquisitions are imminent.
Consistent with the competitive hypothesis, PRIORP is signi®cant and
negative. Hence, foreign acquirers that have already established a US presence
evidently adversely aect their US rivals. This relationship may occur if the
inherent disadvantages from operating abroad that Caves (1971) discusses have
been alleviated through their experience, allowing the foreign acquirer to now
focus on their competitive advantages.
In the regression results, FLEV is not shown to be statistically signi®cant. 22
However, when only those ®rms with the greatest degrees of ®nancial leverage
are examined, a signi®cant and negative relationship is detected. Thus, there is
some support for the competitive eects of ®nancial leverage.
5. Summary
This study provides evidence that, on average, the stock price response of
the rivals of US targets of foreign acquisitions is positive and signi®cant. These
results indicate that the information-signaling eects resulting from crossborder acquisitions outweigh the competitive eects. These empirical results
are informative since the DFI theories argue that industry rivals may either
bene®t or be adversely aected by direct foreign investment.
The cross-sectional analysis ®nds evidence supporting both the information-signaling and competitive hypotheses. Consistent with the informationsignaling hypothesis, the stock price response of target rival ®rms is more
favorable for relatively small rivals, for rivals with stock returns that are
strongly correlated with the stock returns of the target, when the targets
experience favorable valuation eects, in technologically-intense industries,
for rivals with poor prior performance, and with related acquisitions. Consistent with the competitive hypothesis, the stock price response of target
rival ®rms is less positive when the industry has a high degree of ®nancial
leverage and when the acquirers already have an established presence in the
US.
22
COMP is also not statistically signi®cant. Although, not detecting a strong relationship
between the degree of industry competition and the rival eects was anticipated due to osetting
forces.
1320
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
References
Atiase, R., 1985. Predisclosure information, ®rm capitalization, and security price behavior around
earnings announcements. Journal of Accounting Research 23, 21±36.
Caves, R.E., 1971. International corporations: The industrial economics of foreign investment.
Economica 38, 1±27.
Cebenoyan, A.S., Papaioannou, G.J., Travlos, N.G., 1992. Foreign takeover activity in the US and
wealth eects for target ®rm shareholders. Financial Management (Autumn), 58±68.
Doyle, P., Saunders, J., Wong, V., 1992. Competition in global markets: A case study of American
and Japanese competition in the British market. Journal of International Business Studies, 419±
442.
Dunning, J.H., 1988. The eclectic paradigm of international products: A restatement and some
possible extensions. Journal of International Business Studies, 1±31.
Eun, C.S., Kolodny, R., Scheraga, C., 1996. Cross-border acquisitions and shareholder wealth:
Tests of the synergy and internationalization hypotheses. Journal of Banking and Finance 20,
1559±1582.
Flowers, E.B., 1976. Oligopolistic reactions in European and Canadian direct investment in the
United States. Journal of International Business Studies, 43±55.
Froot, K.A., Stein, J.C., 1991. Exchange rates and foreign direct investment: An imperfect capital
markets approach. Quarterly Journal of Economics 106 (November), 1191±1217.
Harris, R.S., Ravenscraft, D., 1991. The role of acquisitions in foreign direct investment: Evidence
from the US stock market. Journal of Finance 46, 825±844.
Hennart, J., Park, Y., 1993. Green®eld vs. acquisition: The strategy of Japanese investors in the
United States. Management Science (September), 1054±1070.
Hertzel, M.G., 1991. The eects of stock repurchases on rival ®rms. Journal of Finance 46, 707±
716.
Hymer, S.H., 1960. International operations of national ®rms ± a study of direct foreign
investment. Dissertation. Massachusetts Institute of Technology, Cambridge, MA.
Kang, J., 1993. The international market for corporate control: Mergers and acquisitions of US
®rms by Japanese ®rms. Journal of Financial Economics, 34 (December), 345±371.
Knickerbocker, F., 1974. Oligopolistic Reaction and Multinational Enterprises. Harvard University Press, Cambridge, MA.
Lang, L.H.P., Stulz, R.M., 1992. Contagion and competitive intra-industry eects of bankruptcy
announcements. Journal of Financial Economics 12, 45±60.
Marquardt, D., Snee, R., 1975. Ridge regression in practice. American Statistician 29, 3±19.
Mikkelson, W.H., Partch, M., 1988. Withdrawn security oerings. Journal of Financial and
Quantitative Analysis (June), 119±134.
Pettway, R.H., Sicherman, N.W., Spiess, D.K., 1993. Japanese foreign direct investment: Wealth
eects from purchases and sales of US assets. Financial Management (Winter), 82±95.
Scholes, M.S., Wolfson, M.A., 1990. The eects of changes in tax laws on corporate reorganization
activity. Journal of Business (January), s141±s164.
Servaes, H., Zenner, M., 1994. Taxes and the returns to foreign acquisitions in the United States.
Financial Management (Winter), 42±56.
Shapiro, A.C., 1996. Multinational Financial Management. Prentice-Hall, Upper Saddle River, NJ.
Slovin, M.B., Sushka, M.E., Bendeck, Y.M., 1991. The intra-industry eects of going-private
transactions. Journal of Finance 46, 1537±1549.
Slovin, M.B., Sushka, M.E., Polonchek, J.A., 1992. Informational externalities of seasoned equity
issues. Journal of Financial Economics 32, 87±101.
Stulz, R.M., 1990. Managerial discretion and ®nancing policies. Journal of Financial Economics
26, 3±27.
Sudharshan, D., 1995. Marketing Strategy. Prentice-Hall, Englewood Clis, NJ.
A. Akhigbe, A.D. Martin / Journal of Banking & Finance 24 (2000) 1307±1321
1321
Szewczyk, S.H., 1992. The intra-industry transfer of information inferred from announcements of
corporate security oerings. Journal of Finance 47, 1935±1945.
Vernon, R., 1966. International investment and international trade in the product cycle. Quarterly
Journal of Economics 81 (May), 190±207.