Khả năng sinh lời của chứng khoán góc độ 2

DISCUSSION PAPER SERIES



ABCD

www.cepr.org


Available online at: www.cepr.org/pubs/dps/DP3998.asp and />
www.ssrn.com/xxx/xxx/xxx









No. 3998

CROSS-BORDER ACQUISITIONS
AND GREENFIELD ENTRY:
PROFITABILITY AND ST OCK
MARKET VALUE

Pehr-Johan Norbäck and Lars Persson


INDUSTRIAL ORGANIZATION



ISSN 0265-8003
CROSS-BORDER ACQUISITIONS
AND GREENFIELD ENTRY:
PROFITABILITY AND STOCK
MARKET VALUE
Pehr-Johan Norbäck, Research Institute for Industrial Economics (IUI), Stockholm
Lars Persson, Research Institute for Industrial Economics (IUI), Stockholm
and CEPR

Discussion Paper No. 3998
August 2003
Centre for Economic Policy Research
90–98 Goswell Rd, London EC1V 7RR, UK
Tel: (44 20) 7878 2900, Fax: (44 20) 7878 2999
Email: , Website: www.cepr.org
This Discussion Paper is issued under the auspices of the Centre’s research
programme in INDUSTRIAL ORGANIZATION. Any opinions expressed here
are those of the author(s) and not those of the Centre for Economic Policy
Research. Research disseminated by CEPR may include views on policy, but
the Centre itself takes no institutional policy positions.
The Centre for Economic Policy Research was established in 1983 as a
private educational charity, to promote independent analysis and public

discussion of open economies and the relations among them. It is pluralist
and non-partisan, bringing economic research to bear on the analysis of
medium- and long-run policy questions. Institutional (core) finance for the
Centre has been provided through major grants from the Economic and
Social Research Council, under which an ESRC Resource Centre operates
within CEPR; the Esmée Fairbairn Charitable Trust; and the Bank of
England. These organizations do not give prior review to the Centre’s
publications, nor do they necessarily endorse the views expressed therein.
These Discussion Papers often represent preliminary or incomplete work,
circulated to encourage discussion and comment. Citation and use of such a
paper should take account of its provisional character.
Copyright: Pehr-Johan Norbäck and Lars Persson
CEPR Discussion Paper No. 3998
August 2003
ABSTRACT
Cross-Border Acquisitions and Greenfield Entry:
Profitability and Stock Market Value*
This Paper studies cross-border acquisitions and greenfield entry in a multi-
firm setting. Acquisition entry is more likely when the acquirer gains a strong
position in the product market, relative to greenfield entrants. We also show
that such acquisitions might have a low profitability, however. The reason is
that the bidding competition over the domestic assets is then so fierce that the
firms involved would be better off not starting a bidding war. Moreover, this
implies that domestic firms will then sell their assets at a substantially higher
price than their reservation price. Implications for stock market values are also
derived.
JEL Classification: F23, G34 and L13
Keywords: FDI, mergers and acquisitions and stock market value
Pehr-Johan Norbäck
Research Institute of Industrial

Economics (IUI)
Box 5501
S-11485 Stockholm
SWEDEN
Tel: (46 8) 665 4522
Fax: (46 8) 665 4599
Email:

For further Discussion Papers by this author see:
www.cepr.org/pubs/new-dps/dplist.asp?authorid=153438
Lars Persson
Research Institute of Industrial
Economics (IUI)
Box 5501
S-11485 Stockholm
SWEDEN
Tel: (46 8) 665 4504
Fax: (46 8) 665 4599
Email:

For further Discussion Papers by this author see:
www.cepr.org/pubs/new-dps/dplist.asp?authorid=134788

*We are grateful for helpful discussions with Jonas Björnerstedt, S-O
Fridolfsson, Chiara Fumagalli, Mattias Ganslandt and Monika Schnitzer,
participants at IUI seminars and the 2001 SAET conference in Ischia.
Financial support from the Marianne and Marcus Wallenberg Foundation, and
Tom Hedelius’ and Jan Wallander’s Research Foundations, is gratefully
acknowledged.
Submitted 10 July 2003

1. Introduction
Multina tion al Enterprises (MN Es) are often associated with firm specific assets.
1
Suc h
assets, which includ e marketing ab ility, brand names, patents o r tec hn ology, enable th ese
firm s to profitably expand production abroad. W hen expanding business abroad, MNEs
canbasicallyusetwotypesofbusinessstrategies; they can e ith er a cquire (or m erg e with)
a firm in the host co untry or invest greenfield, i.e. set up a new p lant in th e h ost coun try.
Sev eral factors ha ve been argued to be important for e xplaining wh y MNEs p refer
to gro w via M &As rather than through organic gro wth: the quest for strategic or com-
plementary assets, such as b ra nd names, the possession s of local permits or d istrib ution
netw or ks and patents. When the time to mark et is vital, the takeover of an existing firm
with an established distributio n system m ig ht also b e preferable.
2
The importance o f
the d om estic assets will d epend on the strength of complementarities between MNEs’
firm -specific assets and the domestic assets. For example, the com b ination of a n MNE ’s
strong brand name and the acquired firm’s knowledge of the market or strength in dis-
tribution, may provide the acquiring M N E with a strong mark et position. In order to
capture th ese aspects, the dom estic assets will be said to b e strategically valuable when
combining an MNEs firm specific asse ts with the do m es tic assets gives t he acqu irer a
stron g position in the product mar ke t, re lative to greenfield entrants.
We present a model where we indeed sh ow that a h igh str ategic value of the dom estic
assets is conducive to acquisitions. Howev er, while acquisition entry is associated with
a high strategic value of the dome stic assets, we also show that su ch a cqu isitions might
hav e a lo w expected profita bility. In the model, a domestic firm is initia lly located in th e
dom estic market in country H. There are also sev eral symm etric M N E s located in the
world market. T h e domestic m arket will now be exposed to intern ational competition.
The interaction ta kes place in thr ee stages. In the firststage,theMNEsmightacquire
the dom estic firm’s assets. In the seco nd sta ge, MNEs h ave the option o f investing

1
See for instance, Caves (1996), Dunning (1977) and Markusen (1995).
2
See World Investment Report (WIR) 2000 and its reference to different studies of cross-border
M&As.
2
greenfield in new assets in country H, where green field entry is associated with a risk of
failure. Finally, in the third stage, firms compete in oligopoly fashion in country H.
Theresultthattheexpectedprofit of the acquirer (i.e. the expected product market
profit net of the acquisition price), ma y decrease in the strategic value of the dom estic
assets seems counter-intuitive a t first sight, since the domestic firm’s assets are then m ore
va luab le to the MNE s when acquired. Howev er , this result is in tu itive, when taking in to
account how the level of stra teg ic value of the domestic assets affects the acquisition
price. The price of the assets is a non-acquiring MN E ’s willingness to pay, whic h consist
of t wo profit terms: t he expected produ ct market profitforthisfirm if it were instead
to obta in the domestic firm’s assets net of the corresponding profit w hen not buying.
Itthenfollowsthatthefirst profit term increases to exactly the same exten t as that of
the acquirer from an in crease in the strategic value of the d om estic assets, and will thus
off-set the acquirer’s increase in profit. M oreover, the second profit term will decrease,
the more stra tegically valuable the domestic assets are, since the non-acquirer will then
face a stronge r competitor in the product market. This implies that the willingness to
pay fur ther increases for the non-acqu irer. Consequently, the a cquisition price in creases
more than the a cquirer’s p roduct market profit, when the dom estic assets becom e more
strategically valuable. We furth er sho w that the bidding co m petition over the domestic
firm implies th at this firm will sell its a ssets at a higher price than its reservation price,
or to a substantially higher price wh en the d om estic assets are su fficiently strategically
va luab le.
We then use our results on how expected profits are affected b y an a cquisition to
derive imp licatio ns for s tock market valu es for t he di fferen t entry m odes. Follow ing
the standard approach in the so-called ev en t studies on M&A performance, assum ing

that the merger com es as a surprise for th e financial markets, the empirical im p licatio n
is then that the ta rget firm’s sha reh olders benefit f r om the acqu isitio n and that th e
takeo ver premium is increasing in t he stra tegic im portance of the d omestic assets. T he
predictions for the poten tial acquiring firms are more in volved. H owev er, w e find that
the share value of both the acquirer and a non-acquirer will decrease w hen an acquisition
is announced, if the domestic assets a r e sufficiently strategically valuab le. This is due to
3
the fact that the bidding com petition is then so fierce that the firms involv ed wou ld be
be tter off not starting a bidding war.
IntheliteratureonMNEs
3
, it has also been argued that one of the main benefits from
acquiring a local competitor instead of en tering greenfield is that the acquisition helps
the firm avoid risks due to lack of knowledge of the specific characteristics of the local
market. By entering by an M & A , a n M N E avoids the individual risk of un successfu l
greenfield entry. Ho wever, w e show that the bidding competition over being successfully
loca ted in the mark et with certaint y may driv e up the acquisition price to suc h a lev el
that being a greenfield entrant is more profitable ex-post. B ut firmsalsofaceamarket
risk as realized product market profitforafirm may differ from the expected one. It
is th en sho w n that if firms a re sufficiently s ym m e tric, the stoc k market value of the
acquirer is reduced relative to that of a successful greenfield entrant, since the market
risk will then be similar between the two entry mod es and only the individual risk differ .
However, if competition becomes s ufficiently softer than expected and the strategic value
is sufficiently high, the aquirer’s stoc k m arket valu e will in crease relative to that of a
successful green field entrant.
The related theoretical lit erature on foreign direct in vestm ent FDI and MNE s is sur-
v e yed in Markusen (1995). Th is literatu re does no t explicitly address th e q uestion of
wheth er en t ry in to a foreign market is greenfield or th ro ugh the acquisition of a ssets
already in the mark et, or both, a n issue wh ich is at focus i n our study, h owev er.
4

There
is also a recent theoretical literature addressing aspects of cross-border mergers in in-
ternational oligopoly mark ets.
5
However, the equilib riu m acqu isition price, wh ich is in
focus in our study, is not determined in those studies.
3
See Caves (1996).
4
See, Das and Sengupta ( 2001), Görg (1997) and Norback and Persson (2003) for papers addressing
the choice of entry mode. However, these papers abstract from the competition between MNEs, which
is at focus in our study. Bjorvatn (2001) allows for competition between MNEs, but studies how the
pattern and profitability of cross-border acquisitions depend on trade cost and greenfield cost.
5
This literature includes papers by, for example, Head and Reis (1997), Horn and Persson (2001),
Lommerud, Straume and Sorgard (2003), Neary (2003), Norbäck and Persson (2002), Saggi and Yildiz
(2002), Straume (2003), Yildiz (2003).
4
The m odel is spelled out in Section 2. In Sectio n 3, we derive the equilibriu m market
structure and the equilibrium net profits fo r different entry modes. Section 4 , derives
implicatio ns for stock market va lues for the different e ntry modes. Section 5 concludes.
Finally, most proofs appear in the Appendix.
2. The M odel
Consider a country H, where the mark et has previously been served b y a single domestic
firm , denoted d, possessing on e unit of domestic assets, denoted k
0
. This mark et will
now be exposed to in ternation al competition . Th ere are several different reasons wh y the
mark et is now suddenly exposed to international competition. For instance, the country
might be investment liberalizing, the exp a nsion might be a natu ral step i n the life cycle

of a product or stem from increasing local d ema nd , or the admin istrative co sts of cross-
bo rder acq uisition s and gree nfield entry may have been reduced in the globalization
process.
We assume there to be M>1 symmetric MN E s in th e world market. A t the o utset,
the MNEs ha ve no assets in Co untry H, but might n ow invest. The interaction takes
place in three stages. In the first stage, t he MNEs migh t a cquire the domestic firm’s
assets. In the second stag e, M N E s have the option of investing green field in n ew a ssets
in country H. Finally, in the third stage, firm s compete in oligopoly fashion in country
H.
The next sections describe the product market intera ction, the greenfield investm ent
game, and the acquisition game.
2.1. Stage three: product mark et in teraction
Theproductmarketprofits in the industry will depend on the distribu tion of asset
o w n ership. An asset o w nership v ector k is d efined as k ≡ (k
d
,k
1
,k
2
, , k
M
),whereentry
one r efers to firm d’s asset holdings, entry two to M N E 1’s asset holdings, etc. In order to
simplify the presentation, we will distinguish between two types of o w nersh ip structures:
(i) the on e where the d om estic assets are so ld to o ne of the MNE s, denoted k
m
,and(ii)
5
the one where the domestic assets remain in the hands o f the dom estic owner, denoted
k

d
.
Vectors k
m
and k
d
are defined as follo w s:
k
m
≡ k
m
(α, N
m
,k
0
,k
G
) ≡ (0,αk
0
,k
G
,k
G,
, k
G,
| {z }
N
m
0, , 0
|

{z }
M−N
m
−1
), α>0 (2.1)
k
d
≡ k
d
(N
d
,k
0
,k
G
) ≡ (k
0
,k
G
,k
G,
, k
G,
| {z }
N
d
0, , 0
|
{z }
M−N

d
). (2.2)
The first en try in each v ector shows the asset o w nership of the domestic firm, the second
en try is the asset ow nership of the, poten tially, acqu iring MNE. The parameter α>0
captures that an MNE a nd the domestic firm may differ in h ow efficien tly they c an use
the a ssets k
0
. We s hall discuss parameter α in detail belo w . The r e m a ining entries
show the asset ow nership o f the no n-a cquiring MNEs, being either successful g reenfield
en trants (hav ing assets k
G
) or ”exporters”, i.e. M N E s which do not succeed in in vesting
greenfield (havin g assets k
E
≡ 0). U nder MN E ownership, there is one acquiring MNE
and N
m
non-acquiringMNEsthatinvestgreenfield, whereas M − N
m
− 1 MNEs do not
in vest. Under domestic ownership, there a r e N
d
MN E s su ccessfu lly investing gr eenfield
and M − N
d
MN E s th at do no t.
Under MNE own ersh ip of th e dom estic assets k
0
,weletπ
A

(k
m
) denote the reduced-
form product market p rofit for the acquiring M NE , π
G
(k
m
) the c orresponding pro fit
for a non-acquiring MN E as a greenfield entrant, and π
E
(k
m
) the corresponding profit
for a non-ac quirin g M NE as an export e r (i.e. a non-inves t ing MNE). Under domestic
o wnership of the assets k
0
, M N E s are either greenfield entrants or exporters, with the
product mark et profits π
G
(k
d
) and π
E
(k
d
), r espectively. The corresponding profits for
the d omestic firm u nder the respectiv e ow nersh ip structures are π
d
(k
l

),l= {d, m}.
We ma ke the f ollowing assumptions a bout profitsintheproductmarketassumma-
rized by t able 2.1:
Assum p tion A1 states th at the product mark et profit f or all types of firms decreases
in the n umber of s uccessful g r eenfield entrants, N
l
.
Assum p tion A 2 states that a firm’sproductmarketprofit increases in its own capital
stock in coun try H. This assum ptio n then forms the basic motive for FDI in terms
6
Table 2.1: Basic assumptions.
Assumption
A1: π
h
(k
l
(·,N
l
) >π
h
(k
l
(·,N
l
+1)
A2: π
h
(k
l
) >π

E
(k
l
) ≡ 0,h= {A, G}
π
d
(k
d
) >π
d
(k
m
) ≡ 0
A3:









∂π
A
(k
m
)
∂α
> 0,

∂π
G
(k
m
)
∂α
< 0,
∂π
h
¡
k
d
¢
∂α
≡ 0,h= {d , G }
of acquisition or greenfield en try, stem m ing from trade cost avoidance or lower factor
costs. To facilitate the readability, but with no loss of generalit y, we norm a lize such that
π
E
(k
l
) ≡ 0, i.e. the product mark et profitinexportingissettozero. Moreover,we
also assume that the domestic firmwillnotmakeanyproductmarketprofit w ithout its
assets, π
d
(k
m
)=0.
The loc al assets k
0

may be used differen tly und er dom estic and foreig n ownership. As-
sumption 3 then states that an increase in the strategic value, α, increases the acquirer’s
profit, whereas the mark et profit for a non-acquirer (i.e. greenfield in vestor) decreases.
Thesizeoftheseeffects depends on the strength of co m plem entarities betw een MN E s’
firm -specific assets and the domestic assets. For example, the com b ination of a n MNE ’s
strong brand name and the acquired firm’s knowledge of the market or strength in dis-
tribution, may pro vide the acquiring MNE with a strong market position. If the brand
nam e of the domestic assets is locally v ery strong, the strategic valu e of the assets will
also be high. Or, if the domestic assets are sold at an early stage, the a cqu irer ma y gain
astrongfirst-mover a d vanta ge, b u ilding up a dom inant position in the produ ct market.
7
2.2. Stage t wo: Greenfield inv estm ents
At this stage, MNE s that did not enter the m arket through the acquisition o f firm d,
can e nter by undertaking a greenfield investment at a fixed cost, G. To sim p lify the
analysis, w e assum e that in vestments in greenfield assets k
G
are ”lumpy”, i.e. they come
in discrete assets or plan ts and the domestic firm does not find it pro fitable to invest in
this stag e due to, fo r instance, financial o r manager ial restrictions.
Assum ption A2 states th at there are locational advantages for MNEs of producing
in coun try H . In the literature on M NE s, greenfield entry is considered risky due to
the lack of knowledge of th e specific c ha racteristics of the local mark et. For exam ple,
Caves(1996)arguesthatoneofthemainbenefits of acquiring a local competitor instead
of entering green field is the avoidance o f such r isks. Moreove r, greenfield in vestm ents
contain a large initial investm ent under uncertain ty, w h ich is likely to be sunk to a large
degree. This is motivated by the fa ct that these a ssets are likely to be designed to fit
the p roduction in a particular industry an d the cost of restructuring them in to suitable
assetsinotherindustriesisassumedtobehigh.
6
To model this, we assum e that eac h

potential greenfield entrant enters suc cessfully with pr obab ility p ∈ [0, 1] and will n ot
enter with p robab ility 1 − p. N
l
is th en simply the MNEs drawn as successfu l in the
greenfield stage.
Note that w e abstract from the possibilit y that the n umber o f greenfield entrants
under domestic and foreign own ership of t he domestic assets is affected b y the strategic
value, α (i.e. we assume p to be independent of α). We will discuss the modelling of the
greenfield un certainty in more d eta il o f the end of Section 4 .
2.3. Stage one: the acquisition game
To focus o n t he b id din g co m petition am on g MNEs as t he d ete rm in ant of the e q uilibrium
buyer, we assume that MNEs post bids for the domestic firm , which that firm m ay accept
or reject. More specifically, the acquisition process is depicted as an auction where M
6
To our knowledge, the only empirical paper studying sector-specific assets is Ramey and Shapiro
(2001), which finds capital to be very sector-specific.
8
MN Es simultaneously post bids and the do m estic firm then either accepts or rejects these
bids. Each MN E i announces a bid, b
i
, for the dom estic firm. b =(b
1
,b
2
, , b
M
) ∈ R
M
is the vector of these bids. Following the announcement of b, the domestic firm ma y be
sold to one of the MN E s at the bid price or remain in the ownership of firm d.

7
The
acquisition is solv ed for N ash equilibria in undominated pure strategies.
It is assume d th at firm d cannot m a ke a bid for th e M N E s. This assum p tion m ight
be motivated by the domestic owner being financially w ea ker or lac kin g the competence
to efficiently run the larger business. M or eov er, it is assum ed that M N E s canno t ma ke
bids on eac h other’s firms. This assumption migh t be suppor ted in two basic ways in
a full merger model. On e is to assume that the profit of a merged en tity is sufficiently
small to imply that no merger takes place between the M N E s,
8
the seco n d is to a ssu m e
that mergers between MNEs wo uld not be perm itted by the com petition authorities.
We now turn to the firm s’ valuations of the dom estic firm’s assets, k
0
.Notethat
when forming its valu ation in stage one, a firm does not know the outcome of the
greenfield gam e i n stage t wo. Hence, since the n u mber of su ccessful greenfield entrants,
N
l
, is stochastic, it fo llows that the asset ow nersh ip structure is also stochastic at the
acquisition stag e. To capture this, define the stochastic v ariable (or function) K
l
(·,N
l
)
with realizations in terms of asset o wn ersh ip structures, k
l
(·,N
l
). The expected produ ct

mark et profitforfirm h when a firm of type l acqu ires the d omestic assets k
0
, ¯π
h
(K
l
),is
then defined as:
¯π
h
(K
l
) ≡ E
£
π
h
(K
l
)
¤
≡
N
l
max
X
N
l
=0
ρ(N
l

)π
h
¡
k
l
(·,N
l
)
¢
(2.3)
where π
h
¡
k
l
(·,N
l
)
¢
is th e product m a rket profitforfirm h when N
l
firm s en ter greenfield
successfully, ρ(N
l
)=
¡
N
l
max
N

l
¢
p
N
l
(1 − p)
N
l
max
−N
l
denotes the joint p robabilit y of observ-
7
If more than one bid is accepted, the bidder with the highest bid obtains the domestic assets. If
there is more than o ne MNE with such a bid, each such MNE obtains the assets with e qual probability.
There is a smallest amount, ε, chosen such that all inequalities are preserved if ε is added or subtracted.
8
For instance, it has been shown by Kamien and Zang (1990) that the hold-up problem in merger
formation might lead to no merger taking place in equilibrium, if the initial number of firmsissuffi-
ciently large. Moreover, mergers might be non profitable since the costs associated with mergers can be
substantial, for example due to problems of melting together differen t company cultures.
9
ing N
l
greenfield entr ants, where N
m
max
= M − 1, N
d
max

= M and
¡
N
l
max
N
l
¢
denotes the
combinato rial function.
There are then three different valuations whic h need to be co nsid ered:
v
m
i
m
j
is the expected value for MNE i of obtainin g k
0
,whenMNEj would otherwise
obtain k
0
. U sin g sym m etry amo ng MNEs, w e will suppress the subindices and simply
write v
mm
.Thefirst term show s the expected product mark et profit when possessing k
0
.
The second term is the expected product market profitwhenarivalMNEobtainsk
0
, in

which c a se greenfield en try in stage 2 takes p la ce with pr obab ility p.
v
mm
=¯π
A
(K
m
) − p[¯π
G
(K
m
) − G] . (2.4)
v
md
is the expected value for MNE i of obtaining k
0
when the domestic firm would
otherwisekeepthem. TheexpectedproductmarketprofitforMNEi when not obtaining
assets k
0
is different in th is ca se, since t he k
0
assets are in the hands o f the domestic firm,
they might be used differen tly from when in the hands of an MNE. This implies that the
expected product mark et profits as a greenfield entrant will typically be different.
v
md
=¯π
A
(K

m
) − p
£
¯π
G
(K
d
) − G
¤
. (2.5)
v
d
istheexpectedvalueforthedomesticfirm of obtaining k
0
.Thisissimply:
v
d
=¯π
d
(K
d
). (2.6)
The firms’ bidding beha vior is dependent o n the relation bet ween their own valuation
of obtaining the assets k
0
and all other firms’ valua tions of o bta inin g these assets. Since
MN Es are symm etric, valuation s v
mm
,v
md

and v
d
can be ordered in six different wa ys,
as shown in table 3.1.
3. The equilibrium ownership structure
In this section, w e derive the equilibrium ownership structure (EO S) and the acquisition
price, A. We also determin e how the level of strategic va lue affects the ac quisition
pattern. M oreo ver, w e study the profitability of the d ifferent en try modes, both after
10
the acquisition and after greenfield in vestment has taken place. T hese findings will be
instrumental in deriving the stock market p red iction s of an a cq uisitio n in Section 4.
The equilibrium ownership structure (EOS) and the a cquisition price, A,arede-
scribed in table 3.1 where it is shown that when one of the inequalities I1, I2, I3,orI4
holds, k
0
is o b tained by one of the MNE s. Under I1, I2 or I3, th e acquiring MNE pays
the acquisition price A = v
mm
,andA = v
d
under I4.WhenI
5
or I
6
holds, the domestic
firm k eep s its assets.
9
Table 3.1: The e quilibr ium ow nership structure and ac quisitio n price.
Inequality: Definition: Ownership structure: Acquisition price
I1: v

mm
>v
md
>v
d
K
m
v
mm
I2: v
mm
>v
d
>v
md
K
m
or K
d
v
mm
(if K
m
)
I3: v
md
>v
mm
>v
d

K
m
v
mm
I4: v
md
>v
d
>v
mm
K
m
v
d
I5: v
d
>v
mm
>v
md
K
d
.
I6: v
d
>v
md
>v
mm
K

d
.
3.1. Post-Acquisition Expected Profits
In th is section, we examine h ow the M N E s’ expected n et profits are affected by th e
acquisition. T h ese profits are described in Figure 3.1. As a point of reference, we first
describe the expected n et profit in the ab sence of an acquisition,
¯
Π
h
(K
d
),whichwillbe
referred to as the Pre-Acquisitio n Expected (Ne t) Profit (depicted in Stage 0 in Figure
3.1), w here, in the following, we shall drop the net abbreviation . If no acq uisition
is expected to occur, the MNE s’ expected pro fits are
¯
Π
h
(K
d
)=p
£
¯π
G
(K
d
) − G
¤
for
h = {A, NA}, wh ereas the domestic firm’s expected profitissimply

¯
Π
d
(K
d
)=¯π
d
(K
d
).
9
When I2 holds, there exist multiple equilibria. In one equilibrium, firm d keeps the assets
andnoMNEpostsabidabovev
d
. There is also an equilibrium where one of the MNE´s obtains
the assets at a price v
mm
− ε and another MNE posts the second highest bid at v
mm
− 2ε.
11
Pre-Acquisition Expected
(Net) Profit:
Post-Acquisition Expected
(Net) Profit:
Post-Greenfield
(Net) Profit:
Acquirer (A):
Non-acquirer
(NA):

Domestic
firm (d):
A
0. Investment
Liberalization
1. Acquisition game 2. Greenfield game and
3. Oligopoly interaction
k
m
EOS:

A
k
m
  A

h
k
m

Note:
A 
v
mm
: I1,I2 or I3
v
d
: I4

G

k
m
  G : entry
0:no-entry (exporting)


h
K
d



h
K
m

K
d
K
m


d
K
d



A
K

m
  A
p


G
K
m
  G
p

G
K
d
  G
p

G
K
d
  G
Pre-Acquisition Expected
(Net) Profit:
Post-Acquisition Expected
(Net) Profit:
Post-Greenfield
(Net) Profit:
Pre-Acquisition Expected
(Net) Profit:
Post-Acquisition Expected

(Net) Profit:
Post-Greenfield
(Net) Profit:
Acquirer (A):
Non-acquirer
(NA):
Domestic
firm (d):
A
0. Investment
Liberalization
1. Acquisition game 2. Greenfield game and
3. Oligopoly interaction
k
m
EOS:

A
k
m
  A

h
k
m

Note:
A 
v
mm

: I1,I2 or I3
v
d
: I4

G
k
m
  G : entry
0:no-entry (exporting)


h
K
d



h
K
m

K
d
K
m


d
K

d



A
K
m
  A
p


G
K
m
  G
p

G
K
d
  G
p

G
K
d
  G
Figure 3.1: Defining profits in differen t stages of the game under I1, I2, I3 or I4.
We th en proceed to examine the expected profits w hen an acquisition take s place,
¯

Π
h
(K
m
). ThisisreferredtoasthePost-Acquisition Expecte d Pr ofit (dep icted in Stage
1 in Figure 3.1). TheacquiringMNE’sexpectedprofitisthen
¯
Π
A
(K
m
)=¯π
A
(K
m
) − A,
whereas a non-acquiring MNE’s expected pro fitis
¯
Π
NA
(K
m
)=p[¯π
G
(K
m
) − G].The
dom estic fir m collects the acquisition price, i.e.
¯
Π

d
(K
m
)=A. We can then derive the
following proposition:
Proposition 1. The Post-Acquisition Expected P rofit for an M NE will: (i) be equal for
theacquirerandthenon-acquirerunderI1,I2orI3,(ii)belowerfortheacquirerthan
the non-acquirer when I4 holds, (iii) fall as comp ar ed to the Pre-Acqu isition Expected
Profit when the acquisition tak es place under I1 and I2, and increase as compared to the
Pre-Acqu isition Expected Profit when t h e acquisition ta kes place und er I3 and I4.
Part (i) in the pro position sho w s that when sev e ral M N E s are potential buyers of
the domestic firm’s assets, the P ost-A cqu isition Expected Profitfortheacquirerand
the non-acquirer will be equal when I1-I3 holds, i.e.
¯
Π
A
(K
m
)=
¯
Π
NA
(K
m
).Thereason
is that the MNE’s willingnes s to pa y is then also driv en by the desire to prevent other
12
MN E s fr om obtaining the asse ts, as illustrated by v
mm
>v

d
being fulfilled. Con sequ ently,
the acquisition pr ice will be such that an MN E is ind ifferent betw een acquiring and not
acquiring.
Part (ii) sho w s that, under I4, howev er, the acquirer’s Po st-A cquisition Expected
Profit w ill be lower than that o f the n on-acqu irer, i.e.
¯
Π
A
(K
m
) <
¯
Π
NA
(K
m
),since
the acq uiring MNE in this case m ainly pay s for eliminating a rival (the domestic firm).
Howev er, non-acquirers will also benefit from this elimination, b ut do not pay the price
for it. T here is thus a free rider problem associated with eliminating the do m estic rival.
Part (iii) sho w s that the P ost-A c qu isition Expected ProfitforallMNEswillfallas
compared to the correspo nding Pre-Acquisition Expected Profit, when the acquisition
takes place under I1 and I2, that is,
¯
Π
A
(K
m
)=

¯
Π
NA
(K
m
) <
¯
Π
NA
(K
d
). This situation
is illustrated i n Stage 0 and 1 in F ig u r e 3.2. To explain this fa ll in e xpected profit, note
that the acquisition price is A = v
mm
>v
md
. This shows that the acquirer’s valuation
of the domestic assets when the domestic firm holds the assets k
0
, v
md
, is lower than the
price this firm pays in the acquisition under I1 or I2, v
mm
. Consequently, the acquirer ’s
expected profits will fall as compar ed to its corresponding pre-acquisitio n value. It also
follows th at a non-acquiring MNEs expected p ro fits must fall since this firm is indifferent
bet ween acquiring and not acquiring.
Under I3 or I4, h owever, the acquisition price is A = v

d
<v
md
.TheMNEs’ex-
pected profits then increase from an acquisition, since t he acquisition price is no w lower
than their willingness to pa y, if firm d w ould otherwise k eep its asset. Hence, we have
¯
Π
A
(K
m
)=
¯
Π
NA
(K
m
) >
¯
Π
NA
(K
d
) under I3 and
¯
Π
NA
(K
m
) >

¯
Π
A
(K
m
) >
¯
Π
NA
(K
d
) under
I4.
3.1.1. Ac quisitio n pattern , Post-Ac q uisitio n Expected Profits and strategic
value
Letusnowrelatetheaboveresultstothelevelofthestrategicvalue,α.First,we
examin e when a foreign acquisition will take place. We then h ave the f ollowing result:
Proposition 2. A foreig n acquisition will take place if a n d only if α>α
∗
.
13
0. Investment
Liberalization
1. Acquisition game 2. Greenfield game and
3. Oligopoly interaction
Profit
k
m
Announcement:
Investment

Liberalization
Announcement:
Acquisition of
domestic firm
Announcement:
Greenfield entry
(Pre-Acquisition
Expected Profit)
(Post-Acquisition
Expected Profit)
(Post-Greenfield Profit:
Greenfield entrant)
(Post-Greenfield Profit:
Aquirer)
(Post-Greenfield Profit:
Exporter)

G
k
m


G
k
m
  G


A
k

m


A
k
m
  v
mm


E
k
m

 0
0
K
d
K
m
 p
G
K
m
  G


h
K
d


 p
G
K
d
  G


A
K
m
 

NA
K
m

0. Investment
Liberalization
1. Acquisition game 2. Greenfield game and
3. Oligopoly interaction
Profit
k
m
Announcement:
Investment
Liberalization
Announcement:
Acquisition of
domestic firm

Announcement:
Greenfield entry
(Pre-Acquisition
Expected Profit)
(Post-Acquisition
Expected Profit)
(Post-Greenfield Profit:
Greenfield entrant)
(Post-Greenfield Profit:
Aquirer)
(Post-Greenfield Profit:
Exporter)

G
k
m


G
k
m
  G


G
k
m
  G



A
k
m


A
k
m
  v
mm


A
k
m
  v
mm


E
k
m

 0

E
k
m

 0

0
K
d
K
m
 p
G
K
m
  G


h
K
d

 p
G
K
d
  G


A
K
m
 

NA
K

m

Figure 3.2: Illustrating the evolution of MN E profits under I1 or I2.
Proof. See the Appendix.
Proposition 2 thus sh ows t hat a high strategic value of the domestic assets is conducive
to foreign acqu isitions. A high strategic value is, how ever, no t necessarily associated with
high expected profits. W hen there are severa l poten tial buy ers of the dom estic firm’s
assets, th e Po st-Acqu isition Expected pro fitoftheacquirer,Π
A
(K
m
), will de crea se in α
when I1, I2 or I3 holds. To see this, first note that the acquisition price is a n on -acq uir ing
MNE’s w illin gn ess to p ay, i.e. A = v
mm
. Then, using (2.4) and Assumption A3, w e hav e:
dA
dα
=
d¯π
A
(K
m
)
dα
− p
d¯π
G
(K
m

)
dα
>
d¯π
A
(K
m
)
dα
. (3.1)
Expression (3.1) shows that when the strategic value α increases, the acquisition price A
increases more than the a cquirer’s product market profit, ¯π
A
(K
m
). Thi s follows directly
from the fa ct that w hen the d omestic a ssets have a larger strategic value, an M NEs
va luatio n of the d o m estic assets k
0
, v
mm
, increases not only because its product m ark et
profitasanacquirerin creases (i.e.
d¯π
A
(K
m
)
dα
> 0), but also since its product market

14
profitasnon-acquirerdecreases (i.e.
d¯π
G
(K
m
)
dα
< 0). C o nsequently, since
¯
Π
A
(K
m
)=
¯π
A
(K
m
) − A,wehave:
d
¯
Π
A
(K
m
)
dα
=
d¯π

A
(K
m
)
dα
−
·
d¯π
A
(K
m
)
dα
− p
d¯π
G
(K
m
)
dα
¸
=p
d¯π
G
(K
m
)
dα
< 0 (3.2)
Moreover, the non-acquirer also faces an iden tical decline in expected profits, since:

d
¯
Π
NA
(K
m
)
dα
=p
d¯π
G
(K
m
)
dα
< 0 (3.3)
We can summarize:
Proposition 3. (i) Under I1, I2 or I3, the Post-Acqu isition E xpected P rofits of all types
of MN Es, including the acquirer, w ill decrease, the more strateg ically valuable the domes-
tic assets are. (ii ) A t a sufficiently high strategic va lue, α>α
∗∗
, th e Post-Acquisition
Expected Profits of the MNEs will fall below their co rrespondin g P re-a cq u isitio n E x -
pe cted Profits(i.e.I1orI2willhold).
Proof. See the Appendix.
Note th at when there a re no externalities exerted on rivals associated with an ac-
quisit io n, en try will become more profitable, the m o re valuable the domestic assets are.
Consequently, it is the fact that there are several potential acquirers present, and that
the acquirer will use the domestic assets competing against the other potential acquirers
in a n oligopoly-interact io n, that drives th e result in the proposition.

3.2. Post-Green field Profits
We here exam ine profits for the differen t entry modes when the greenfield uncertaint y
is reso lved, Π
h
(k
m
),wherewemaynotethatk
m
is a realization of K
m
.Thisisreferred
to as the Post-Gr eenfield P rofit (depicted as dotted lines in comb in ed Stages 2 and 3
in F igure 3.2). An acquiring M N E’s profitisthenΠ
A
(k
m
)=π
A
(k
m
) − A,whereasa
non-ac quirin g MNE’s p rofitisΠ
G
(k
m
)=π
G
(k
m
)− G, giv en th at it su cceeds in g reen field

en try an d Π
E
(k
m
)=0if not. We shall here comp ar e the Post-G r eenfield Profits of the
acquirer, Π
A
(k
m
) to those of a successful greenfield entrant, Π
G
(k
m
).UnderI1,I2orI3,
15
the a cqu isition price is A = v
mm
=¯π
A
(K
m
) −
¯
Π
NA
(K
m
) an d thus:
Π
A

(k
m
) − Π
G
(k
m
)=[π
A
(k
m
) − ¯π
A
(K
m
)] +
£
¯
Π
NA
(K
m
) − Π
G
(k
m
)
¤
(3.4)
In general, (3.4) is ha rd to sign. To see this, note that a non-acquirer faces an
individual risk of not being able to en ter g reen field. An a cq u is itio n avoid s t h is individual

risk, bu t the value of a voiding this risk is incorporated in the acquisition price. Therefore,
the Post-G r eenfield Pr ofit for the acquirer tends to be lower than the Post-Greenfield
Profit for a s uccessfu l green field entrant. However, both types of firms also face a market
risk, since the r ealized product m arket profit, π
h
(k
m
) is given from a particular realizat io n
of the number of successful greenfield en trants, N
m
, and ma y therefore differ from the
expected product market profit, ¯π
h
(K
m
).Howsensitiveafirm’sproductmarketprofitis
to cha ng es in the n umber of competitors will then depend on its position on the produ ct
market, which is directly related to the strategic valu e of the domestic assets, α.This
implies that the Po st-G ree nfield p rofit for the acquirer can be lower or higher than that
of a successful greenfield entrant, depending on their respective product m a rket positions
and t he outcome in the greenfield en try stage.
Makin g use of symmetry, we can deriv e some analytical results. To see this, note
that the mark et risk solely d eterm ines th e first com ponent of (3.4), whic h show s the
difference in realized and expected product mark et profits for the acquirer, whereas
the seco nd term , wh ich shows the difference in expected an d realized product mar ket
profits fo r a successful non-acquirer, is jointly determined from the market risk and the
individual risk faced in g reen field entry. Ho wever, note that if MNEs ha ve symmetric
market shares in the oligopoly interaction, i.e. if k
A
= αk

0
= k
G
,wehaveπ
A
(k
m
)=
π
G
(k
m
) and π
A
(K
m
)=π
G
(K
m
). Then, the market risk is the same for both types of
firms, and only the individual risk differs. N oting that
¯
Π
NA
(K
m
)=p[¯π
G
(K

m
) − G] and
Π
G
(k
m
)=π
G
(k
m
) − G, (3.4) simplifies to:
Π
A
(k
m
) − Π
G
(k
m
)=− (1 − p) [¯π
G
(K
m
) − G] ≤ 0 (3.5)
We th us hav e the following result:
16
Proposition 4. (i) If the mar ket share of a successful greenfield entrant in the oligo poly
in tera ction is the sam e as that of the MN E entering b y acquisition, then the P ost-
Greenfield Profit of an MN E entering greenfield is at least as high as the P ost-Greenfield
Profit of an MNE ente ring by acquisition.

4. Stoc k mark et value and e ntry m ode
In this section, we w ill use the results derived in sections 3.1 a n d 3.2 t o derive im plicatio ns
for s tock mark et values for the different en try modes. Mor e precisely, we will study how
the stockmar ket value is a ffected by the announcement of th e acquisition and by outcome
in the greenfield in vestment game. We assume that the financial markets tak e all relevant
information into accoun t in their valuations. H o wever, following the standard approac h
in the so-called event studies on M& A performance, w e assum e that the merger comes
as a surprise for the financial markets.
10
We will discuss this assumption at the end of
this sectio n.
4.1. Stock mark e t effects when the acquisition is announced
We here comp are the stock market value of the differen t t ypes of firmsatthetimeofthe
announcement to its p re acquisition announcemen t value. This corresponds to comparing
the Pre-Acquisition E x pected p rofits for a firm of type h,
¯
Π
h
(K
d
), to its P ost-A cqu isition
Expected valu e,
¯
Π
h
(K
m
).
It follo w s th at the d om estic seller’s stock m arket value increases when th e acqu isition
is announced, i.e. since

¯
Π
d
(K
m
)=A > ¯π
d
(K
d
)=
¯
Π
d
(K
d
). Moreover, in Section 3.1.1,
we showed in Proposition 3(i) t hat due to b idding competition among M N E s both over
the benefits as an acqu irer - as well as avoiding the negative extern alities a s a non-
acquirer - the acq u isition price, A, was increasing in the strateg ic value o f t he domestic
assets, α. Consequently, we h ave the follo w ing result:
10
See Scherer and Ross (1990).
17
Corollary 1. (i) The stock ma rket value of the selling dom estic firm increases when the
acquisition is announced and (ii) under I1, I2 or I3, th e stock market valu e of the selling
dom estic firm is increasing in the s trategic value o f the dom estic assets.
Let us now turn to the M N E s. To infer the effect on stock ma rket value when the
acquisition is announced, we once more compare the difference in the Pre- to Post-
Ac quisition E xpec t e d pro fit, i.e. we e x amine
¯

Π
h
(K
d
) −
¯
Π
h
(K
m
) for the acquirin g MNE
and a non-acquiring M NE, h = {A, NA}. As illustrated in Figure 3.2, Proposition 1(iii)
showed that under I1-I2, the Pre-Acquisition Expected profit for both types of MNEs
exceeds their corresponding Post-A cqu isition Expected profit, i.e.
¯
Π
h
(K
d
) >
¯
Π
h
(K
m
),
whereas t he opposite is true under I3 or I4. Mo reover, as showed in Proposition 3(ii), the
bidding competition among MNEs implied that a high strategic value of th e domestic
assets leads to an acquisition under inequa lities I1 or I2. We thus ha ve the following
result:

Corollary 2. ( i) T he stockmarket value for the acquiring and the non-acquiring MNE ,
respectively, falls when the acquisition is announced under I1 and I2, and increases under
I3 and I4. (ii) The stockma rket value for the acquiring and the non-acquiring MNE falls
when the acquisition is anno unced , if and only if the stra teg ic value of t he domestic
assetsissufficiently high.
Wh at can then be said w h en comparing different types of MNEs? It follows directly
fromProposition1(i)thatnorelativechangeinstockmarketvalueoccursbetweenthe
acquiring- an d non-acquiring M N E s under I1,I2 or I3, since th e bidding competition
among MNEs implies that firms are indifferent between taking on the role as acquirer
and non-acquirer, i.e.
¯
Π
A
(K
m
)=
¯
Π
NA
(K
m
). How ever, under I4, we showed that non-
acquiring MNEs cou ld free-ride, i.e.
¯
Π
A
(K
m
) <
¯

Π
NA
(K
m
). Hence, we ha ve th e follow ing
results:
Corollary 3. The stoc kma rket value for the acquiring MN E is unchanged at the time
of the announcement as compared to a non-acquiring MNE under I1, I2 and I3, but
decreases un der I4.
18
4.2. Stock mark e t effe cts in the long run
Let us finally examine the stock market value after greenfield en try uncertaint y has been
resolv ed for the acquirer and a successful greenfield entrant. This can be examined by
inv estiga ting the MNE s’ P ost-G r een field Profits, Π
h
(k
m
), as illustrated in Figure 3.2. In
Section 3.2, we show ed in Proposition 4 that if the acquirer and a successful greenfield
en trant are sufficien tly symmetric in the oligopoly int era ction, the profits in g reen field
entry alw ays exceed t hose of an acquisitio n, i.e. Π
G
(k
m
) > Π
A
(k
m
), the intu ition be-
ing that the acquisition price discounts that an ac q uisition avoids the individual risk

associated with green field entry. H ence, w e have the follo w in g proposition :
Corollary 4. The stock market value of an MNE successfully entering green field, w ould
in th e long run when the greenfield uncertainty h as been resolved, in crease relativ e to the
stock m a rket value o f th e acquirer, if firms are su fficiently symmetric in m a rket shares.
Howev er, we also sho wed tha t firms face a market risk as well since the realized prod-
uctmarketprofit, π
h
(k
m
) ma y differ from the expected product market profit, ¯π
h
(K
m
).
How the market r isk affects a firm’s stoc km ar ket value will then depend on its ex pected
market position in the oligopoly interaction . To also e xam ine the impact o f th e market
risk, we simu lated the m odel using Courno t competitio n with lin ear deman d.
11
In F igure
4.1, it is sho w n h ow the difference in Post-Greenfield profits fr om (3.4), Π
A
(k
m
)−Π
G
(k
m
),
indicating the relative change in stockm arket valu e betw een the acquirer and a success-
ful gr e en field entran t, depends on th e s trateg ic value of the d om e stic a ssets, α,andthe

outcome in t he greenfield gam e, N
m
. Figure4.1showsthatiffirms are sufficiently sym-
metric on the product mark et, the stock market v alue of the acquirer is reduced, relativ e
to that o f a successful greenfield entrant, i.e. Π
A
(k
m
) < Π
G
(k
m
). How ever, th e simu-
lation also illustrates that if competition becom es sufficien tly softer than expected (i.e.
N
m
<
¯
N
m
,where
¯
N
m
is the expected n umber of greenfield entrants) and the strategic
value is sufficiently high, the aquirer’s stockmarket value will increase relative to that
of a successfu l greenfield entrant, i.e. Π
A
(k
m

) > Π
G
(k
m
). Intuitively, in such a case,
11
See the Appendix.
19
4
1
2
3
5
6
7
8
9
10

A
k
m
 
G
k
m


A
k

m
 

G
k
m

Competition
more intense
than expected
Competition
less intense
than expected
2
0.4
EOS  k
d
k
G
 2
1.2 2.8
3.6
Number of greenfield
entrants, N
l
k
0
 1
Strategic value, i.e.
k

A
 k
0
N

m
4
1
2
3
5
6
7
8
9
10

A
k
m
 
G
k
m


A
k
m
 


G
k
m

Competition
more intense
than expected
Competition
less intense
than expected
2
0.4
EOS  k
d
k
G
 2k
G
 2
1.2 2.8
3.6
Number of greenfield
entrants, N
l
k
0
 1
Strategic value, i.e.
k

A
 k
0
N

m
Figure 4.1: The linear Cournot model.
the acquirer has a large expected market share, an d therefo re gains more from a price
increase associated with less than expected entry, than does a smaller greenfield entrant.
Let us end this section of stoc km arket effects with som e discussion of the ev ent
study app roa ch. In the ev ent studies it is assum ed th at th e acquisition c om es as a
surprise for the stockmar ket. Ho wever, Fridolfsson and Stennek (1999) arg ue that, if an
efficien t stoc kmark et an ticipate the acquisition, th e new information in the acquisition
announcement is which firms are insiders and whic h are outsiders.
12
Howev er, consider
a situation where the stoc km arket has difficulties in evaluating the strategic value α of
the dom estic assets for the MN E s. Here, the m erg er sh ould come as a p artial su rprise for
the market. The stockm arket effects of the a nno uncem ent will then be more involved,
since the financialmarketmustthenupdateitsbeliefsoftheentrygameaswellasthe
product m ark et. The main effects found shou ld, however, still be valid.
12
Under this assumption t hey show that preemptive m ergers could explain the empirical evidence that
mergers reduce profits and raise share prices. Fridolfsson and Stennek (2000), u sing the same approach,
show the limits of using effects on rivals’ share prices to determine the competitive effect of a merger.
20
If the stockmark et is instead assumed to be perfectly inform ed, the stoc kmark et
effects when the merger is announced will look different. Stock market values should
now c ha nge before the merger ann ouncem ent, for instance w hen the local mark et is
libera lized, since the a cquisition is a nticipated b y the financial m ark et. Consequently,

Corollary s 1 and 2 (under I1-I3) should then come when the liberalization and not when
the acquisition, is announced. Ho wever, there will still be an effect at the announcem ent
of the acquisition, namely w h ich firm is the acquirer and which a re non-acquirers. If
we are in I4, it then fo llo w s that the acquirer’s stockmarket will decrease and the non-
acquiring MNE’s stockmarket value w ill in crease.
5. Concluding discussion
In this p aper, it has been sh own that the biddin g c o m petition over the d o m estic target
firm implies t hat the d om estic firm will sell its assets at a higher, an d possibly substan-
tially higher, p rice than its reservation price. T he empirical implication is then that the
target firm’s sha rehold ers benefit from the acquisition. The predict io ns for the acquirer
are more in volved. H owev er, an interesting finding is that the share value of both the
acquirer and the non-acquirer will decrease when a n acquisition is announced, if t he
dom estic assets a r e su fficiently strategically importa nt. This is d u e to the fact that t h e
bidding competition is then so fierce that the firms involved would be better o ff not
starting a bidding w ar.
An indication that the mechanism s iden tified here are em pirica l relevant c an be found
in the emp irical event study literature on cross-border M&A performa nce. It has there
been sh own that there is a takeo ver prem ium in cross-border M& A s an d that the takeov er
premium differ between industries.
13
Moreov er, Cebeno yan et al (1992) documen ts a
signific ant po sitive link between the share of dollar volu m e of fo reign acquisitions in a
particular industry ov er the total value of foreign acquisitions and the target shareholder
wealth g ain in that in du stry. This finding seems consistent with the results in ou r model:
acquisitions takes place in industries where domestic assets are strategically important
13
See, for instance, Cebenoyan et al (1992), Dewenter (1995) and Harris and Ravenscraft (1991).
21
and in those industries targets benefits the m o st fr om the sev ere bidding com petition
between th e foreign entrants over these strategically important assets.

14
Moreover, the mo tive for pa y ing a high price for importa nt com plemen tarity assets
seems to have been important in several recent large acquisitions. One examp le is th e
bidding competition over Banco do Esta do de Sao Paulo (Banesp a), the seventh-largest
bank in Brazil. I n Novem ber 2000, Banco San tander Central Hispanio (BSCH) won
a controlling minor ity stake in Banespa, in competition with several other large ba nk s,
including its Span ish rival Ban co Bilbao Vizcay a Argentar ia (BBVA). According to B u si-
ness Week (April 23, 2001): “ It cost an astronomical $3.55 billion, but it put B SC H back
on top” (before BBVA - au thors´ comment). The assets of Banespa were considered
strategically valuable as ind icate d by the following qu ote ”An yone w ho can add Banespa
to their existing structu re w ill take a gigantic leap forward,” says Elio D ua rte, director
of institutional relations at the Brazilian subsidiary of Britain’s HSB C Holdings PLC ,
one of the nine banks qualified to take part in the a uctio n.” (Business Week, November
20, 2000). According to Business Week (No vember 20, 2000), this means that ” bidders
will pay a premium not just to get their hands on Banespa but also to stop rivals from
doing so.”
If there is risk associated with greenfie ld entry our empirical predic tion is that, in
the lo ng run when th e greenfield uncertaint y has been resolved, th e share value of a
successful greenfield en trant sh ou ld perform better than th e share value of the acquirer,
if the firms are sufficiently symm etric in the produ ct market. To test this hypoth esis, w e
need to be able to d istinguish between successful and unsuccessful non-acquirers. One
po ssibility would be to use data fro m markets opened u p by an in vestm ent liberalization.
It shou ld then be po ssible to identify the MNE s active in the industry: acquiring firms,
greenfield entran t s, exporters and firms not active in the mar ket.
One poten tial problem with testing the results from this model is tha t M N E s are
t ypical m ulti-product firm s only generating a small fraction of their revenues from the
14
That specificity of assets are important for the acquisition pattern are found in Blonigen (1997). He
finds support for the hypothesis that real dollar depreciations make Japanese acquisitions more likely
in U.S. industries, particular those which more likely have firm-specificassets.

22
market where the ac quisitio n tak e s place. Using profit flows in affiliates might be a way
of handling this problem.
The m odel u sed is obviously restrictiv e in sev eral respects. If MNEs are ex-ante asym-
metric, a cr oss-border acquisition should lead to higher profita bility and share prices for
the acquirer, com p ared to its rival firm s not involved in acquisitions, since the acquiring
firm w ill then pay a lower price than its valuation of th e firm,therebyleadingtoasurplus
for the acquir er .
Itcanalsobeshownthatthemainresultswouldholdalsoiftherewerefreeentryas
long entry does not take place in suc h a small scale that all rents are competed awa y.
15
Howev er, if free entry i m plies zero p ro fits in expectation for green field entran ts, the
profitability and stock market value of the acquirer and the non-acquirer will not be
affected w hen the acqu isition is announced.
The m a in r esults in th e pa per would hold if the ac quisitio n an d g reen field decisions
were assu m ed to take place simultaneo usly. To see th is, note that as long as the dom estic
assets a re scarce a nd their use by an M N E shifts profits from greenfield investors to th e
acquiring MNE , v
mm
mightbehigherthanv
md
and v
d
,and
dA
dα
> 0, and th us the results
in Section 3.1 holds. More over, the results in Section 3.2 would also hold since acquisition
en try is still certain a nd greenfield entry uncertain. H o wever, i f greenfield in vestm ent
were to take place first and acquisition after wards, the implication would be different.

First,theargumentthatanacquisitionispreferablesinceitallowsforanearlyentry
would naturally disappear. Moreover, if ”over in vestment” in greenfieldentrycouldbe
used as an entry deterrin g (p reda tory ) s trateg y, the dom estic assets might be worthless
in the acquisition game. How ever, if the dom estic assets are sufficiently unique and suc h
”o ver investments” are not profitab le, the dom estic assets might actually incr ease in
valu e. To see this, note that the unsuccessful greenfield entrants outsid e option is now
export profit, whic h amounts to less than the expected profit of a potentia l greenfield
en trant. Consequently, their willin gness to p ay may increase and th us, the acquisition
pric e mig ht be highe r in s uc h situa tions.
15
See Norbäck and P ersson (2002).
23