Asian Economic and Financial Review

Transcription

1 Aian Economic and Financial Review, 05, 5(): Aian Economic and Financial Review ISSN(e): -6737/ISSN(p): journal homepage: CROSS SHAREHOLDING AND INITIATIVE EFFECTS Yauhiro Arikawa --- Atuhi Kato Graduate School of Finance, Accounting and Law, Waeda Univerity; Nihonbahi -Chome Building, Nihonbahi, Chuoku, Tokyo, Japan School of Buine, Aoyama Gakuin Univerity, Shibuya, Shibuya-ku, Tokyo, Japan ABSTRACT Cro hareholding can make takeover more difficult but may be beneficial for hareholder if the manager private benefit align with hareholder benefit. Cro hareholding i more likely to take place a the congruence of interet between manager and hareholder increae, the manager private benefit become greater, the manager reervation utility i lower, and the hareholder hare of the takeover become maller. Due to the lack of monitoring, corporate value tend to be maller with cro hareholding. 05 AESS Publication. All Right Reerved. Keyword: Cro hareholding, Initiative effect, Takeover, Control right, Monitoring, Agency cot. JEL Claification: G3, G34. Contribution/ Originality Thi paper primary contribution i to clarify the condition under which cro hareholding i likely to occur and it i beneficial for hareholder.. INTRODUCTION It i known that cro hareholding i widely oberved outide the U.S. and the U.K., for example, in Germany and France (Shleifer and Vihny, 997; Adam, 999; Allen and Gale, 000), a well a in Eat Aian countrie (La Porta et al., 999; Claeen et al., 000). The cro hareholding tructure can make hotile takeover le likely. Thi inulate controlling manager with mall take from dicipline, and could generate large agency cot. What then explain the Clayton and Jorgenon (005). provide example of cro hareholding in the U.S. Correponding author DOI: /journal.aefr/05.5./ ISSN(e): -6737/ISSN(p): AESS Publication. All Right Reerved. 305

2 Aian Economic and Financial Review, 05, 5(): pervaivene of cro hareholding tructure given their large agency cot? In thi paper, we how the condition under which cro hareholding i likely to occur and why not only manager but alo other hareholder would approve of cro hareholding. Our model relie upon initiative effect, uch a thoe analyzed by Burkart et al. (997). Initiative effect imply that if there were le intervention by other hareholder, controlling manager with very mall take would work harder becaue they would be able to purue private benefit. In thi paper, we aume that manager obtain non-verifiable private benefit from project, while hareholder (including raider) receive monetary return. We further aume their interet diverge, but not completely. If a raider take over a firm, they will monitor a (new) manager and, with ome poitive probability, will force the manager to chooe a project beneficial for hareholder but le o for manager. Thu, the manager incentive to work hard i lower in a takeover. In the cae of cro hareholding, hareholder of firm do not intervene. Thi can lead to controlling manager working harder to eek private benefit. When the benefit of manager and hareholder are to ome extent accordant, cro hareholding can alo be beneficial to hareholder. In our model, cro hareholding i a kind of commitment by hareholder not to intervene or, more preciely, not to accept tender offer. Through thi commitment, manager can eek private benefit, which may be good for hareholder. We how that cro hareholding i more likely to occur a the congruence of interet between controlling manager and hareholder rie, the manager private benefit become larger, the manager reervation utility get lower, and the hareholder pie in the cae of takeover become maller. Due to lack of monitoring, the hareholder value of a firm get maller in cro hareholding, while ocial welfare could be higher or lower depending on ome condition. The main argument of our model i that any ownerhip tructure that effectively deter hotile takeover may induce initiative effect that may increae hareholder value. Claeen et al. (000) how that in nine Eat Aian countrie, a mall number of hareholder, indeed often a ingle hareholder, enhance their control through cro hareholding and pyramid tructure in a large number of companie, and in thee cae voting right exceed formal cah flow right. We uually conider thee ownerhip tructure a undeirable. However, our model indicate that if the benefit of minority hareholder are highly correlated with thoe of manager, minority hareholder might alo benefit. For intance, when the economy i growing rapidly, a in emerging countrie, minority hareholder could alo obtain large capital gain if controlling manager with mall take purue private benefit through expanding firm ize. Thi may explain why, depite apparently exploitative ownerhip tructure in Eat Aian countrie, foreign minority invetor are till willing to purchae the tock of local companie. A ubtantial literature ha analyzed cro hareholding and, a a cloely related concept, partial hareholding. Some reearcher examine partial and cro hareholding a a mean of reolving the hold-up problem (Riorden, 990;99; Dagupta and Tao, 000). Other reearcher In the financial literature, it i conidered that manager have the tendency to expand firm ize regardle of hareholder value. 05 AESS Publication. All Right Reerved. 306

3 Aian Economic and Financial Review, 05, 5(): have invetigated the relationhip between cro hareholding and product market competition (Reynold and Snapp, 986; Farrel and Shapiro, 990; Flath, 99); (Malueg, 99; Hanen and Lott, 995; Gilo et al., 006; Mathew, 006); (Clayton and Jorgenon, 005). All of thee tudie focu on cah flow right, while another line of reearch pay more attention to control right (Berglof and Perotti, 994; Oano, 996). However, thee previou tudie do not fully explain why hareholder agree to cro hareholding at the beginning. Our model hed light on the channel through which cro hareholding could provide higher return to hareholder by inducing greater effort from manager eeking private benefit. The initiative effect, a analyzed by Burkart et al. (997), i applied to the analyi of cro hareholding for the firt time, and thi viewpoint on cro hareholding differ from all previou work in thi area. The remainder of the paper i organized a follow. In Section, we explain the model and olve the equilibria. In Section 3, we how that cro hareholding reduce corporate value. Section 4 conclude the paper.. MONITORING VS INITIATIVE EFFECTS.. The Model We extend Burkart et al. (997) model a follow. There are two firm, firm and firm. The manager of firm i i called manager i, where i i or. Firm i face N i + potential project. Project j of firm i, where j {0,,,,N i }, may yield non-verifiable private benefit b j to manager i and verifiable pecuniary benefit j, which are eventually attributed to hareholder. Project 0 i ditinguihable, and it i known that 0 = b 0 = 0. The other N i project are not differentiable unle further invetigation i undertaken. For N i project, j < 0 and b j < 0, and at leat one yield j = b j =. The remaining two project, indexed N i and N i, yield tochatic payoff a per the following table. Probability Project N i Project N i i { i, b i } { i, b i } i { i, 0} {0, b i } If project N i i undertaken, pecuniary benefit i accrue to firm i and private benefit b i to manager i with probability i, and no benefit to firm i and private benefit b i to manager i with probability i. From the table, it i obviou that hareholder prefer project N i but manager i prefer to chooe project N i. The parameter i [0,] repreent the congruence of interet between the hareholder and manager of firm i. A higher i implie that the benefit of manager i are more cloely aligned with thoe of the hareholder. In line with Burkart et al. (997), we aume that if the manager make effort e, they will obtain payoff from a project. In addition to the manager and hareholder of the firm, there are many identical potential raider, who attempt to take over firm or firm. Unlike Burkart et al. (997), we aume that the raider make a dicrete deciion between monitoring and not 05 AESS Publication. All Right Reerved. 307

4 Aian Economic and Financial Review, 05, 5(): monitoring once the raider take over a firm. If the raider decide to monitor, the project payoff will be found with probability E, where E i a fixed number between 0 and. Thi monitoring cot the raider c > 0. There i no time dicount. We begin with the ituation where dipered hareholder currently own the hare of the two firm. The equence of event i a follow. In period, manager make a propoal to iue new hare and exchange them between the two firm. At the ame time, manager i propoe to make a monetary tranfer m i to hareholder of firm i. In period, hareholder decide whether to accept the propoal of cro hareholding cum monetary tranfer. In period 3, and baed on the outcome in period, the raider decide whether to make a tender offer to hareholder. We aume numerou identical potential raider who attempt to take over each of the two firm. In period 4, and if tender offer are made in period 3, hareholder decide whether to accept. In period 5, each manager, and the raider in cae of takeover, make effort and realize the profit and private benefit. We aume that if a takeover take place, a new manager replace an old manager. We aume the abilitie of the new manager are identical to the old manager, though the aumption of a more competent manager can be eaily included. Figure depict the time tructure of the game. The payoff to each player are explained when we olve for the ubgame perfect equilibria. Fig-. The time equence of the game Several note on the tructure of the game are in order. Firt, the time equence indicate that the manager propoe cro hareholding in order to avoid a takeover. Thi i becaue raider may attempt takeover later if there i no cro hareholding. If portion of hare i neceary to dimi exiting management, then at leat portion mut be cro hared among the firm to protect themelve againt takeover. Second, the aumption that manager make a monetary tranfer to hareholder may ound awkward. However, thi etup i not crucial. A become clear, the condition for cro hareholding to occur amount to the condition that the total value accruing to both hareholder and manager i higher with cro hareholding than in takeover. In uch a ituation, there hould arie ome method of diviion that make both better off. Note alo that manager are willing to make a monetary tranfer to hareholder becaue cro hareholding can effectively block takeover, and there are no concern about intervention in management under cro hareholding due to dipered hareholding. Otherwie, hareholder may act in an opportunitic manner, i.e., they both receive the monetary tranfer and accept a tender offer. Shareholder approval of cro hareholding i a kind of commitment not to ell hare. 05 AESS Publication. All Right Reerved. 308

5 Aian Economic and Financial Review, 05, 5(): The incentive of hareholder to repond to tender offer i ubtle. If a takeover take place, the raider will buy a certain portion of hare. A dicued in the literature, unle the hareholder who accept tender offer obtain the ame gain a hareholder who do not, none would accept the offer and takeover would not ucceed (Groman and Hart, 980). Therefore, the raider mut offer the ame premium to all hareholder. It then turn out that in term of the payoff that each player obtain, it doe not matter how many hare the raider would capture. Rather, what matter i how much the raider and the hareholder obtain. In thi paper, thi ratio i aumed to be exogenouly determined before the game begin. Shareholder will get [0,] out of the realized profit, and the raider will obtain. In thi tudy, we refer to a the dilution rate. We alo aume that the raider ha many other potential takeover target, and give up a takeover attempt if thi dilution rate i not acceptable. Finally, in thi article we aume ymmetry, and will omit the index of a firm hereafter... Equilibrium Next, we olve for the ubgame perfect equilibria through backward induction. Firt, we olve for the equilibrium in each ubgame in period 5 in the takeover cae and in the cro hareholding cae.... Takeover Cae We aume that if the takeover take place in period 4, the raider will dimi the current manager, and the outed manager would obtain reervation utility of r. Denoting the payoff that the dimied manager would get in the cae of takeover a V mt, then: V mt r The raider will hire a new manager. The new manager make an effort e at a cot e /. Thi effort reveal the payoff of all project to the manager with probability e, where 0 e. A poitive parameter capture the difficulty of raiing the probability of uccefully finding the payoff of the project. A we aume that the new manager ha the ame competence a the old manager, take the ame value for both the new manager and the dimied manager. The raider then decide whether they will monitor the manager or leave free. If the raider chooe to monitor, they exert effort level E at cot c, which enable the raider to ee the payoff of all project with probability E, but only when the manager find the payoff of the project. If the raider decide to monitor, the new manager (hereafter manager) chooe e to maximize their payoff, denoted V nmt : V nmt e[ E b ( E) b] e Project N i elected if the raider uccefully monitor. Then, with probability, private benefit b accrue to the manager. If the raider fail to monitor, project N i choen o that private 05 AESS Publication. All Right Reerved. 309

6 Aian Economic and Financial Review, 05, 5(): benefit b accrue to the manager with a probability of one. The firt-order condition yield the manager optimal effort level: e b [ ( ) E] () Lemma With a takeover, the new manager effort level increae a b and increae and a and E decreae. Proof. Omitted A private benefit b increae, the manager effort naturally rie. A increae, even if the raider uccefully monitored, the manager would till obtain a higher benefit, and thu make a higher effort. Converely, a rie, the manager effort i more cotly, and le effort i expended. Finally, a E increae, the new manager will be monitored with higher probability, and thu exert le effort. Therefore, in a takeover the poibility of monitoring reduce the manager effort. Thu, a takeover adverely affect initiative effect. Expecting a new manager to exert thi effort level, the raider will chooe whether to monitor. V rt denote the raider payoff in a takeover. If the raider doe not monitor, E = 0 in (), o the manager effort will be e = b/. The payoff will then be: V rt ( ) b ( ) e () On the other hand, if the raider chooe to monitor, the payoff will be: V rt ( ) e[ E ( E) ] c The term in quare bracket i the expected return when the manager find the payoff of project. In thi cae, if the raider uccefully monitor, project N i choen, and i realized with a probability of one. If the raider fail to monitor, i realized with probability. A the raider mut concede portion of return to hareholder, the whole term i multiplied by ( ). Finally, it cot c for the raider to monitor. Given the manager optimal effort level, the raider payoff i: V rt ( ) b [ E ( E)][ E ( E) ] c From () and (3) the raider will monitor the manager if the following condition hold: ( ) b Condition E( E)( ) c 0 Note that the raider make no monitoring effort with violation of thi condition. Thi i becaue, under the aumption of the identical abilitie of manager, there i no incentive for hareholder to accept tender offer. A a reult, the new manager effort level and the realized corporate value are the ame a for the cro hareholding cae. The raider will imply take away ome portion of the realized corporate value without any monitoring. If o, hareholder will not (3) 05 AESS Publication. All Right Reerved. 30

7 Aian Economic and Financial Review, 05, 5(): accept tender offer, and by expecting thi manager do not propoe cro hareholding. Thu, the tatu quo continue. From (3) we can ee that V rt increae a b and increae, while it decreae a, and c increae. Incidentally, the new manager payoff V nmt i: e b [ E ( E)] V nmt e[ E b ( E) b] The payoff of the hareholder V t i expreed a follow: Plugging e into V t, we obtain the following: V t e ( ) E b{ E ( E)}{ E ( E) } V t (4) Lemma V t increae a, b and increae or a decreae. With repect to E, V t increae up to ½ and then decreae. Proof. Omitted The intuition i a follow. The change in doe not affect manager effort, only the portion of corporate value given to hareholder. Conequently, the increae in raie V t. Further, the increae in b enhance the effort level of the manager, the increae in raie expected corporate value, and both make V t larger. A effort become le cotly (i.e., the decreae in ), the manager exert more effort and increae V t.... Cro Shareholding Cae Auming ymmetry between firm and, they iue the ame amount of hare and imply exchange thee without payment. We aume that each manager doe not monitor the other under cro hareholding. Furthermore, becaue of the diperion of the ownerhip tructure, hareholder would not monitor the manager. The manager payoff V m i: V m e eb (5) If the manager uccefully find the payoff of project, they will chooe project N. Then, b i realized with a probability of one. The manager mut pay e / to expend effort level e. From the firt-order condition for maximization with repect to e, the manager effort i: Lemma 3 b e (6) The manager effort level increae with b or a decreae. 05 AESS Publication. All Right Reerved. 3

8 Aian Economic and Financial Review, 05, 5(): Proof. Omitted Intuitively, a the private benefit get larger or effort get le cotly, the manager make more effort. Propoition The manager exert a higher level of effort with cro hareholding than in a takeover. Proof. In the takeover cae, the manager effort i b { E ( E)} / while in the cro hareholding cae it i b /. It i eay to ee: b{ E ( E)} b Q.E.D. Thi i the ource of the initiative effect: when manager are unmonitored, they are willing to work harder. Subtituting e in (6) back into V m yield: From thi, we can ee that V m increae a b increae, or a decreae. If portion of the hare of firm i owned by firm, the return to hareholder i only ( ) of the total hare price of firm. Suppoe firm hold portion of the hare of firm and firm hold portion of the hare of firm. If firm realize profit and firm realize, the hare price of firm under cro hareholding,, i: For thi derivation ee Flath (996). Under cro hareholding, e and.let V be the payoff to hareholder of firm i with cro hareholding. Then: e i e b V m eb V e e ( ) Subtituting e = b / and e = b / yield: Under the ymmetry aumption, = =, = =, = =, b = b = b, = = and V V V, we have: V b ( ) b 05 AESS Publication. All Right Reerved. 3

9 Aian Economic and Financial Review, 05, 5(): Note that V doe not depend on in the ymmetric cae. It i then jut the expected value of pecuniary benefit accruing to firm i. Lemma 4 V increae a,, and b increae, or a decreae. Proof. Omitted. b V..3. Statu Quo If neither cro hareholding nor takeover occur, hareholder payoff equal V, the manager payoff equal V m, but no monetary tranfer take place. Thu, manager prefer the tatu quo..3. Takeover v Cro Shareholding To ee whether hareholder accept tender offer in period 4, we compare V and V t. If = 0, V > 0 = V t. On the other hand, if = : V V t b E( E)( ) Thi mean V i maller than V t. Becaue V t monotonically increae in, a * exit uch that if < *, V > V t, while if > *, V < V t. We thu get, from V = V t : * { E ( E)}{ E ( E) } In period 4, hareholder would accept tender offer if >* and not otherwie. We refer to thi a condition : 0 Condition In period 3, the raider decide whether to make a tender offer. Knowing that hareholder will accept tender offer in period 4 if >*, the raider will compare the payoff in each cae. If the raider make a tender offer, it will get V rt. If not, the raider will get an outide payoff. Let thi outide payoff be V o. Then, the raider make a tender offer to firm or firm if V rt > V o. We aume that thi inequality hold. In period, hareholder are willing to accept cro hareholding if the following condition i met: V t m V (7) The left-hand ide of the inequality i the payoff to hareholder if they accept the propoal of cro hareholding with monetary tranfer, while the right-hand ide i the payoff when they reject thi but accept tender offer in period AESS Publication. All Right Reerved. 33

10 . 3 (B) From condition and, the event from period 3 are the ame a in cae (A). However, Aian Economic and Financial Review, 05, 5(): Further back in period, the manager decide to propoe cro hareholding with monetary tranfer m. If the following inequality hold, the manager i willing to make thi propoal: V m m r (8) Thi inequality ay that even if the manager made monetary tranfer m to hareholder, the utility would till be higher than the reervation utility r. Combining (7) and (8), cro hareholding i propoed in equilibrium if condition 3 i atified: Condition 3 V m V V t r Condition 3 implie that the total value accruing to manager and hareholder in cro hareholding are higher than in a takeover. We uppoe that a long a the total value to hareholder and manager are higher in cro hareholding, they will manage to figure out a way to hare them o that both are better off. For example, on average, manager of Japanee companie receive low alarie relative to their U.S. counterpart. We can conider thi a a kind of tranfer mechanim from manager to hareholder. A mentioned earlier, the following analyi would not involve the tranfer of m. Now, we fully decribe a ubgame perfect equilibrium. Propoition The following contitute a ubgame perfect equilibrium in thi model. (A) If condition, and 3 are atified, then cro hareholding i propoed in period. Shareholder then accept it in period and the raider doe not make a tender offer in period 3. (B) If condition and are atified but condition 3 i violated, then the manager doe not make a cro hareholding propoal in period. The raider make a tender offer in period 3 and hareholder accept it in period 4. The raider monitor a new manager in period 5. (C) If either condition or i violated, then the manager doe not make a cro hareholding propoal in period, the raider doe not make a tender offer in period 3, and the manager make effort in period 5 without any intervention. Proof (A) From condition, if the raider ucceeded in a takeover, it would monitor the manager in period 5. Expecting thi, and from condition, hareholder would accept a tender offer in period 4 if offered. By aumption, the raider would make a tender offer in period 3 if there wa no cro hareholding. A the threat of a takeover i real from condition 3, the manager find it more deirable to make a cro hareholding in period, and hareholder will accept it in period becaue condition 3 i violated, the manager doe not make a cro hareholding propoal in period. The takeover will then ucceed. 3 The diagram of game tree correponding to each cae are available from author upon requet. 05 AESS Publication. All Right Reerved. 34

11 Aian Economic and Financial Review, 05, 5(): (C) If condition i not atified, it i never beneficial for hareholder to accept tender offer, and they would not do o in period 4. Expecting thi, the manager doe not make a cro hareholding propoal in period. If condition doe not hold, hareholder would not accept tender offer in period 4. Expecting thi, the manager doe not make a cro hareholding propoal in period. In both cae the tatu quo remain. Q.E.D. We can eaily how that condition and can hold imultaneouly. Temporarily, we focu on the et of parameter value that atify condition and. Thi implie that either poibilitie (A) or (B) would occur. Then, we inquire into the condition that make cro hareholding more likely relative to takeover, i.e., thoe that make condition 3 more likely to be atified. From condition 3, cro hareholding occur if: Define a new function: b b b{ E ( E)}{ E ( E) } r b b b{ E ( E)}{ E ( E) } F( b,,, E,,, r) r The increae in F looely indicate the higher poibility of cro hareholding. Propoition 3 F(b,,, E,,, r) increae a r and decreae, and a and b increae. Proof. It i obviou that F increae a r and decreae and a b increae. Rewriting F yield: b F( ) b b { E ( E)}{ E ( E) } r Denote the numerator of the firt term on the right-hand ide a G( ). Then, differentiate G with repect to : G 4b E( E) b b { E So G i concave in and greatet at: ( E) } G, 4bE ( E) 0 { E ( E) E ( E) } E ( E) Thu, for [0,] F increae a increae. Q.E.D. Note that ince condition 3 include r, which i independent of condition and, there alway exit a low that atifie condition 3. Theoretically, there i no retriction on r. Thu, if the exiting manager fall into a terribly mierable ituation once dimied (i.e., a very mall r), then they are willing to make large monetary tranfer to hareholder. Hence cro hareholding would occur. 05 AESS Publication. All Right Reerved. 35

12 Aian Economic and Financial Review, 05, 5(): Thu the quetion i not whether condition 3 can occur, but what parameter value make condition 3 more likely to be met. Given any value of, and b, there i a value of r that make pecific value of, and b threhold value for condition 3. Then thoe change decribed in propoition 3 may become critical. Intuitively, other thing being equal, the decreae in make takeover le attractive for hareholder. On the other hand, a private benefit b rie, the manager i likely to make a greater effort. Then, the hareholder benefit (which are correlated with the manager benefit) become greater, meaning that hareholder will not have to depend on the raider intervention. In addition, lemma implie that up to ½, the increae in E make cro hareholding le likely; beyond ½, it i more likely. Thi can be een by rearranging F: F b ( ) ( E ) Therefore, F reache it minimum at E = ½. [ b b ( ) b ( ) Now, we pay pecial attention to the effect of the change in. Propoition 3 implie that a increae, cro hareholding i more likely to occur. Intuitively, a increae, the congruence of interet between manager and hareholder rie. In thi cae, the benefit from relying on the raider monitoring i mall for hareholder. Converely, if i low, hareholder find cro hareholding le attractive than a takeover. Although we do not make a rigorou analyi of how hareholder would behave when become low and cro hareholding i already in place, we could looely tate that for a mall hareholder have le reaon to agree to cro hareholding. ] Fig-. Player action in repone to the value of and c Next, we focu on the relationhip between the parameter and c. A higher c mean that the raider face greater difficulty in monitoring the manager. In Figure, i meaured on the vertical axi and c on the horizontal axi. Condition include both c and. A c increae, mut decreae in [0,] to make the following maller. ( ) b E( E)( ) Therefore, the relationhip between and c i expreed by a downward-loping curve in Figure, which we call curve. If monitoring cot c i o high that it i in the top right-hand area 05 AESS Publication. All Right Reerved. 36

13 Aian Economic and Financial Review, 05, 5(): of curve, the raider will not monitor. Unle the raider monitor the manager, it exitence imply mean the robbery of rent from hareholder. Thu, hareholder would not accept tender offer. Expecting thi, the manager doe not make a cro hareholding propoal. Then, regardle of the value of, neither a takeover nor cro hareholding would take place. Hence, we argue that for thoe firm where a raider would have great difficulty in monitoring the manager, cro hareholding and takeover would not take place. 4 A condition and 3 do not include c, the curve derived from condition (V = V t ) and condition 3 (V + V m = V t + r) are two horizontal line. We refer to the former a line and the latter a line 3. With regard to the relative location of the two line, from V + V m = V t + r, we get: b b b{ E ( E)}{ E ( E) } r 0 The left-hand ide i F in propoition 3. A we have een, F increae a increae in the domain [0,] and (b / r) > 0. Therefore, compared with that atifie condition, that atifie condition 3 mut be maller. In the area above line, hareholder would not accept tender offer in period 4. Therefore, a imilar argument to the above implie that the tatu quo continue. In the area below line, hareholder would accept tender offer in period 4. Thu, the threat of a takeover i real to manager. In the area below line and to the bottom left-hand area of curve, two cae can occur. If it i above line 3, cro hareholding occur, but if it were below line 3, then a takeover would occur. From Figure, we ee that takeover would occur if goe down from the cro hareholding area to the takeover area. Intuitively, a the correlation between the manager benefit and the hareholder benefit decreae, hareholder would rather rely on the raider monitoring than the manager initiative effect. In contrat, if i very high, hareholder would not accept tender offer becaue higher effort by manager eeking private benefit would ufficiently increae hareholder benefit. Neverthele, by expecting thi, the manager doe not propoe cro hareholding. From thi argument, we would claim that the manager of a good performing firm doe not require cro hareholding a a defenive device.. 3. A COMPARISON OF PERFORMANCE In thi ection, we compare performance in the cro hareholding and takeover cae. Note that the tatu quo i equivalent to cro hareholding in term of corporate value and ocial welfare. Here, we conider the monetary value each firm realize a our criteria. In a takeover cae, the value of a firm, CV t, can be hown to be: CV t { Eb ( E) b}{ E ( E) } When the raider ucceed in monitoring, the monetary value i realized with a probability of 4 If a potential raider conit only of foreign capital, thee firm could be thoe that only manufacture very traditional good, uch a kimono or hanafuda (a Japanee card game) becaue foreign raider find it difficult to locate an appropriate trategy in thee indutrie. 05 AESS Publication. All Right Reerved. 37

14 Aian Economic and Financial Review, 05, 5(): one. When the raider fail, thi only happen with a probability of. Similarly, in the cro hareholding cae: CV c b ( ) e e Propoition 4 Cro hareholding alway yield a lower monetary value than a takeover. Proof. CV t b b{ E ( E)}{ E ( E) CV b E( E)( ) 0. Q.E.D. c } Intuitively, when there i no monitoring, the manager would chooe project N, which lead to a lower expected value of monetary payoff, even if the effort level are the ame. Thi effect dominate, even though the manager exert higher effort under cro hareholding. 5 A we have een, cro hareholding may occur under certain condition. Hence, even if the corporate value of a firm i lower in cro hareholding cae, cro hareholding till occur. 4. CONCLUSION Thi paper how that in contrat with the traditional view, cro hareholding may be beneficial for hareholder. Thi reult depend on initiative effect that imply that the manager make greater effort if they face no intervention by hareholder in eeking private benefit. A long a hareholder benefit correlate with the manager private benefit, hareholder may alo benefit from no intervention due to cro hareholding. We how that the corporate value of a firm tend to be maller in cro hareholding due to the lack of monitoring. However, if we include manager private benefit in the ocial welfare function, it i poible that ocial welfare i higher with cro hareholding. REFERENCES Adam, M., 999. Cro holding in Germany. Journal of Intitutional and Theoretical Economie, 55(): Allen, F. and D. Gale, 000. Corporate governance and competition, in X., Vive, (Ed).: Corporate governance. Cambridge, U.K.: Cambridge Univerity Pre. Arikawa, Y. and A. Kato, 004. Cro hareholding and initiative effect. RIETI Dicuion Paper Serie, No. 04-E-07. Berglof, E. and E. Perotti, 994. The governance tructure of the Japanee financial keiretu. Journal of Financial Economic, 36(): If we take ocial welfare a our criterion, which i the corporate value of a firm plu private benefit to manager le the cot of effort, we can how that the ocial welfare in cro hareholding cae i more likely to be larger than that in takeover cae, a decreae, c increae and b/ increae. Pleae refer to the earlier verion of thi paper Arikawa and Kato (004). 05 AESS Publication. All Right Reerved. 38

15 Aian Economic and Financial Review, 05, 5(): Burkart, M., D. Gromb and F. Panunzi, 997. Large hareholder, monitoring, and the value of the firm. Quarterly Journal of Economic, (3): Claeen, S., S. Djankov and L. Lang, 000. The eparation of ownerhip and control in Eat Aian corporation. Journal of Financial Economic, 58(-): 8. Clayton, J.C. and B.N. Jorgenon, 005. Optimal cro holding with externalitie and trategic interaction. Journal of Buine, 78(4): Dagupta, S. and Z. Tao, 000. Bargaining, bonding, and partial ownerhip. International Economic Review, 4(3): Farrel, J. and C. Shapiro, 990. Aet ownerhip and market tructure in oligopoly. Rand Journal of Economic, (): Flath, D., 99. When i it rational for firm to acquire ilent interet in rival. International Journal of Indutrial Organization, 9(4): Flath, D., 996. The keiretu puzzle. Journal of the Japanee and International Economic, 0(): 0. Gilo, D., Y. Mohe and Y. Spiegel, 006. Partial cro ownerhip and tacit colluion. Rand Journal of Economic, 37(): Groman, S. and O. Hart, 980. Takeover bid, the free-rider problem, and the theory of the corporation. Bell Journal of Economic, (): Hanen, R.G. and J.J.R. Lott, 995. Profiting from induced change in competitor market value: The cae of entry and entry deterrence. Journal of Indutrial Economic, 43(3): La Porta, R., F. Lopez-de Silane and A. Shleifer, 999. Corporate ownerhip around the world. Journal of Finance, 54(): Malueg, D.A., 99. Colluive behavior and partial ownerhip of rival. International Journal of Indutrial Organization, 0(): Mathew, R., 006. Strategic alliance, equity take, and entry deterrence. Journal of Financial Economic, 80(): Oano, H., 996. Intercorporate hareholding and corporate control in the Japanee firm. Journal of Banking & Finance, 0(6): Reynold, R.J. and B.R. Snapp, 986. The competitive effect of partial equity interet and joint venture. International Journal of Indutrial Organization, 4(): Riorden, H.M., 990. Aet pecificity and backward integration. Journal of Intitutional and Theoretical Economic, 46(): Riorden, H.M., 99. Ownerhip without control: Toward a theory of backward integration. Journal of the Japanee and International Economic, 5(): 0 9. Shleifer, A. and R. Vihny, 997. A urvey of corporate governance. Journal of Finance, 5(): View and opinion expreed in thi article are the view and opinion of the author, Aian Economic and Financial Review hall not be reponible or anwerable for any lo, damage or liability etc. caued in relation to/ariing out of the ue of the content. 05 AESS Publication. All Right Reerved. 39