Japan and the World Economy 41 (2017) 71–86

Empirical evidence of coercive tender offers in Japan

Marc Bremera,*, Kotaro Inoueb, Hideaki Kiyoshi Katoc

a School of Business Administration, Nanzan University, Japanb School of Engineering, Tokyo Institute of Technology, JapancGraduate School of Economics, Nagoya University, Japan

A R T I C L E I N F O

Article history:Received 3 October 2015Received in revised form 22 August 2016Accepted 18 January 2017Available online 4 February 2017

JEL classifications:G32G34

Keywords:Mergers and acquisitionsTender offersCorporate governanceJapan

A B S T R A C T

This is an empirical investigation of the degree to which legal rules impact the welfare of minorityshareholders in acquisitions. While an efficient market for corporate control is vital for an economy’sgrowth and development, insufficient legal standards may permit coercive takeovers that have negativeimplications for capital markets. This research focuses on tender offers in Japan, where legal rules provideacquirers with the opportunity to make coercive takeovers that expropriate minority shareholder wealth.Japan’s legal system changed in 2006 to introduce cash mergers to freeze-out remaining shareholdersafter successful takeovers, and in 2007 to require bidders making tender offers that seek more than two-thirds of the voting securities of a target to offer to buy all the shares. However, acquirers with the statedaim of securing less than two-thirds of voting securities have no such obligation. We find evidence thatthese acquirers tend to make coercive two-tier offers that expropriate the interests of minorityshareowners. Our results suggest that avoiding coercive takeovers requires that laws force acquirers toprovide full information concerning the clean-up merger conditions as well as to pay an equivalentamount in the clean-up to minority shareowners as was offered in the initial tender offer withoutambiguity. These conclusions have relevance for all countries that have not fully considered theappropriate level of protection for minority shareholders.

© 2017 Elsevier B.V. All rights reserved.

Contents lists available at ScienceDirect

Japan and the World Economy

journal homepa ge: www.elsev ier .com/locate / jwe

1. Introduction

This research examines the expropriation of minority share-holders by analyzing stock price after corporate takeovers. Oursample is tender offers in Japan, where the legal protection ofminority shareholders is highly rated. Yet Japan’s rules provideacquirers with the opportunity to make coercive takeovers thatexpropriate minority shareholder wealth. Japan recently changedits takeover rules in an effort to better protect minority share-holders; nevertheless, this change may not sufficiently protectshareholders from coercive tender offers. A takeover, or tenderoffer bid (TOB) is an important transaction whereby control of atarget company is transferred to a purchaser who is expected to

* Corresponding author at: Nanzan University, School of Business Administration 18E-mail addresses: bremerm@ic.nanzan-u.ac.jp (M. Bremer), inoue.k.aq@m.titech.ac.j

1 The term “tender offer bid” or TOB is commonly used in Japan. In other countries, thinterchangeably throughout this paper.

2 The U.S., the U.K., and Japan have highly advanced stock markets and active mergers asince the late 1990s. Although the three countries advocate active markets for corporate coto minority shareholders in takeovers. Japan’s protection of minority shareholders is not wrights index than the U.S. Spamman (2010). ranks Japan at the same level as the U.K. Neveduty of block shareholders to minority shareholders.

http://dx.doi.org/10.1016/j.japwor.2017.01.0010922-1425/© 2017 Elsevier B.V. All rights reserved.

more effectively use the firm’s resources.1 Executing a TOB at areasonably low cost from the acquirer’s perspective is important todevelop an active market for corporate control, and an appropriatelegal system is essential to improve the efficiency of financialmarkets. Yet, the interests of minority shareholders must beconsidered as well. This is where legal systems dealing withtakeovers and the protection of minority shareholders becomeimportant.2 It is essential to design a system that balances the twogoals of fostering an active market for corporate control andprotecting minority shareholders.

On the whole, we find that Japanese corporate takeovers createvalue. However, this research demonstrates that the post-TOBvalue of target firms in Japan drops on average and that there is

Yamazato-cho, Showa-ku, Nagoya 466-8673 Japan.p (K. Inoue), kato@soec.nagoya-u.ac.jp (H.K. Kato).e terms “tender offer” and “takeover bid” are more often used. We use these terms

nd acquisitions markets, yet Japan’s market for corporate control became active onlyntrol to achieve economic efficiency, they differ in the details of protection providedeaker. For example, Djankov et al. (2008) rank Japan as having a higher anti-directorrtheless, the details of Japan’s protection differ, particularly in terms of the fiduciary

6 ; Burkart (1999) and Burkart et al. (1998) discuss the implications of freeze-outregulations in detail.

7 Grossman and Hart argued that dilution of minority owners’ interests providesa solution to the free-rider problem. Their research was extended by Burkart et al.(1998), who arrived at very different conclusions. They argued that free-ridingbehavior by minority shareholders has two results. The equilibrium supply of shareswill be increasing in bid price. This happens because shareholders will beindifferent between selling at a low price and retaining their shares only when theyexpect a low minority share value; consequently, only a small proportion of theseshareholders will sell. Yet, as the bid price increases, the minority share value thatmakes them indifferent increases as well. Similarly, the fraction of shares that willbe tendered will increase. Secondly, because the improvement in value will be

72 M. Bremer et al. / Japan and the World Economy 41 (2017) 71–86

coercion on minority shareholders to participate in the TOB evenwhen they are not satisfied with the terms. We find that the drop inthe target’s share price around the expiration date of the TOB is aslarge as 7 percent on average. The price drop is large when there isno information during the TOB about the price conditions of theclean-up merger following the TOB. In addition, for partial bids theprice drop becomes large when prorationing occurs. Thus, lack ofinformation about clean-up mergers and partial bids are probablythe main causes of the target’s stock price fall at the expiration ofTOBs. We argue that these two causes result from coercion onminority shareholders in Japanese TOBs. Our results suggest thatthis coercion is due to Japan’s insufficient legal protection ofminority shareholder interests. This incomplete protection forminority shareholders arises from a lack of a fiduciary duty of blockshareholders and incomplete legal restrictions on two-tieracquisitions. Fiduciary duty means that boards and influentialcontrolling shareholders have a responsibility to act in theinterests of all shareholders, not of only controlling shareholders.Controlling shareholders may not profit from their relationshipwith the firm; they may not favor themselves over othershareholders. One of the causes of this insufficient protection ofminority shareholders in Japan is that there is no clear rule norcourt precedent that establishes the fiduciary duty of controllingshareholders in the country. We argue that practical fiduciary dutymeans that minority shareholders must be protected against lossthrough expropriation by large block holders, must be givensufficient information to make value-preserving decisions withregard to clean-up mergers and must be provided with theopportunity to exit from takeovers at a fair value.

This research documents that Japanese TOBs have a dispropor-tionate number of failures by these criteria. Our research suggests adifferent policy prescription from the now-classic reasoningoffered by Barclay and Holderness (1992) that the premium overthe market price for block shareholders need not be shared withminority shareholders through legislation. We argue that thedifferent protection for minority shareholders in Japan justifies theconcern that minority shareholders are unfairly exploited by largeblock shareholders.

Two significant concerns arise from our results. The first is thehold-up problem under incomplete contracts involving minorityshareholders in public firms.3 This reduces the willingness ofoutside investors to invest in the Japanese stock market,potentially raising the cost of capital to Japanese companies.Second is the possibility that acquirers will adopt the two-tier bidformat that facilitates the expropriation of the interests of theremaining minority shareholders after the TOB and leads to loweracquisition costs compared to any-or-all bids. We find that thisproblematic partial bid format accounts for about half of oursample of takeovers. This is quite a large portion compared to theTOBs in the U.S, or U.K. Inoue and Ikeda (2016) report that theportion of partial bids for TOBs are less than 4 percent in the U.S.,and less than 1 percent in the U.K. between 2010 and 2014.4 Thesepartial bids make the conflict of interest between large share-holders and minority shareholders worse and limit the efficiencygain from the acquisition.5 Consequently, to make acquisitionsmarkets efficient and eliminate the concerns of outside investors, itis desirable to strengthen the legal protection of minorityshareholders in takeovers in Japan to at least match the levelsof the U.S. and the U.K. We suggest this strengthening be done withdexterity because taken too far, some value enhancing acquisitions

3 See Becht et al. (2007).4 Inoue and Ikeda (2016) analyzed takeovers which had values of more than USD

10 million.5 See Burkart et al. (1998).

might be discouraged. The logic of this argument also applies toother countries that have not yet introduced regulations thatattempt to balance the interests of minority shareholders againstbenefits of increasing the number of value enhancing takeovers.

The next section outlines the literature concerning takeoversand the legal protection of shareholders; it also explains the logicof coercive takeovers. The third section discusses the current legalsituation of takeovers in Japan. The fourth section describes ourhypothesis and methodology. The fifth section describes oursample of tenders and shows our empirical analysis. The lastsection offers conclusions and suggests changes to Japan’s takeoverlaws.

2. Theoretical discussion and the literature

The tender offer price presented by the acquirer in a TOB is animportant indication of an acquirer’s potential managerialcapabilities. The minority shareholders of a company targetedby a TOB may choose one of several options: sell their shares for anacquisition premium, reject the offer to wait for a better bidder toappear who is willing to pay a higher premium, or free ride on thevalue created by the acquirer without tendering their shares. Ifmany minority shareholders choose the last option, the number oftendered shares may not reach the number sought by the acquirer,and thus prevent the execution of a value enhancing takeover. Afreeze-out could restrict free riding on the part of minorityshareholders.6 Conversely, if the judgment of minority share-holders in regard to tendering their shares is somehow distorted,the acquisitions market itself may be prevented from functioningefficiently. In a typical case, if the acquirer makes changes to lowershareholder value after the TOB, tendering shares even at a pricemuch lower than their reservation price becomes the optimalchoice for minority shareholders. If the TOB is a success and theacquirer accumulates enough shares to control the target firm, theacquirer may then engage in opportunistic behavior that dilutesthe interests of the remaining minority shareholders. In such cases,minority shareholders are pressured into tendering their shares,thus distorting their decision with respect to the tender.

Research by Grossman and Hart (1980) is the foundation ofmodern analysis of the regulation of transfers of control ofcompanies. They showed that free riding by target firm share-holders can prevent value-enhancing transfers of control. Theessential idea is that shareholders of poorly managed firms will notsell their shares unless the offer price matches the post-takeovershare value. An acquirer who can better manage the firm will haveto share the increase in firm value with these free-ridingshareholders.7 Burkart et al. (1998) extended their analysis butreached a very different conclusion. Acquirers will maximize their

shared with the minority owners, the bidder will rationally attempt to maximize hisprivate gains from the takeover. He will do this by acquiring the minimumpercentage of voting shares necessary for control of the target firm (say 51 percentof voting shares). Burkart et al argued that this is an inefficient equilibrium with asmaller efficiency gain than that which would be expected if the acquirer gets fullownership rights of the target after the TOB and concentrates on maximizing thevalue of the target firm.

M. Bremer et al. / Japan and the World Economy 41 (2017) 71–86 73

private gains by securing the minimum proportion of sharesnecessary to control the target. In this case, acquirers can extractprivate benefits from the target at the expense of minorityshareholders. There is good reason to suspect that large share-holders extract substantial private benefits in Japan. Consider anexample from 2008: An investor claimed in the shareholdermeeting of NEC Electronics, which is a listed subsidiary of NEC, thatNEC had forced NEC Electronics to supply large amounts ofcomponents at unprofitable prices to NEC (Nikkei Business Watch,2008). Shareholders of companies which are controlled by a blockshareholder may have to contend with this kind of expropriationby controlling shareholders through non-arm’s length transactionsor capital transactions between the parent and the subsidiary. Infact, there are more than three hundred listed subsidiaries of otherlisted firms in Japan (roughly nine percent of all listed firms) whichsuggests that there are substantial private benefits that can berealized by block holders of public firms.8

Comment and Jarrell (1987) provided an analysis of thedifference between the tender offer price and the share priceimmediately after the expiration of the offer in the U.S. for 210deals executed between 1981 and 1984. Their results show that theshare price of target firms after TOBs with any-or-all bids (70percent of their sample) fell only 3.4 percent; however, thedecrease for two-tier bids was 17.9 percent (20 percent of theirsample consisting of mainly tenders with a limit on the maximumnumber of shares to be purchased followed by clean-up mergers).For partial bids (10 percent of their sample), the fall was 20.9percent. Comment and Jarrell analyzed deals from a period beforethe introduction of strong legal protections for minority share-holders in two-tier bids. Yet, on the basis of their any-or-all bidsample, it seems that minority shareholder protection has workedreasonably well in the United States. And as mentioned in theintroduction, Inoue and Ikeda (2016) report that the portion ofpartial bids of TOBs are less than 4 percent in the U.S. between 2010and 2014. This is consistent with Subramanian (2007) whosuggested that the stronger legal protections for minority share-holders in the U.S. introduced in the last part of the previouscentury are responsible for the lower proportion of two-tier bidsand partial bids.

Barclay and Holderness (1989) analyzed large block sharetransfers in the U.S. and found that minority shareholdersbenefited from block trades in the absence of legislation thatrequires sharing the control premium with minority shareholders.They argued that such legislation is not necessary since it woulddiscourage a more efficient manager from acquiring block sharesand improving the efficiency of the firm. They also noted that thefiduciary responsibility of managers to all shareholders mayexplain why minority shareholders in the U.S. do not seem to beexploited.

As described above, the topic of protecting minority sharehold-er interests from the influence of large shareholders is explored inmany articles in law and finance.9 These articles fall into twocategories. One concerns the degree to which large shareholderscan exploit minority investors through opportunistic behavior,thus curbing acquisition effectiveness and justifying legal

8 Chernenko et al. (2012) find empirical evidence justifying this concern. Yet, adifferent result was reported by Dyck and Zingales (2004) who analyzed thedifference between the acquisition price of block shares and remaining share valueafter the announcement of block share transfers in 39 countries using data from the1990s. They interpreted this difference to be private benefits acquired by blockshareholders. In 21 cases involving Japan, they report a �4 percent averagedifference. They concluded that Japan was a country without large private benefitsfor block shareholders. We argue that it is not appropriate to draw a conclusion fromsuch a small sample.

9 See Burkart (1999).

protections for minority shareholders.10 The other category iswhen investors anticipate expropriation by large shareholdersprior to the fact and rationally protect themselves by discountingthe target’s shares. This discount potentially results in ex antefairness. In addition, anticipation of exploitative behavior byminority investors may induce large shareholders who have ahigher valuation of the target to make a binding, voluntarycommitment to protect minority shareholders. Consequently, iflarge share ownership lowers the agency costs of management,legally forcing large shareholders to guarantee ex post fairness mayreduce their ability to discipline management. Such counter-productive regulation is, of course, undesirable. For example,Shleifer and Vishny (1997) argued that minority shareholders areprotected because of the desire of large Japanese shareowners topreserve their reputations, or through the existence of implicitcontracts that assure ex post fairness.

From the simple rational perspective of self-interest, if theacquirer considers the TOB to be an initial step towards fullacquisition of the target, that acquirer cannot be expected tocommit wholeheartedly to minority shareholders’ ex postinterests. Japan imposes little fiduciary responsibility on theacquirer toward the minority owners. In addition, minorityshareholders cannot protect their interests by revising theirinvestment in the firm to discount anticipated expropriation bythe acquirer because takeovers are not easy to predict. Thus, in aTOB, it is desirable to ensure ex ante fairness by legally restrictingthe expropriation of minority shareholders or allowing minorityshareholders to seek remedies in a legal system that acknowledgesthe fiduciary responsibility of controlling owners to minorityowners.

3. The logic of coercive takeovers and hypotheses

3.1. Takeovers

The logic of coercive takeovers can be appreciated with anumeric illustration. Suppose that value of the target firm prior tothe tender offer is 100. Suppose also that the acquirer canpotentially increase the target’s value to 160. Thus, the synergyvalue of the acquisition is 60 (=160�100). Suppose further that theacquirer offers to purchase 100 percent of the target’s shares withan any-or-all bid. Assume that the acquirer has to offer 130 toconvince at least two-thirds of the target’s shareholders to selltheir shares (which corresponds to a 30 percent acquisitionpremium—about what we actually observe in our sample).Ownership of two-thirds of the target’s shares is the thresholdfor full control in Japan. In this case, the net value of the acquisitionfor the acquirer is 30 (=160�130), and the net value for the target’sshareholders is 30. If some of the target’s shareholders choose notto tender their shares, the acquirer can transfer part of the synergyvalue from the target to itself through non-arm’s length businesstransactions and/or through a clean-up merger at some unspeci-fied future date. The incentives for the acquirer to transfer wealthfrom the target to itself increases in proportion to the size of theremaining target shareholders’ stake, as was pointed out byBurkart et al. (1998). Thus, even after an any-or-all bid, we expectto observe a drop in the target’s stock price at the expiration date ofthe offer, and in particular, when clean-up merger terms are notclear during the tender offer period.

10 See La Porta et al. (2002) and the other studies in their research series. See alsoDyck and Zingales (2004).

74 M. Bremer et al. / Japan and the World Economy 41 (2017) 71–86

Moreover, if the acquirer can gain effective control of the targetwith 50 percent ownership, the acquirer might prefer to purchasethis proportion with a partial bid. Assume that the acquirer canpurchase 50 percent of the target’s shares with bid of 120 (whichcorresponds to the approximate 20 percent premium we observein our sample) due to the coercive nature of partial bids. Supposethat the acquirer can transfer wealth of 50 (from the 60 synergyvalue of the acquisition) through non-arm’s length businesstransactions between the acquirer and the target in the post-acquisition period. The target share price will drop to 110(=160�50). If the target’s shareholders rationally anticipate thiswealth expropriation, they will attempt to sell their shares at thetender offer price of 120, even though they understand that thesynergy-enhanced value of the firm is 160. However, only 50percent of the target’s shareholders can sell their shares at 120; andafter the tender offer, the target share price will drop to 110.Although non-tendered shares can receive 10 in additional valuewith the partial bid, at least 10 [=50% � (130 � 110)] is expropriatedby the acquirer compared to the case of an any-or-all bid in whichthe acquirer purchases all of the target’s shares. With the partialbid with a maximum of 50 percent of shares, the gain for theproportion of shares that are tendered is 60 in total (=50% � 120)and the gain for the proportion of shares that could not be tenderedis 55 (=50% � 110). Hence, tendered shares receive a greater valuethan non-tendered shares, and the tender offer has coerciveness.11

Thus, target shareholders will receive 115 (=60 + 55) in total, whichis smaller than the 130 received in the case of the any-or-all bid. Onthe other hand, the acquirer’s gain becomes 45 (=160 � 115) whichis larger than the 30 which was the acquirer’s gain in the any-or-allbid in the absence of gains extracted via non-arm’s lengthtransactions. Although it is true that target shareholders arebetter off even with partial bids, expropriation by the acquirerremains possible. This makes partial bids potentially coercive.12

Table 1 is an overview of the events and stock price indicatorssubject to analysis. These events are depicted in the figure shownin Panel A of the table, where the horizontal axis represents timeand the vertical axis represents stock price. There are two keyevent dates: the announcement date of the deal and the day afterthe expiration date of the tender offer. The announcement date ofthe deal is the day on which the TOB price conditions are initiallydisclosed (if the announcement occurred after the closing of thestock market on that day, we use the next trading day as theannouncement date). We assume that the tender offer priceinformation is captured in the target’s stock price on the date thedeal is announced. We define the Acquisition Premium to be theoffer price less the target’s stock price five days before theannouncement date divided by the target’s price five days for theannouncement. After the expiration of the TOB, the stock price ofthe target firm shows the value held by the post-TOB minorityshareholders. This is the Securities Benefit; it is defined as thetarget’s stock price five days after the expiration of the TOB less thestock price five days before the announcement date of the offer

11 Note that it would not be logical for target shareholders to decide not to tenderin this kind of partial bid. In such a case, they will clearly be worse off. Shareholderswho do not tender would receive only 110. Shareholders who were able to tender allof their shares would receive 120. In effect, the no-tender decision transfers 10 fromnon-tendering shareholders to tendering shareholders. As a practical matter, whenmore than the maximum number of shares is tendered in response to a partial bid,the purchased number of shares is split equally among all tendered shares. Eachstockowner ends up with the same proportionate shares and payment.12 This illustration has the feature that minority shareholders are assumed not tobe able to prevent wealth transfers from the target to the acquirer. This was the real-world case in the NEC example mentioned in the previous section. It should benoted that in cases where wealth transfers are limited, the coercive nature of partialbids may be less. Minority shareholders were actually able to sell as many shares asthey like in about half of the partial bid deals in our sample.

divided by the price before the announcement. On the other hand,the tender offer price shows the value of shares from theperspective of the acquirer. If the price of the target’s stock falls,the amount of the decrease (Premium over Post-TOB Price) is thedifference between the value of one share priced by the acquirer inthe TOB and the share priced by the remaining minorityshareholders after the TOB. We interpret this difference torepresent the acquirer’s private benefits. For example, if a clean-up merger is executed by the acquirer at a price lower than thetender offer price, then the diluted portion of the remainingshareholders’ interests is the acquirer’s private benefit. In addition,by exercising his influence over the management of the target firm,the acquirer may engage in self-dealing behavior to obtain benefitsfrom transactions with the target firm after the acquisition. Theseare private benefits to the acquirer at the expense of the targetfirm’s minority shareholders.

3.2. Hypotheses

Our base-line interest concerns Japanese takeovers in general.Based on previous research and the arguments outlined above weassert that Japanese takeovers are value enhancing investments foracquirers:

Hypothesis 1. Japanese Takeovers are value increasing. TheAcquisition Premium is positive.

Similarly, Japanese takeovers are beneficial for target share-holders:

Hypothesis 2. Target shareholders benefit from the greatervalue created by the takeover. The Securities Benefit is positive.

Our next and primary interest, concerns whether Japanesetakeovers have a coercive nature in the sense that the remainingminority shareholders receive less than shareholders who tendertheir shares in TOBs. Opportunistic behavior by Japanese acquirers,or anticipation of this behavior, is expected to lower the value ofminority shareholders’ holdings after takeovers. This value is thedifference between the tender offer price presented by theacquirer, which is based on his own valuation of the target firm’scontrolling stake, and the post-takeover share price that repre-sents the value for the remaining minority shareholders after thetakeover. This difference is value obtained by the acquirer at theexpense of the minority shareholders. This occurs because there isno price rule for clean-up mergers following TOBs, nor rulespreventing coercive two-tier takeovers in Japan, nor a fiduciaryduty by target management and controlling shareholder to act inthe interests of minority shareholders. They do not have to protectthem against loss and they can benefit at their expense:

Hypothesis 3. Opportunistic behavior by acquirers will causethe stock price of the target firm to decline at the expiration dateof the TOB. The Premium over Post-TOB Price will be positive.

Our next hypotheses concern the degree of coercion of Japanesetakeovers. As demonstrated in the illustration in the previoussection, partial and two-tier takeovers might give acquirers theopportunity to coerce target shareholders to tender their shares atoffer prices below their estimation of the shares’ intrinsic value.The possibility that minority shareholders will be frozen out atlower prices in a clean-up acquisition, or that the acquirer willextract private benefits is greater with the partial and two-tierformats. These arguments give rise to the following:

Hypothesis 4a. Takeovers using the partial and two-tier formatwill have a lower Acquisition Premium than takeovers that usethe any-or-all format.

Table 1Description of indicators, Panel A: TOB time line, Panel B: TOB return indicators.

Panel A: TOB time line

Premium ov er Post-TOB Price

Acquisition Premium

Sec urity Benefit

TOB period: 20 trading days at minimum

TOB Expiration Date (ED)TOB Annou nce ment Date

(AD )

Panel B : TOB re turn indicators

[Tend er off er price] − [Stock price on day 5 before the Anounce ment Date]Stock price on day 5 before the Annou nce ment Date

[Stock price on day 5 after the Expiration Date] − [ Stock price on day 5 before the Annouce ment Date]Stock price on day 5 before the Annou nce ment Date

[Tend er off er price ] - [Stock price on day 5 after the Expiration Date]Stock price on day 5 before the Annou nce ment Date

Annou nce ment Return Cumulative Abnormal Return fr om day 3 before to day 3 after the Announce ment Date *3

Exp iration Return Cumulative Abnormal Return fr om day 3 before to day 3 after the Expiration Date *3

*1Acqu isition Premium

Sec urity Benefit

Premium over Post-TOB Price

*1 *2

*1. The Ann oun cement Date is the first ann oun cement date of the tend er offer price . *2 . The Expiration Date is the expiration date of the tend er off er bid. *3 CumulativeAbn ormal R etur n is o btained with the mark et mod el in which the r isk parameters are calculated over a period of 200 days that end s 2 1 d ays prior to the Ann oun cement Date.

M. Bremer et al. / Japan and the World Economy 41 (2017) 71–86 75

Hypothesis 4b. Takeovers using the partial and two-tier formatwill have lower Security Benefit than takeovers that use the any-or-all format.

Hypothesis 4c. The Premium over Post-TOB Price will be higherfor takeovers that use the partial and two-tier format thantakeovers that use the any-or-all format.

As was argued by Burkart et al. (1998), acquirers can maximizetheir private benefits by securing the minimum proportion ofownership that achieves control. This and the fact that acquirers donot have a clear fiduciary duty to minority shareholders suggeststhat the partial and two-tier takeover formats with a limit on themaximum number of shares to be purchased are most likely to beexploitative and coercive. Since in many cases, target shareholderscannot distinguish between the partial format and what willultimately be the two-tier format at the initiation of the tenderoffer, if a tender offer is proposed with a limit on the maximumnumber of shares to be purchased, we term these deals as “two-tier” takeover format from this point. Hence, we argue thatJapanese two-tier takeovers with pro-rationing are exploitative

relative to any-or-all takeovers. In addition, partial and two-tiertakeovers that seek lower share proportions (Burkart et al. (1998))and lack information about the price conditions of the clean-upmergers will have a higher probability of being coercive and also ahigher probability of prorationing. In these prorationed takeovers,shareholders who cannot sell their shares will be coerced to selltheir shares in the market even at a lower price than the TOB priceto avoid expropriation by acquirers in the post-TOB period.

Hypothesis 5a. Pro-rationed two-tier format takeovers havelower Security Benefit than non-pro-rationed takeovers.

Hypothesis 5b. The Premium over Post-TOB Price will be higherfor pro-rationed takeovers that use the two-tier format than fornon-prorationed takeovers.

Fiduciary duty traditionally focuses on the value maximizingadministration of the interests of the principals. Yet, we assert thatanother aspect of this duty for both target firm management andcontrolling shareholder of a target firm is the responsibility of thefiduciary to disclose sufficient information to the principals tomake value preserving decisions. Under Japanese law and in

Table 2Major differences in the regulation of takeovers in the U.S., U.K., and Japan.

Equal Opportunity Rule Mandatory BidRule

Restriction onTwo-tier Bid Price

Freeze-out Rule Duty to Purchase Non-Tendering Shares

Fiducial Duty ofControllingShareholder

Provide participation rights fortransfer of controlling shares tonon-controlling shareholders

Provide fairvalue exit for allshareholders

Fair price rule toprevent coercivetwo-tier bids

Enable freeze-out of minorityshareholders and prevent free-riding by minority shareholders

Provide remainingshareholders an exitoption in the Post-TOBperiod

Prevent self-dealing bycontrollingshareholders

U.S. None None Yes >90% No YesU.K. >30% >30% Yes >90% Yes YesJapan >33.4% >66.7%1 None None2 No None

1 This description of the Mandatory Bid Rule in Japan applies since 2006 when the Financial Instruments and Exchange Law was enacted.2 Japan introduced a freezeout rule without the requirement of a shareholder’s meeting from April 2015.

13 From 2015, squeeze-out transactions without the approval of shareholders at ameeting are permitted under the revised Corporate Law if the bidder acquires morethan 90 percent of the shares of the target. Still, bidders are allowed to squeeze-outthe remaining shareholder at any time after the tender offer if approved in ashareholders’ meeting.14 See Ito (2008). In the U.S., controlling shareholders do have fiduciary duties tominority shareholders. See Sinclair Oil Corporation v. Levien, 280 A.2d 717 (Del.1971).15 The Tokyo Stock Exchange (2006) asked listed firms to disclose the acquirer’splans to deal with minority shareholders, with such things as the plan for freeze-outs and clean-up mergers in a formal request issued on August 28, 2006 (TSE No.1338). However, there is no obligation for listed firms to disclose price details for thefreeze-out and clean-up mergers. Freeze-outs and clean-up mergers without ex

76 M. Bremer et al. / Japan and the World Economy 41 (2017) 71–86

certain circumstances, block shareholders are not obligated toshare information about the details of the price conditions in theclean-up mergers following the TOBs. When detailed priceconditions under the clean-up mergers are not disclosed at theinitial announcement of the TOB, target shareholders mustevaluate the tender offer, and decide whether to subscribe whiledealing with several unknowns including the exact terms of thefreeze-out transaction. They are not assured that they can sell theirshares at the tender offer price in the post tender offer period ofsuccessfully completed TOB. This means that partial and two-tierdeals are more likely to be exploitive and coercive.

Hypothesis 6a. Security Benefit will be smaller for takeoversthat do not reveal details about the clean-up merger.

Hypothesis 6b. The Premium over Post-TOB Price will be higherfor takeovers that do not reveal details about the clean-upmerger.

In a related way, partial and two-tier takeovers that offer agreater threat of wealth transfers are more likely to be coercive andexploitive. So we also hypothesize that partial and two-tier formattakeovers by acquirers who are in the same or related industries asthe target will have lower Security Benefit and a higher Premiumover Post-TOB Price. We will also consider these hypotheses withan examination of the abnormal returns of the target and theacquirer. Finally, and as discussed below, Japan recently changedthe laws that concern the protection of minority shareholders. Weexplore how these legal changes impact the welfare of minorityshareholders in targets in the context of these hypotheses.

4. Legal protection of minority shareholders facing tenderoffers in Japan

Tender offers in Japan are regulated by the Securities andExchange Law. All bidders who would own more than one-third ofthe voting rights of the target after off-market purchases arerequired to make a tender offer. This is different from the rules inthe U.S. that allow private transfers of block shares in the upstairsmarket. Japan’s Financial Instruments and Exchange Law whichwas enacted in 2006 mandated an any-or-all rule that requires abidder who will own two-thirds or more of the voting rights in thetarget firm to make an offer to purchase all classes of votingsecurities. This is different from a related rule under the City Codein the U.K., where a tender offer that seeks to acquire 30 percent ormore of the voting rights in the target is required to offer topurchase all the remaining equity after the tender offer. In the U.S.,bidders are allowed to make partial bids, but it is costly for biddersto gain further control due to a series of fairness rules that apply ifthey do not squeeze-out all remaining shareholders with thetender offer. Thus, in recent years, partial and two-tier formattakeovers have become less popular in the U.S. Table 2 summarizes

the major differences between the legal systems protectingminority shareholders in the U.S., the U.K., and Japan as of 2015.In essence Japan’s takeover rules do not limit coercive behavior bycontrolling shareowners. Further, controlling shareholders do nothave a fiduciary duty to minority shareholders.

Mergers and share swaps were regulated by the CommercialCode until April 2006 and have been regulated by the CorporateLaw from May 2006. Both laws require approval by two-thirds ormore of the shareholders present at the shareholders’ meetings,both for the acquirer and for the target, to execute mergers orstock-for-stock acquisitions. Under the new Corporate Law,acquirers are allowed to use cash as a payment method formergers instead of the acquirer’s shares for the first time in Japan.In addition, the Corporate Law introduced innovative ways torecapitalize a company. These changes enable controlling share-holders to freeze out minority shareholders easily.13

The intent of the mandatory any-or-all bid rule was tocounterbalance the relaxation of freeze-out constraints. Becausea dominant shareholder who controls two-thirds or more of thevoting shares of the target can freeze out the remaining minorityshareholders at his discretion, minority shareholders are providedan exit in the form of a tender offer. However, there is no regulationthat prohibits coercive two-tier bids in which an acquirer pushesminority shareholders to sell their shares by setting the freeze-outprice lower than the initial tender offer price, or by simply notdisclosing his plan for how the remaining shareholders will betreated after the completion of the tender. Although, under boththe previous Commercial Code and the new Corporate Law,shareholders who object to the proposed merger have the right torequest that the firm repurchase their shares at a “fair price”, theprocess of appraisal and remedy is costly. Further, the outcome isextremely uncertain if contested in the courts. Under Japanese lawcontrolling shareholders do not have fiduciary responsibility tominority shareholders.14 There are no provisions concerning thetreatment of minority shareholders after the takeover and theirprotection is a matter covered by the Corporate Law.15 Thus, thereis the valid concern that a controlling shareholder might engage inself-dealing behavior to expropriate minority shareholders. This

post disclosure are not legally prohibited.

M. Bremer et al. / Japan and the World Economy 41 (2017) 71–86 77

may compel minority shareholders to sell their shares even whenthey are dissatisfied with the tender offer price.

The Financial Instruments and Exchange Law expanded the scopeof the bidder’s disclosure obligation. Under the amended rules, thebidder is required to disclose the purposes of the tender offer, thebusiness plans after the completion of the takeover (including anyplans to reorganize or delist the target), and the basis for the offerprice in the tender registration statement. Stock exchanges also haveequivalent disclosure requirements for listed companies. In practice,a bidder must disclose his plan to freeze out minority shareholdersafter August 2006. Yet, he is not legally required to pay the sameamount to minority shareholders in the freeze-out transaction thathe paid in the initial tender offer. Since there is no law that clearlydefines and regulates freeze-out transactions or two-tier bids in theanalyzed period, acquirers can make “coercive two-tier bids” byseparating the tender offer and the subsequent minority freeze-out.Thus, although under the current tender offer rules minorityshareholders are protected by the mandatory any-or-all bid ruleto some extent, they still face the possibility of coercive two-tier bidsor ex post expropriation through self-dealing behavior by theacquirer. For these reasons, we anticipate that the acquirer willexpropriate minority shareholders, or that the stock market willanticipate expropriation. Hence, the interests of minority share-holders who did not subscribe to the TOB will systematicallydeteriorate compared to those who successfully tendered theirshares. In such a case, minority shareholders will be pressured intotendering their shares.

There are many cases that justify concern about coercive two-tier bids in Japan.16 An example is the acquisition of HanshinRailway by Hankyu Holdings in 2006. Hankyu declared itsintention to acquire Hanshin through a combination of a TOBand a follow-up stock-for-stock acquisition. Hankyu set its tenderoffer price at 930 yen per share, which was slightly lower than themarket price of Hanshin Railway stock at the time. This wasbecause of Hankyu’s intention to purchase only the 47 percentstake held by a hostile activist fund. The theoretical acquisitionprice of Hanshin shares in the subsequent stock-for-stock clean-up,calculated from the previously announced stock-for-stock ratioand the market price of Hankyu Holdings, was 732 yen. This was amore than 20 percent discount from the tender offer price. At theexpiration date of the tender offer, the chief executive officer ofHankyu Holdings said in an interview that the proportion of sharesacquired was more than he expected.17 We suspect that the head ofHankyu was not genuinely surprised because this deal was anunambiguous coercive two-tier bid.

5. Empirical analysis

5.1. Sample

Following the approach of Comment and Jarrell (1987) andDyck and Zingales (2004), the hypotheses described in sectionthree are tested by analyzing the difference between the tenderoffer price, the post-TOB share price and where appropriatecumulative returns for targets and acquirers.18 We use the Nikkei

16 The Nikkei Shimbun (2007) reported that among the twelve two-tier bidsexecuted in the year, the freeze-out prices were set lower than the tender offerprices in nine of the deals.17 See Nihon Keizai Shimbun (2006a) and Nihon Keizai Shimbun (2006b). See alsoInagaki (2013) and Lewis and Inagaki (2017) for other vivid examples.18 Dyck and Zingales (2004) excluded tender offers from their sample. Since, incountries with legal rules to restrict coercive two-tier bids, no difference betweenthe tender offer price and the post-TOB share price is expected, excluding thesetenders is a reasonable procedure. Because this is not the case in Japan, we focus onthe price difference between the tender offer price and post-TOB price.

NEEDS M&A database which is provided by Nikkei, a Japanesemedia group focusing on economic and business news. We use thisdatabase because it provides the most comprehensive and reliabletender data in Japan. Among the many databases that provide M&Ainformation related to the Japanese market, only this databaseprovides the information necessary for this study. These datainclude the number of shares sought by acquirers, the number ofshares actually tendered by target shareholders and the amountactually purchased in the tender offer. We also collect informationrelating price conditions in the clean-up mergers, if any, from TOBfilling information. We collect stock price data, financial data andownership data for each firm from the Nikkei NEEDS database.

The sample consists of the all Japanese takeovers announcedbetween January 2000 and March 2007 whereby shares wereactually obtained by the acquirer and excludes tender offersmotivated by stock repurchases.19 We obtained data on 230transactions from the database. Five of these deals failed tocomplete and 52 of these deals had tender offer prices less than theshare prices five days before the TOB announcements. Thesediscount tenders were actually pre-arranged block share trans-fers.20 After excluding these failed and rescue deals our sample sizeis reduced to 173 takeovers.

We separate the takeover deals into two sub-groups in Table 3,using an approach similar to that of Comment and Jarrell (1987).One group is “TOB with Maximum Limit” which are two-tierformat, or partial bids, in which there is a maximum limit on thenumber of shares to be purchased. The other group, “TOB withoutMaximum Limit” is deals without a maximum limit on the numberof shares to be bought. As described in the literature andhypothesis sections, we anticipate that coercion of target share-holders will be less of a problem for TOB without Maximum Limitdeals. Target shareholders are free to sell, or wait for a better offer(or even attempt to free ride on the value created by the acquirer).Takeovers with a Maximum Limit on the number of shares to bepurchased present a much thornier decision. Under the two-tierformat, target shareholders must evaluate the tender withoutknowing whether the offer will be pro-rationed, the likelihood thatthe acquirer will extract private benefits and the conditions of thepotential clean-up deal. This is far more complicated in Japan thanthe evaluation of tenders in the U.S. and the U.K., since there is nobest price rule as in the U.K. or restrictions on coercive two-tierbids as in the U.S. There is also the real possibility the acquirer willuse his controlling position to extract value from the post-takeovertarget in related party transactions. We anticipate that TOB withMaximum Limit deals are likely to be subject to coercion, and thatthis coercion effect will be captured by the stock price at the timeof the tender offer.

As shown in Table 3, TOBs with maximum limit result in lowerpost-deal ownership by the acquirer (59 percent on average and 53percent as a median). This is consistent with the Burkart et al.(1998) prediction that, for cases where private benefits can beacquired by controlling shareholders, the minimum required tocontrol the target becomes the equilibrium for the acquirer.Minority shareholders cannot sell all the shares that they desire tosell in about half of the TOBs with a Maximum Limit. This suggeststhat minority shareowners face genuine expropriation risk in

19 We end our sample in 2007 to avoid tender offers that may have beeninfluenced by the market volatilities associated with the Lehman shock. As Bakeret al. (2012) show for U.S. cases and Ozawa et al. (2016) show for Japanese cases,prior stock price peaks have significant effects on acquisition premiums. As apractical matter, we observe large premiums for tender offers in 2008 and 2009. Inaddition, the substantial market movements in 2008 and 2009 make it very difficultto estimate abnormal returns using the market model (and other models).20 Although not shown in the tables, these tender offers do not have a significanteffect on target share prices.

Table 3Descriptive statistics of Japanese tender offers.

Hostile AcquisitionPremium

AcquirerToehold

MaximumTOB Shares

PostTOBShare

TargetMarketCap (MilJPY)

TargetPBR

DeficitTarget

ProrationingDeal

PublicAcquirer

AcquirerMarketCap (MilJPY)

HorizontalAcquirer

FundAcquirer

TOB withMaximum Limit (N = 97)

Mean 3% 23% 31% 37% 59% 16,356 1.65 29% 44% 82% 917,191 55% 7%Median 19% 42% 36% 53% 4188 1.13 207,817

TOBwithoutMaximum Limit (N = 76)

Mean 1% 28% 37% 62% 91% 46,654 1.50 24% 0% 28% 574,029 42% 33%Median 26% 36% 57% 95% 10,761 1.12 67,568

This table shows descriptive statistics of Japanese tender offers from January 2000 to March 2007 from the Nikkei NEEDS Mergers and Acquisitions Database. This database isthe most complete and reliable record of acquisitions in Japan. Deals categorized as “TOB with (without) Maximum Limit” are takeovers in which the maximum number ofshares purchased is specified (is not specified). Hostile is the proportion of the deals categorized as hostile by the Japanese financial press. The Acquisition Premium is thetender price divided by the stock price 5 days before the TOB less one. Acquirer Toehold is the proportion of the TOB deals in which the acquirer already owns some of thetarget’s shares. Post TOB Share is the proportion of ownership that the acquirer holds after the TOB. Target Market Cap is the market capitalization measured in millions ofJapanese yen of the target in the fiscal year before the TOB. Target PBR is the price to book value of the target based on the previous year’s market and book values. ProrationalDeal is the proportion of deals that are subject to prorationing when the number of tendered shares exceeds the maximum limit to be purchased. Deals categorized as PublicAcquirer, Horizontal Acquirer and Fund Acquirer are the proportions of deals in which the acquirers are public firms, firms in the same industry and investment funds.Acquirer Market Cap is the market capitalization measured in millions of Japanese yen of the acquirer in the fiscal year before the TOB.

78 M. Bremer et al. / Japan and the World Economy 41 (2017) 71–86

clean-up deals, or through related-party transactions. “MaximumTOB Shares”, which is the maximum percentage of shares thatacquirers intend to purchase under the TOB, is also higher for TOBswithout maximum limit. The percentage is not 100 percent sinceacquirers usually have a toehold ownership stake in the target’sshares. The mean of the acquirer’s toehold is 31 percent in TOBswith a maximum limit and 37 percent in deals without a maximumlimit.

In terms of market capitalization, TOB without Maximum Limitdeals involve larger acquirer firms and target firms. Comparingtotal market capitalization (median) of acquirers and target firms,for TOB with Maximum Limit deals, the acquirer is typically over20 times larger than the target.21 These conditions are inconsistentwith the idea that acquirers opt for the TOB with Maximum Limitformat due to capital constraints.

Further, there are many cases where TOB without MaximumLimit deals are used by investment funds, whereas listedcompanies tend to use TOB with Maximum Limit deals.Acquisitions by firms in the same industry as the target areapproximately balanced between the maximum limit format andany-or-all format. These horizontal acquisitions are interestingbecause they have the potential to create substantial value fromsynergies and economies of scale. Perhaps in these horizontaldeals, minority shareholders will be less concerned with post-dealexpropriation through non-arm’s length transactions; indeed, theymay prefer not to tender their shares in the hope of free-riding onthe greater value created by the acquisition in spite of post-dealexpropriation. And by similar reasoning, acquisitions by fundsmight be expected to have few synergies, but that targetshareholders may have less concern about post-deal expropriation.One reason (aside from value extraction) for selecting theMaximum Limit format may be that the parent company’s listedstock can be used in the clean-up merger. On the other hand, casesof partial acquisition often result in parent-subsidiary dual stock

21 The relatively small size of the targets compared to the acquirers suggests thatthe value created (or the amount overpaid) by the acquisition may be too small to beobservable from examination of the stock price of the acquirer. Based oninformation for listed acquirers, the average market capitalization of acquirersfor takeovers with a maximum limit on shares to purchase was 917,191 million yen;this value is far larger than the average market capitalization of targets at 16,356million yen.

market listings. Although the average percentage share of controlafter a Maximum Limit deals is 59 percent, 85 percent (not shownin the table) of the corresponding target firms maintained theirstock market listing (as of March 31, 2009).

5.2. TOB transactions in Japan

Panel A of Table 4 shows stock price data for the 173 TOBs in oursample. Clearly the takeover announcement is good news. Onaverage, target firms’ abnormal return rises by 21 percent when thetakeover becomes public information. The acquisition premiumaverages a significant 25 percent. So, there is good reason to thinkthat these deals create value. This is clear evidence in favor ofHypothesis 1. And this is also consistent with the idea that aneconomy such as Japan’s that is approaching maturity cangenuinely benefit from the corporate restructurings made possibleby an active market for corporate control. The acquirer’s privatebenefit, as measured by the Premium over Post-TOB, is a significant10.5 percent. On average the target’s abnormal return falls after theexpiration day of the TOB. This supports Hypothesis 3. Yet, targetshareholders appear better off in the sense that the SecuritiesBenefit averages a significant 14.8 percent. This result supportsHypothesis 2. Inoue and Ikeda (2016) report that a statisticallysignificant price fall at the expiration dates of TOBs is only observedin Japan among TOBs that occurred between 2010 and 2014 in theU.S., U.K., Germany and Japan.

Panel B of Table 4 shows stock price performance around TOBevent dates for the sample split into the TOB with Maximum Limitand TOB without Maximum Limit categories. As anticipated, targetstock returns appreciate significantly when TOB without Maxi-mum Limit deals are announced (24.5 percent). This is significantlylarger than the appreciation for TOB with Maximum Limit deals(18.5 percent). The Acquisition Premium is larger for TOB withoutMaximum Limit than for TOB with Maximum Limit; however, thisdifference is statistically significant only for the median, not themean.22 TOB deals experience a significant share price drop after

22 This is consistent with the argument by Burkart et al. (1998); the acquisitionpremium is increasing with the target control ratio of the acquirer. Thus, acquirershave a strong motivation to avoid bidding for all the shares outstanding; this allowsthem to avoid paying large acquisition premiums.

Table 4Changes in stock price near tender offers dates.

Panel A All TOBs

N Mean Median t-stat Z-stat

Target Announcement Return (�3, +3) Mean 173 21.1% 16.942***Median 19.7% 11.150***

Target Expiration Return (�3, +3) Mean 173 �6.9% �7.470***Median �4.5% �8.192***

Target TOB Return (�5 of AD, +5 of ED) Mean 173 13.3% 7.442***Median 9.1% 7.534***

Acquisition Premium Mean 173 25.3% 16.762***Median 22.2% 11.382***

Premium over Post-TOB Mean 173 10.5% 6.477***Median 4.7% 8.492***

Security Benefit Mean 173 14.8% 8.834***Median 12.3% 8.160***

Acquirer Announcement Return (�3,+3) Mean 101 �0.4% �0.361Median �0.2% 0.239

Acquirer Expiration Return (�3, +3) Mean 101 �0.3% �0.574Median �0.3% �0.994

Panel B: TOB with and without Maximum Limit of shares to Purchase.

TOB with Maximum Limit TOB without Maximum Limit Difference

N Mean/Median N Mean/Median

Target Announcement Return (�3, +3) Mean 97 18.5%*** 76 24.5%*** �6.1%**Median 14.4%*** 23.3%*** �8.8%***

Target Expiration Return (�3, +3) Mean 97 �6.7%*** 76 �7.1%*** 0.4%Median �4.6%*** �4.3%*** �0.4%

Target TOB Return (�5 of AD, +5 of ED) Mean 97 13.0%*** 76 13.7%*** �0.7%Median 7.7%*** 11.6%*** �3.9%

Acquisition Premium Mean 97 23.3%*** 76 27.9%*** �4.6%Median 18.7%*** 25.7%*** �7.0%***

Premium over Post-TOB Mean 97 12.0%*** 76 8.7%*** 3.4%Median 6.7%*** 3.9%*** 2.8%***

Security Benefit Mean 97 11.3%*** 76 19.2%*** �8.0%**Median 7.4%*** 17.3%*** �9.8%***

Acquirer Announcement Return (�3,+3) Mean 80 �0.9% 21 1.5% �2.4%Median �0.9% 3.0% �3.9%

Acquirer Expiration Return (�3, +3) Mean 80 �0.5% 21 0.3% �0.8%Median �0.4% 0.5% �1.0%

This table shows abnormal returns and price relationships associated with the tender offers described in Table 3 which are based on the Nikkei NEEDS mergers andacquisitions database for Japan over the period from 2000 to 2007. Mean shows the average value for the associated indicator. Median is the median value for the associatedindicator. The columns with asterisks show results for tests of the hypotheses that the mean or median are not importantly different from their expected zero value. The meanhypothesis test uses a t-test; the median hypothesis test uses a Wilcoxon test. Full descriptions of the indicator values are reported in Table 1. Target (Acquirer) Announcement(Expiration) Return is the cumulative abnormal return of target (acquirer) firms from 3 days before the announcement (expiration) date of the TOB to 3 days after. AcquirerAnnouncement (Expiration) Return is calculated only when an acquirer is listed on a stock exchange in Japan. Acquisition Premium is the tender offer price divided by thetarget’s stock price 5 days before the announcement date of the TOB less one. Premium over Post-TOB is the difference between the tender offer price and the stock price of thetarget 5 days after the expiration date of the TOB divided by the stock price 5 days before the announcement date. When the Premium over Post-TOB Price is positive, it meansthat the stock prices of target firms have fallen from the tender offer price. Security Benefit is the stock price 5 days after the expiration date divided by that 5 days before theannouncement date of the TOB less one. Deals categorized as “TOB with (without) Maximum Limit” are tenders in which the maximum number of shares to be purchased isspecified (is not specified). ***, ** and * indicate statistical significance at the 1, 5 and 10 percent levels of confidence.

23 See Roll (1986).

M. Bremer et al. / Japan and the World Economy 41 (2017) 71–86 79

the TOB. If we look at the Premium over Post-TOB Price, for TOBwithout a Maximum Limit, the fall is the smallest at 8.7 percentand larger for TOB with Maximum Limit deals at 12 percent. Thedifference is statistically significant in the nonparametric test ofmedians. As hypothesized, the Securities Benefit is larger for TOBwithout Maximum Limit; this result is statistically significant forboth mean and medians. Thus, the average Security Benefit for TOBwithout Maximum Limit deals is significantly greater (19.2percent) than TOB with Maximum Limit deals (11.3 percent). Onthe whole, the results in Panel B provide modest, if not strong,support for all parts of Hypothesis 4a–4c. The market puts a lowervalue on deals that may be subject to coercion.

Comment and Jarrell (1987) report an average 3.4 percent dropin stock price after any-or-all tenders while the average fall forJapanese TOBs without Maximum Limit is much larger. Oneplausible explanation for this difference is the fact that there aremany deals where the clean-up merger details are unclear at thetime of the tender. We discuss this point in detail below. In general,for minority shareholders, waiting for the clean-up merger is

disadvantageous on average, compared to tendering their shares tothe initial offer. As a matter of fact, the average target stock priceone day before the target is delisted in clean-up mergers that wereconducted within two years of the TOB is lower than the target’sstock price 5 days after the expiration date of the TOB. The target’sstock price is down on average �7.4 percent for TOBs with amaximum limit on shares to be purchased, and �2.5 percent forTOBs without a maximum limit on shares to be purchased (notreported in the tables). This fact is consistent with our argumentthat Japanese TOB are coercive due to lack of sufficient legalprotection for minority shareholders.

The cause of the post-TOB drop in stock price may not be solelya result of the fear of expropriation of the interests of minorityshareholders by the acquirer. It could come about from anoverestimate of the value of the target firm and the resultantsynergy effects of the takeover on the part of the acquirer.23 To

Table 5Test of acquirer overpayment.

Sample: Premium over Post-TOB Price >Median (=5.2%) =<Median (=5.2%) Difference

N 50 51Mean t-stat Mean t-stat Difference t-stat

Acquisition Premium 28.50% 9.219*** 16.13% 8.297*** 12.37% 3.386***Acquirers: Announcement Return (�3, +3) 0.32% 0.232 �1.11% �0.630 1.43% 0.639Acquirers: Expiration Return (�3, +3) �0.03% �0.041 �0.61% �0.784 0.58% 0.506

This table shows analysis of mergers and acquisitions in Japan over the period from 2000 to 2007. This table shows results of t-tests of the cumulative abnormal returns ofacquiring firms to examine the hypothesis that the fall in the target’s stock price relative to the tender offer price means that the market anticipates that the acquirer is payingtoo much. The sample is all cases where the acquirers are public firms. Acquisition Premium is the tender offer price divided by the target’s stock price 5 days before theannouncement date less one. Announcement Return and Expiration Return are cumulative abnormal returns from 3 days before to 3 days after the initial announcement dateof the tender offer price and the expiration date of the tender offer respectively. *** indicates statistical significance at the 1 percent level.

80 M. Bremer et al. / Japan and the World Economy 41 (2017) 71–86

explore this possibility we follow the approach of Dyck andZingales (2004) to divide the sample into two subsamples on thebasis of the size of acquirer’s private benefit (Premium over Post-TOB Price) and compare the returns of bidding firms in each group.The results of this analysis are shown in Table 5; results are shownonly for cases where acquirers are public firms. If the stock marketbelieves that the acquirer overpaid, a wealth transfer from theshareholders of the acquirer firm to the shareholders of the targetfirm will be reflected in the stock price of the acquirer. Its stockprice should decrease. However, the results show that the returnsto acquirers in both groups are not significantly less than zero. Inaddition, the difference between the two groups is not statisticallysignificant. Therefore, there is no evidence that supports the hubrisinterpretation of overpayment by the acquirer.24

On the other hand, we do not observe positive returns foracquirers both on the announcement date and at the expirationdate even for deals with a Maximum Limit. At first thought, thisseems to suggest that acquirers pay a fair price for value-enhancingdeals. It also seems to contradict the view that acquirers benefitfrom tender offers with a coercive nature. Nevertheless, there arenumerous and fairly convincing stories in the business press aswell as research by Chernenko et al. (2012) that suggest privatebenefits are extracted by controlling shareholders. Relative sizemight explain the lack of observable evidence in our results. Inmost deals, the marginal benefit for acquirers is relatively smallsince the acquirer actually purchases only a small fraction of theshares in these deals. The scale of these deals might be too small toobserve as statistically significant. Note that the median acquirer’stoehold is 42 percent and the post-tender offer share is 53 percent.Furthermore, these target companies are relatively small capitali-zation firms compared to acquirer firms (the median ratio ofmarket capitalizations of targets to acquirers is around 7 percent)that experience TOB with Maximum Limit deals. All thingsconsidered, we are unable to absolutely prove that acquirersextract private benefits. Yet, we can state that target stock behavioris consistent with the possibility that private benefits will beextracted.

5.3. Analysis of the causes of post-TOB stock price decline

In a well-functioning, efficient market, in the absence of newinformation, a zero abnormal return is expected. Conversely, ifthere is a systematic price fall around the expiration date of thetender offer, new information has probably been revealed thatnegatively affects the interests of the remaining minority share-holders. During the subscription period of a tender offer without a

24 The CARs of very large acquirers might not be sufficiently impacted byoverpayment for small targets to obtain an observable effect. In work not reportedhere, we find no evidence of overpayment in a sample constructed to match theapproximate size of targets with acquirers.

limit on the maximum number of shares to purchase, arbitragersmay trade using the tender offer information and, after theexpiration date, reverse their trades at prices based on the clean-upmerger conditions. Thus, the difference between the tender offerprice and the price based on information about the clean-upmerger should explain the post-TOB price decline. On the otherhand, in a TOB with a limit on the maximum number of shares to bepurchased, information is revealed at the expiration date of theTOB concerning the number of oversubscribed shares on a pro ratabasis. If a large number of shares is not accepted by the acquirer,holders of the unsold shares might then sell these shares in theopen market, perhaps even at lower prices than the tender offerprice. To verify this effect for tender offers with share purchaselimits, we analyze the Premium over Post-TOB Price measure fordeals separated into a prorationed sample and a non-prorationedsample. The results of this analysis are shown in Table 6.

Table 6 shows CARs for target firms near the days around theexpiration date of the tender offer for the sample of TOBs with amaximum limit on the number of shares to be purchased.Although, there is little difference between the AcquisitionPremium for the prorationing and non-prorationing samples,Target Announcement Returns are significantly lower for TOBswith prorationing. Yet, further examination shows that proration-ing is indeed an important cause of the stock price decline after theTOB. For TOBs in which an apportionment occurred, the Target TOBReturn (�5 of AD, +5 of ED) is a significant 13.9 percent less than forTOBs without prorationing. Because of this, the average SecurityBenefit of TOBs with prorationing is significantly smaller. TheSecurity Benefit for prorationed TOBs are positive on average, butstatistically significant at only the 10 percent level. Furthermore,the median value of the Security Benefit for prorationed TOBs is notstatistically significant. This is consistent with Hypothesis 5a. ForTOB with Maximum Limit deals in transactions with prorationing,the price conditions (or the freeze-out price) in the clean-upmergers also provide substantially less value compared to thetender offer price. This indicates that the fall in stock price causedby prorationing affects the clean-up merger conditions directly.25

In other words, in cases in which prorationing occurred, minorityshareholders (who are provided only a limited exit opportunity inthe tender offer) were forced to sell their shares under unfavorableconditions. This value-destroying effect is captured by the lowerprice of the target’s stock after the tender offer period ends. Finally,the Premium over Post-TOB Price is significantly larger forprorationed deals. This is consistent with Hypothesis 5b.

25 Among the twelve cases of TOB with maximum limit in which prorationingoccurred in our sample, the price conditions of the clean-up merger were disclosedbefore the expiration date of the tender offer in only one deal. Thus, we can say, atleast in this case, that the disadvantageous price conditions under the clean-upmerger did not cause the prorationings, but were rather caused by an actual stockprice drop associated with the prorationing.

Table 6Analysis of the effects for the TOB with a maximum limit sample with prorationing.

TOB without Prorationing TOB with Prorationing

N Mean/Median N Mean/Median Difference

Target Announcement Return (�3, +3) Mean 54 22.4%*** 43 13.5%*** 8.9%**Median 18.7%*** 11.4%*** 7.2%**

Target Expiration Return (�3, +3) Mean 54 �6.2%*** 43 �7.4%*** 1.3%Median �5.3%*** �4.5%*** �0.8%

Target TOB Return (�5 of AD, +5 of ED) Mean 54 19.2%*** 43 5.3%** 13.9%***Median 12.2%*** 4.8%** 7.3%**

Acquisition Premium Mean 54 22.9%*** 43 23.9%*** �1.0%Median 18.3%*** 18.7%*** �0.4%

Premium over Post-TOB Mean 54 5.8%* 43 19.9%*** �14.1%***Median 3.7%*** 12.1%*** �8.4%*

Security Benefit Mean 54 17.1%*** 43 4.0%* 13.0%***Median 11.2%*** 2.4% 8.8%***

Acquirer Announcement Return (�3,+3) Mean 41 �2.2% 39 0.5% �2.7%Median �1.2% �0.6% �0.6%

Acquirer Expiration Return (�3, +3) Mean 41 �0.5% 39 �0.5% 0.0%Median �0.5% �0.3% �0.2%

This table shows analysis of mergers and acquisitions in Japan over the period from 2000 to 2007. This particular table examines stock price when there is prorationing fortendered shares that exceed the maximum limit on shares to be acquired. The full descriptions of the abnormal returns and price relationships around tender offers aredescribed in Table 1. Mean shows the average value for the associated indicator. Median is the median value for the associated indicator. The columns with asterisks showresults for tests of the hypotheses that the mean or median are not importantly different from their expected zero value. The mean hypothesis test uses a t-test; the medianhypothesis test uses a Wilcoxon test. Target (Acquirer) Announcement (Expiration) Return is the cumulative abnormal return of target (acquirer) firms from 3 days before theannouncement (expiration) date of the TOB to 3 days after. Acquirer Announcement (Expiration) Return is calculated only when an acquirer is listed on a stock exchange inJapan. Acquisition Premium is the tender offer price divided by the target’s stock price 5 days before the announcement date of the TOB less one. Premium over post-TOB is thedifference between the tender offer price and the stock price of the target 5 days after the expiration date of the TOB, all divided by the target stock price 5 days before theannouncement date. When the Premium over Post-TOB Price is positive, it means that the stock prices of target firms have fallen from the tender offer price. Security Benefit isthe stock price 5 days after the expiration date divided by the stock price 5 days before the announcement date less one. Target TOB Return is the stock price 5 days before theannouncement date divided by the target stock price 5 days after the expiration date less one. Deals categorized as “TOB with Prorationing” are deals which are subject toprorationing because the number of tendered shares exceeded the limit. “TOB without Prorationing” are deals which are not subject to prorationing. ***, ** and * indicatestatistical significance at the 1, 5 and 10 percent levels of confidence.

M. Bremer et al. / Japan and the World Economy 41 (2017) 71–86 81

We also consider another approach to analyze deals in which TOBprorationing occurs. We conduct a binomial logit regressionanalysis with a rationing dummy variable that takes the value onewhen prorationing occurs, as the dependent variable. The mainexplanatory variables are the Acquisition Premium, the announcedmaximum number of shares to be purchased and a “Clean-upMerger Condition Disclosure Dummy” variable that takes the valueone if details about the clean-up merger were disclosed before theexpiration of the tender offer and zero otherwise. Other controlvariables are included in these regressions. The results of thisanalysis are shown in Table 7. The coefficient of the Clean-upMerger Condition Disclosure Dummy variable is negative andstatistically significant. This suggests that clean-up details are lesslikely to be revealed in prorationed deals. The coefficient ofMaximum TOB Share, (the maximum proportion of shares to bepurchased) is negative (�5.272) and statistically significant. Thissuggests that the smaller the maximum number of shares to beacquired in the tender, the greater the likelihood of the deal beingprorationed. This is consistent with Hypothesis 6. Of course, this isintuitive; and it confirms that these prorationed deals follow theinsight of Burkart et al. (1998). The minimum proportion requiredto control the target is the equilibrium level for the acquirer toextract private benefits. We also find a negative and significantcoefficient on the horizontal acquirer dummy variable. Weinterpret this to mean that, to the extent that horizontalacquisitions have potentially larger synergy effects througheconomies of scale and scope, acquiring firm shareholders areless likely to attempt to limit their post-deal ownership. Owning alarger share of the synergy-enhanced post-deal target will bebeneficial.26 In general, we observe that the probability of

26 This is consistent with research by Morck et al. (1990) and Bhagat et al. (2005)on U.S. companies who found that acquisitions of targets in related industries havelarger value creation.

prorationing is higher in cases where the clean-up mergerconditions are unclear during the tender offer. We interpret thisas evidence that the fiducial duty of either the target’s manage-ment or of controlling shareholders is not fully exercised.

Since Table 6 shows shareholder value increased in transactionswhere prorationing occurred, we argue that the monitoring andsynergy effects associated with the acquirer do benefit theinterests of the remaining minority shareholders. However,minority shareholders are at a disadvantage when the acquirer’sownership goal is the minimum necessary for control and whenclean-up merger conditions are unclear. The remaining minorityshareholders are left without a way to obtain adequate benefitsfrom the acquisition; and, especially under TOB with MaximumLimit format, they are forced to sell under disadvantageousconditions compared to the tender offer price. This situation meansthat the minority shareholders of target firms face the hold-upproblem. To solve this problem and avoid distorting the decisionsof minority shareholders, it is necessary to provide them with anexit from their investment. To summarize, we show thatprorationing in a TOB is one of the causes of the post-TOB stockprice decrease.

5.4. Analysis of legal revision effects

The recent introduction of the new Corporate Law simplifiesfreeze-outs of minority shareholders by introducing callable class-shares and the Financial Instruments and Exchange Law requiresmandatory bids for all shares outstanding in TOBs where the ratioof ownership of shares after the acquisition is anticipated to betwo-thirds or more. This series of legal revisions gives all minorityshareholders a chance to exit takeovers that seek to control morethan two-thirds of the voting rights of the firm. Although theFinancial Instruments and Exchange Law took effect fromSeptember 2007, it was enacted in June 2006. That was the time

Table 7Logit regression of the occurrence of prorationing.

Sample TOB with a maximum limit on the number of shares to purchase

Dependent Variable Prorationing = 1, Others = 0

Model (1) (2)N 97 97Cox & Snell R2 0.177 0.253Nagelkerke R2 0.238 0.338

Coefficient Wald Significance Coefficient Wald Significance

Constant �2.202 1.739 0.187 �0.289 0.015 0.902Clean-up Merger Disclosure Dummy �1.814 6.102 0.014** �1.465 3.857 0.050**Acquirer Toehold 0.545 0.400 0.527 �1.664 1.102 0.294Target Size 0.232 1.909 0.167 0.215 1.491 0.222

Maximum TOB Share �5.272 3.893 0.048**Post-Corporate Law Dummy 0.798 1.934 0.164Public Acquirer Dummy 0.634 0.818 0.366 0.964 1.244 0.265Horizontal Dummy �0.984 4.273 0.039** �1.094 4.619 0.032**Equity Fund Dummy �2.099 2.971 0.085* �2.160 2.439 0.118Acquisition Premium 1.451 1.408 0.235

This table shows results of logistic regression tests of the probability of occurence of prorationing in TOB with maximum limit on the number of shares to be purchased. Clean-up Merger Disclosure Dummy is dummy variable which takes one when the acquirer or the target discloses the price conditions of the clean-up merger before the expirationdate of the TOB and zero otherwise. Acquirer Toehold is the acquirer’s stock holding in the target before the takeover offer. Target Size is the log of the market capitalization ofthe target. Maximum TOB Share is the maximum number of shares as a proportion to be purchased that is announced in the TOB. Post-Corporate Law Dummy is a dummyvariable that takes the value one when the TOB is announced after the effective date of New Corporate Law (May of 2006) and zero otherwise. Public Acquirer Dummy is adummy variable that takes the value one when the acquirer is a listed firm on a stock exchange in Japan and zero otherwise. Horizontal Dummy is a dummy variable that takesthe value one when the acquirer and the target are in the same industry and zero otherwise. Equity Fund Dummy is dummy variable that takes the value one when theacquirer is a private equity fund and zero otherwise. Acquisition Premium is the tender price divided by the stock price 5 days before the TOB less one. ** and * indicatestatistical significance at the 5 and 10 percent levels of confidence.

Table 8TOB deal category.

All Prior to Corporate Law Introduction Post Corporate Law Introduction

Non-Discount TOB without Maximum Limit 76 51 25Freeze-out Disclosure 52 68.4% 27 52.9% 25 100.0%

Cash Freeze-out 33 43.4% 15 29.4% 18 72.0%Other 24 31.6% 24 47.1% 0 0.0%

Freeze-out in Two Years without Disclosure 3 3.9% 3 5.9% 0 0.0%Non-Discount TOB with Maximum Limit 97 71 26

Freeze-out Disclosure 16 16.5% 11 15.5% 5 19.2%Cash Freeze-out 0 0.0% 0 0.0% 0 0.0%

Other 81 83.5% 60 84.5% 21 80.8%Freeze-out in Two Years without Disclosure 21 21.6% 13 18.3% 6 23.1%

This table shows analysis of mergers and acquisitions in Japan over the period from 2000 to 2007. This particular table shows the proportions of deals where the acquirerdisclosed his intention to freeze-out the remaining shareholders after a successful TOB in the context of changes in Japan’s laws. The sub-samples labeled “Freeze-outDisclosure” show the proportion of transactions in which the acquirer disclosed his intention to freeze-out the remaining shareholders after a successful TOB. “Freeze-out inTwo Years without Disclosure” shows the proportion of transactions in which the freeze-out occurs within two years after the completion of the TOB without the disclosure ofthe freeze-out plan at the beginning of the TOB.

82 M. Bremer et al. / Japan and the World Economy 41 (2017) 71–86

when practitioners started to adjust to the new legal requirementsto avoid potential disputes. In addition, the Tokyo Stock Exchangeformally asked listed acquirers to disclose their plans to deal withminority shareholders by providing information on such things asfreeze-outs and clean-up mergers, for the first time in August 2006(see footnote 15). To test the influence of the series of legalrevisions as a whole, we divide the sample by separating dealsannounced before and after May 2006, when the new CorporateLaw was promulgated. Table 8 summarizes the TOB data across theperiods before and after the new corporate law and in terms offreeze-outs. The law does, indeed, have an impact. TOB withoutMaximum Limit deals are implicitly two-thirds or greater bids, andthus subject to the law. The table shows that all TOB withoutMaximum Limit deals announced freeze-out information (25disclosures for 25 deals) in the period after the law becameeffective. Yet, for TOB with Maximum Limit, only 5 of 26 dealsdisclosed freeze-out information. The question is whether thingschanged for TOBs with Maximum Limit deals after the new law.

Table 9 shows how stock prices change around the time thattakeovers occur; the sample splits TOBs into those that occurredbefore and those that occurred after the change in the corporatelaw. There is a notable difference between the results for TOB witha Maximum Limit deals and TOB without a Maximum Limit deals.TOB without a Maximum Limit deals are effectively subject to thetwo-thirds legal requirement that acquirers provide a full-information exit for minority shareholders. The Target TOB Returnis significantly larger at 21.1 percent after the law change (as shownin Panel B of Table 9). Similarly, the Security Benefit is larger, at 24.1percent, after the law change; yet, this value is not significantlydifferent (t-statistic at 1.43) from before the change in the law.Target Expiration Returns at the end of the takeover period declinesignificantly less, at �1.6 percent after the law change. ThePremium over Post-TOB Price is only 1.8 percent after the lawchange, which is significantly smaller than before. This is evidencethat the change in the corporate law changed takeover behavior.Minority shareholders face less uncertainty about freeze-outs.

Table 9Effects of revisions in Japanese takeover laws from May 2006.

Panel A TOB with Maximum Limit

Before Corporate Law Post Corporate Law

N Mean/Median N Mean/Median Difference

Target Announcement Return (�3, +3) 71 Mean 17.7%*** 26 20.5%*** 2.8%Median 14.4%*** 16.2%*** 1.7%

Target Expiration Return (�3, +3) 71 Mean �6.5%*** 26 �7.3%*** �0.8%Median �4.2%*** �5.8%*** �1.5%

Target TOB Return (�5 of AD, +5 of ED) 71 Mean 14.3%*** 26 9.7%* �4.6%Median 9.1%*** 5.2%*** �3.9%

Acquisition Premium 71 Mean 23.5%*** 26 22.6%*** �0.9%Median 18.2%*** 20.8%*** 2.6%

Premium over Post-TOB 71 Mean 10.9%*** 26 15.2%*** 4.4%Median 4.9%*** 12.7%*** 7.7%

Security Benefit 71 Mean 12.7%*** 26 7.4% �5.3%Median 8.7%*** 3.0%*** �5.7%

Acquirer Announcement Return (�3,+3) 56 Mean �2.4% 24 2.5%** 4.9%Median �1.6%** 2.5% 4.1%

Acquirer Expiration Return (�3, +3) 56 Mean �0.5% 24 �0.5% �0.1%Median �0.2% �1.0% �0.8%

Panel B TOB without maximum Limit.

Before Corporate Law Post Corporate Law

N Mean/Median N Mean/Median Difference

Target Announcement Return (�3, +3) 51 Mean 25.5%*** 25 22.5%*** �3.0%Median 24.4%*** 18.1%*** �6.2%

Target Expiration Return (�3, +3) 51 Mean �9.8%*** 25 �1.6%*** 8.2%***Median �6.2%*** �1.7%*** 4.5%***

Target TOB Return (�5 of AD, +5 of ED) 51 Mean 10.1%*** 25 21.1%*** 11.0%**Median 8.4%* 14.9%*** 6.5%*

Acquisition Premium 51 Mean 28.9%*** 25 25.8%*** �3.1%Median 28.0%*** 16.9%*** �11.1%

Premium over Post-TOB 51 Mean 12.0%*** 25 1.8%*** �10.3%***Median 6.2%*** 1.7%*** �4.5%***

Security Benefit 51 Mean 16.9%*** 25 24.1%*** 7.2%Median 17.5% 16.3%*** �1.1%

Acquirer Announcement Return (�3,+3) 13 Mean 1.1% 8 2.2% 1.1%Median 0.6%** 3.4% 2.8%

Acquirer Expiration Return (�3, +3) 13 Mean 1.7% 8 �2.0% �3.7%Median 1.2% �1.3% �2.5%

This table shows analysis of mergers and acquisitions in Japan over the period from 2000 to 2007. This particular table shows results of tests of whether the revision of theCorporate Law as of May 2006 improved the potential coercive nature of tender offers in Japan. Deals categorized as “TOB with (without) Maximum Limit” are deals in whichmaximum number of shares purchased in the TOB is specified (is not specified). Deals categorized as “Before (Post) Corporate Law” are the subsample of tender offersannounced before (after) May 2006. Mean shows the average value for the associated indicator. Median is the median value for the associated indicator. The columns withasterisks show results for tests of the hypotheses that the mean or median are not importantly different from their expected zero value. The mean hypothesis test uses a t-test;the median hypothesis test uses a Wilcoxon test. Full descriptions of the indicator values are reported in Table 1. Target (Acquirer) Announcement (Expiration) Return is thecumulative abnormal return of target (acquirer) firms from 3 days before the announcement (expiration) date of the TOB to 3 days after. Target TOB Return is the cumulativeabnormal return of the target from 5 days before the annoucement date of the TOB to 5 days after then expiration date of the TOB. Acquirer Announcement (Expiration) Returnis calculated only when an acquirer is listed on a stock exchange in Japan. Acquisition Premium is the tender offer price divided by the target’s stock price 5 days before theannouncement date of the TOB less one. Premium over Post-TOB is the difference between the tender offer price and the stock price of the target 5 days after the expirationdate all divided by the stock price of the target 5 days before the annoucement date. When the Premium over Post-TOB Price is positive, it means that the stock prices of targetfirms have fallen from the tender offer price. Security Benefit is the stock price 5 days after the expiration date divided by the stock price 5 days before the announcement dateof the TOB less one. ***, ** and * indicate statistical significance at the 1, 5 and 10 percent levels of confidence.

M. Bremer et al. / Japan and the World Economy 41 (2017) 71–86 83

Concerning the acquirer, the Premium over Post-TOB Price for TOBwithout Maximum Limit deals falls significantly after the lawchange. Yet the acquirers’ announcement returns are (notsignificantly) positive.

TOB with Maximum Limit deals are less likely to be subject tothe two-thirds legal requirement of disclosure of clean-up details.Hence, we expect that coercive takeover behavior will not changeafter the introduction of the new law. Indeed, this is the case. ForTOB with Maximum Limit deals, the Target Announcement Return(17.7 percent before versus 20.5 percent after, as shown in Panel Aof Table 8), Target Expiration Return (�6.5 percent before versus�7.3 percent after), Premium over Post-TOB Price (10.9 percentbefore versus 15.2 percent after), Acquisition Premium (23.5percent before versus 22.6 percent after) and Security Benefit (12.7percent before versus 7.4 percent after) are not significantly

different. These values are clearly less advantageous for minorityshareholders. Further, the Acquirer Announcement Return doesincrease significantly from �2.4 percent to 2.5 percent.

We conclude that TOBs where the target control ratio is overtwo-thirds are directly affected by the recent legal revisions andthe effect on minority shareholder protection is positive. However,for TOBs where the target control ratio is less than two-thirds (andthus not directly affected by the legal revisions), minorityshareholders are not protected, even after the legal revision. Inaddition, since the proportion of takeovers with the goal of morethan two-thirds of the target’s shares decreased in the period afterthe legal amendments, we suspect that some of the acquirersreduced their target control ratio to less than two-thirdsspecifically to avoid mandatory any-or-all bids as required bythe new law.

Table 10Expiration returns.

Model (1) (2) (3) (4) (5)

Dependent Variable Target Expiration Return (-3, +3)

Adjusted R2 0.194 0.210 0.230 0.243 0.156

Coefficient t-stat Coefficient t-stat Coefficient t-stat Coefficient t-stat Coefficient t-stat

Constant �0.167 �2.604** �0.135 �2.064** �0.105 �1.640 �0.124 �1.939* �0.218 �3.412***Any-or-all Bids Dummy �0.013 �0.656 �0.032 �1.519 0.019 0.767 0.007 0.272 �0.005 �0.200Any-or-all bids Dummy � No Clean-upMerger Disclosure Dummy

�0.085 �2.789*** �0.086 �2.838*** �0.101 �3.207***

Prorationing Ratio �0.333 �2.175** �0.284 �1.849* �0.289 �1.927* �0.450 �2.926***Corporate Law dummy 0.032 1.695* �0.005 �0.188 �0.014 �0.583 �0.005 �0.215 0.004 0.144Any-or-all Bids Dummy � Corporate LawDummy

0.078 2.088** 0.051 1.322 0.040 1.030 0.026 0.630

Acquirer Toehold 0.012 0.419 0.016 0.577 0.018 0.666 0.017 0.619 0.032 1.101Target Size 0.013 2.052** 0.011 1.611 0.006 1.007 0.009 1.420 0.015 2.233**Horizontal Dummy 0.035 2.014** 0.039 2.278** 0.043 2.531** 0.042 2.495** 0.033 1.847*Equity Fund Dummy 0.015 0.796 0.011 0.603 0.005 0.273 0.006 0.312 0.005 0.250Acquisition Premium �0.206 �4.567*** �0.210 �4.703*** �0.216 �5.009*** �0.195 �4.435***

This table shows analysis of mergers and acquisitions in Japan over the period from 2000 to 2007. This particular table shows results of OLS regressions of seven-daycumulative abnormal returns around the expiration date of the TOB. These CARs are calculated over the period from 3 days before the expiration date (ED) to 3 days after. Wetest if lack of disclosure of the conditions of the clean-up merger (No Clean-up Merger Disclosure Dummy) at the announcement of tender offers and the corporate lawamendment of May 2006 (Corporate Law Dummy) improve shareholder returns over the seven days around the expiration date of the tender offers. The Any-or-all BidsDummy is a dummy variable that takes the value one when the TOB is categorized as an Any-or-all Bid and zero otherwise. The No Clean-up Merger Disclosure Dummy is adummy variable that takes the value one if the acquirer or the target do not disclose the price conditions of the clean-up merger before the expiration date of the TOB and zerootherwise. Prorationing Ratio is the proportion of shares not purchased due to oversubscription by the tendered shares. The Corporate Law Dummy is a dummy variable thattakes the value one when the TOB is announced after the effective date of the New Corporate Law (May of 2006) and zero otherwise. Acquirer Toehold is the proportion of thetarget’s shares held by the acquirer before the TOB. Target Size is the log of the market capitalization of the target. The Horizontal Dummy is a dummy variable that takes thevalue one when the acquirer and the target belong to the same industry. The Equity Fund Dummy is a dummy variable that takes the value one when the acquirer is privateequity fund and zero otherwise. Acquisition Premium is the tender offer price divided by the target’s stock price 5 days before the announcement date of the TOB less one. ***,** and * indicate statistical significance at the 1, 5 and 10 percent levels of confidence. The number of observations is 173.

84 M. Bremer et al. / Japan and the World Economy 41 (2017) 71–86

5.5. Regression analysis

In previous sections, we showed that prorationing under partialTOBs and insufficient legal protection of minority shareholders aremajor reasons for the post-TOB stock price drop. To confirm ourresults while controlling for other factors, we further analyze thepotential causes of the stock price drop associated with theexpiration of the tender offer. Consider that the relationshipbetween the acquirer and the target may influence whether theacquirer structures the TOB as an any-or-all deal, as a two-tier dealand indeed how large the proportion of the target is sought in atwo-tier deal while at the same time this relationship may impactthe synergy value created by the takeover as well as the privatebenefits potentially extracted. For example, we saw in the analysisin Table 7 that horizontal acquisitions are significantly less likely toexperience prorationing. It seems reasonable that partial takeoversby acquirers in the same industry that potentially create greatersynergy value may influence part of the post-TOB stock price drop.These endogeneity concerns can be addressed (at least in part)with ordinary least squares regressions. The dependent variable isthe target’s return at the expiration of the tender offer period.27 Wepredict both TOBs with a maximum limit on the number of sharessought and the opacity of the handling of remaining minorityshareholders are responsible for the stock price drop after the TOBin addition to other aspects of the takeover such as the relationbetween the target and the acquirer. As direct indicators ofopportunistic behavior by acquirers, we add: an “Any-or-all Bids”dummy variable that takes the value one if the transaction iscategorized as a TOB without Maximum Limit deal and zerootherwise; a “Prorationing Ratio” variable that is the number oftendered shares rejected by the acquirer as a proportion of thenumber of outstanding shares of the target firm; and, a “No Clean-

27 In work not reported here, we conduct similar analysis using the Premium overPost-TOB Price and Securities Benefit variables. The results are essentially the same.

up Merger Disclosure” dummy variable that takes the value one ifthe price conditions of the clean-up merger were not disclosedprior to the expiration date of the tender offer and zero otherwise.In addition, we add a “Corporate Law” dummy variable that takesthe value one if the takeover is announced after May 2006, themonth when the new Japanese Corporate Law took effect, to test ifminority shareholders are better protected under the new law. The“Horizontal Dummy” variable captures the relationship betweenthe target and the acquirer; it is set to one if the firms are in thesame or related industries and zero otherwise. The “Fund Acquirer”variable is a dummy variable that takes the value one if the acquireris a private equity fund and zero otherwise. We also include thesize of the target firm as a proxy of the negotiation power of targetboards and shareholders against the acquirer. The results of thisanalysis are shown in Table 10.

Model 1 analyzes whether any-or-all bids alone reduce theconcerns of the remaining shareholders of target firms at theexpiration of the TOB. Model 2 further analyzes whether theamendment of the corporate law and other TOB rules have adifferent effect on any-or-all and other bids. Model 3 analyzes if thelack of disclosure of the price conditions of the clean-up merger(“No Clean-up Merger Disclosure dummy”) increase the concernsof target shareholders. Model 4 tests whether the results of Model3 continue to hold in the presence of prorationing. Finally, Model 5tests whether the results of Model 4 continue to hold withoutcontrolling for the Acquisition Premium.

First, note that the negative and significant coefficients on the“Prorationing Ratio” variable indicate that the higher the propor-tion of prorationed shares rejected by the acquirer, the larger thepost-TOB stock price drop. This suggests that the differencebetween TOB with Maximum Limit and TOB without MaximumLimit deals stems from the situation where minority shareholderswish to sell their shares, yet are not given a fair exit from theirinvestment. Second, in all the models, the coefficients on the Any-or-all Bids dummy variable are not significant. Third, the

28 See Subramanian (2007) for an example that describes how biased regulationchanges acquirer behavior.29 This suggestion is consistent with the recommendations of Osugi (2010)concerning the current debate on revisions to the Corporate Law and the FinancialInstruments and Exchange Act.

M. Bremer et al. / Japan and the World Economy 41 (2017) 71–86 85

introduction of legal protections under both the new CorporateLaw and the Financial Instruments and Exchange Law did makeminority shareholders better off in any-or-all bids. This is shown bythe positive coefficient for the Corporate Law dummy variable inModel 1 (0.032, and statistically significant at the 10 percent level).On the other hand, in Model 2 where we include an interactionvariable for the “Any-or-all Bids dummy” and the “Corporate Lawdummy,” the coefficient for the Corporate Law dummy variable isinsignificant (�0.005 with a t-value of �0.188). The interactionvariable is positive and significant at the 5 percent level. Thisshows that shareholders were effectively protected by the law inTOBs that were any-or-all format, but not protected by the law inTOBs that were partial or two-tier format.

In Model 3, the coefficient on the “Any-or-all bids dummy � NoClean-up Merger Disclosure dummy” interaction variable isnegative and significant. These results are consistent withHypothesis 6 that anticipation of opportunistic behavior byacquirers after the TOB due to the lack of clear information aboutthe clean-up merger lowers the value of minority shareholders’interests after the takeover. The results for Model 3 hold even whenwe include the prorationing ratio as an explanatory variable. Apredictable exit option for minority shareholders in a TOB isessential to avoid coercion. These results do not change even whenwe exclude the Acquisition Premium as an independent variable inModel 5. In addition, if the size of the target firm is fairly large,suggesting that the negotiating power of minority shareholders issignificant, then expropriation of minority shareholder interests bythe acquirer is mitigated to some degree. These results areconsistent with Hypothesis 3 that the post-TOB stock price drop isbased on market fears of expropriation of minority shareholderinterests by the acquirer. In addition, in all the models, thehorizontal dummy is positive and significant. The results areconsistent with our interpretation relating to the results of Table 7.In total, the regression results are consistent Hypothesis 3,Hypothesis 5a, Hypothesis 5b and Hypothesis 6.

We deduce that, when the proportion of shares rejected by anacquirer is large and the clean-up merger conditions are at adisadvantage for minority shareholders compared to the tenderoffer price, a significant drop in post-TOB stock price occurs. Both ofthe above tend to occur when the post-TOB percentage of votingrights controlled by acquirer is not high. Consequently, if measuresto increase the proportion of post-TOB shares held by the acquirerare introduced, the acquirer’s private benefits will be internalizedand the expropriation of minority shareholder interests will besmaller.

6. Suggestions for legal system changes based on results of theanalysis

This research finds evidence of coercive takeovers; this coercionis related to legal rules. Because minority shareholder protection islimited for certain types of tenders in Japan, acquirers can engagein opportunistic behavior and minority shareholders are coerced tosell their shares. This may be the main reason that almost alltakeovers are successful, excluding hostile takeovers. Minorityshareholders are not able to cooperate with one another, nor tonegotiate complete contracts with acquirers. Hence, minorityshareholders face a hold-up problem due to the ex post power ofacquirers. This suggests the need for changes to protect minorityshareholders. Specifically, it should be the legal duty of a bidder todisclose in detail all conditions concerning the clean-up merger, ifsuch is planned within a certain period from the completion of thetakeover. It is also desirable to set prices for clean-up mergers thatare not lower than tender offer prices. Simultaneously, in order tosecure a fairly valued exit for minority shareholders, the acquirershould have to make a mandatory bid for all outstanding shares in

takeovers with the goal of securing at least 51 percent of thetarget’s voting rights. For example, an obligatory extension of thetender offer period in cases where the acquirer successfullypurchases his targeted portion of shares, as was introduced in theU.K., is a practical way to achieve the goal.

Still, there is the possibility that, in strengthening the protectionof minority shareholders, acquisition costs borne by acquirerswould increase to the point of hindering value increasingacquisitions. This was one of Burkart’s (1999) key points. Onesolution to this problem is to impede the free-riding behavior ofminority shareholders by setting clear post-takeover freeze-outrules and to keep acquirer costs at a reasonable level in any-or-allbids. This balances two important goals: minority shareholderprotection and efficiency in the market for corporate control.However, these proposed takeover regulations may not besufficient. Gilson and Gordon (2003) point out the importanceof making the legal system’s policies on the private benefits of largeshareholders consistent on the whole. Merely mandating clean-upmergers for a limited period after the takeovers may result inclean-up mergers simply happening later, after the regulatedperiod, or through stock swaps rather than takeover bids.28 Thismeans any country that hopes to improve its market for corporatecontrol by avoiding coercive tender offers should also considerintroducing an entire fairness rule such as that used in the UnitedStates for all mergers and stock-for-stock acquisitions by control-ling shareholders.29

Acknowledgments

The authors thank James Bremer, Joan Collins-Lund, EdmundSkrzypczak, Wataru Tanaka, Yasuhiko Tanigawa, Eriko Yamamoto,Tsung-ming Yeh and an anonymous referee. This work wassupported by JSPS KAKENHI Grant No. 15H03375. The authorshave no relevant or material financial interests that relate to theresearch described in this paper.

References

Baker, M., Pan, X., Wurgler, J., 2012. The effect of reference point prices on mergersand acquisitions. J. Financial Econ. 106 (1), 49–71.

Barclay, M.J., Holderness, C.G., 1989. Private benefits from control of publiccorporations. J. Financial Econ. 25 (2), 371–395.

Barclay, M.J., Holderness, C.G., 1992. The law and large-block trades. J. Law Econ. 35,265–294.

Becht, M., Bolton, P., Roell, A., 2007. Corporate law and governance. Handbook of Lawand Economics, Vol. 2. North-Holland, pp. 829–943 (Ch. 12).

Bhagat, S., Dong, M., Hirshleifer, D., Noah, R., 2005. Do tender offers create value?New methods and evidence. J. Financial Econ. 76 (1), 61–98.

Burkart, M., Gromb, D., Panunzi, F., 1998. Why higher takeover premia protectminority shareholders. J. Polit. Econ. 106 (1), 172–204.

Burkart, M., 1999. Economics of Takeover Regulation Stockholm School ofEconomics Working Paper (December). .

Chernenko, S., Foley, F., Greenwood, R., 2012. Agency costs, mispricing andownership structure. Financial Manage. 41 (4), 885–914.

Comment, R., Jarrell, G.A., 1987. Two-tier and negotiated tender offers: theimprisonment of the free-riding shareholder. J. Financial Econ. 19 (2), 283–310.

Djankov, R., La Porta, R., Lopez-del-Silanes, F., Shliefer, A., 2008. The law andeconomics of self-dealing. J. Financial Econ. 88 (3), 430–465.

Dyck, A., Zingales, L., 2004. Private benefits of control: an international comparison.J. Finance 59 (2), 537–600.

Gilson, R.J., Gordon, J.N., 2003. Controlling controlling shareholders. Univ. Pa. LawRev. 152 (2), 785–843.

Grossman, S.J., Hart, O.D., 1980. Takeover bids, the free-rider problem, and thetheory of the corporation. Bell J. Econ. 11 (1), 42–64.

Inagaki, K., 2013. Minority investors cry foul in Japan. Wall Street J. (January 29).

86 M. Bremer et al. / Japan and the World Economy 41 (2017) 71–86

Inoue, K., Ikeda, N., 2016. A cross-country analysis of shareholder protection in thegoing private transactions. JSDACapital Market Forum Research Report (inJapanese). Japan Securities Dealers Association, Tokyo, Japan.

Ito, Y., 2008. The protection of minority shareholders of subsidiaries. Commer. LawRev. (in Japanese) 1841, 26–34.

La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R., 2002. Investor protectionand corporate valuation. J. Finance 57 (3), 1147–1170.

Morck, R., Schleifer, A., Vishny, R.W., 1990. Do managerial objectives drive badacquisitions? J. Finance 45 (1), 31–48.

Lewis, L., Inagaki, K., Japan pension fund pressed over controversial deal, Panasonictakeover’s pricing sparks call for investor GPIF to take a stand on governance,Financial Times, 2017 (January 22).

Nihon Keizai Shimbun, 2006a. Hanshin Share Price Becomes Lower than TOB Pricefor the First Time (in Japanese) (June 16). .

Nihon Keizai Shimbun, 2006b. Hankyu and Hanshin to Integrate by October (inJapanese) (June 20). .

Nikkei Business Watch, (2008) NEC [NEC’sstake in a separately listed subsidiary], Downloaded on March 8, 2016 from theWorld Wide Web: http://business. nikkeibp.co.jp/article/money/20080612/161812/(in Japanese).

Nikkei Shimbun, 2007. M&A and Individual Investors: Little Choice Left Even whenThey Are Not Satisfied (November 20). .

Osugi, K., 2010. An urgent task is to rectify ‘disregard of shareholders’ rights’. NikkeiShimbun (April 14).

Ozawa, N., Ikeda, N., Inoue, K., 2016. The effect of reference points of individualinvestors on tender offer prices. Japan J. Finance (forthcoming).

Roll, R., 1986. The hubris theory of corporate takeovers. J. Bus. 59 (2), 197–216.Shleifer, A., Vishny, R.W., 1997. A survey of corporate governance. J. Finance 52 (2),

737–783.Sinclair Oil Corporation v. Levien (1971) 280 A.2d 717 (Del. 1971).Spamman, H., 2010. The antidirector rights index revisited. Rev. Financial Stud. 23

(2), 467–486.Subramanian, G., 2007. Post-Siliconix freeze-outs: theory and evidence. J. Legal

Stud. 36 (1), 1–26.Tokyo Stock Exchange, (2006) Directive No. 1338 (August 28).

Further reading

Siliconix, Inc. S’holders Litig. (2001) No. 18700, 2001 Del. Ch. LEXIS 83 (DEL. Ch. June19, 2001).