Method of Payment and Target Status: Announcement Returns to Acquiring Firms in the Malaysian Market

Transcription

1 Method of Payment and Target Status: Announcement Returns to Acqurng Frms n the Malaysan Market Mansor Isa Faculty of Busness and Accountancy, Unversty of Malaya Lembah Panta, Kuala Lumpur, Malaysa Tel: E-mal: mansor@um.edu.my Sew-Peng Lee (Correspondng author) Faculty of Accountancy and Management Unversty of Tunku Abdul Rahman, Lot PT 21144, Jalan Sunga Long Bandar Sunga Long, Kajang, Malaysa Tel: E-mal: leesp@utar.edu.my Receved: November 10, 2010 Accepted: December 21, 2010 do: /jef.v3n3p177 Abstract Ths paper examnes the mpact of acquston announcements on the acqurer s returns n the Malaysan market for the perod from 2000 to Usng the market model wth nfrequent tradng adjustments, our results reveal that acqurng frms obtan postve abnormal returns on the announcement day. When acqustons are analyzed based on target frm status, we fnd that publc acqustons provde hgher abnormal returns to the acqurng frms than prvate acqustons. In terms of method of payment, the evdence suggests that cash acqustons generate postve abnormal returns for the acqurng frms regardless of the target frm status. Fnally, the sze of the acqured asset value relatve to the acqurng frms s postvely assocated wth the acqurng frm s abnormal returns. Keywords: Acquston, Acqurer s return, Method of payment, Target status 1. Introducton Most studes on mergers and acqustons seem to have agreed that target frms receve a sgnfcantly postve abnormal return around the announcement perod, whle acqurng frms reveal mxed fndngs. Certan studes reported that acqurng frms experence postve abnormal returns, whle others found negatve abnormal returns and some even found non-sgnfcant abnormal returns. In a comprehensve paper, Jensen and Ruback (1983) revew 13 emprcal studes on mergers and acqustons that analyze abnormal returns around the mergers and takeover announcements. Ther general concluson s that target frms gan whle acqurng frms do not lose. One mportant element n corporate acquston s the method of payment used to settle the transacton. The most commonly used methods of payment are full cash, full share-exchange or a combnaton of both. Prevous studes fnd that the methods of payment have an mpact on the announcement returns. Fuller, Netter and Stegemoller (2002) support the early fndngs that acqurng frms wth cash offers show hgher returns than those wth share offers. However, the evdence on the mpact of share offer acqustons s nconclusve, some studes (for example, Moeller, Schlngemann and Stulz (2004) and Ismal (2008)) fnd sgnfcant gans whle others (for example, Andrade, Mtchell and Stafford (2001) and Fuller, Netter and Stegemoller (2002)) fnd sgnfcant loss. Another ssue relatng to acquston announcement return s the status of the target, that s, whether the targets are prvate and unlsted frms or publc lsted frms. There s very lttle research done on the returns of acqurng frms n relaton to target status. Two of the latest studes on ths topc were those of Fuller, Netter and Stegemoller (2002) for the U.S. market, and Draper and Paudyal (2006) for the U.K. market; both studes fnd that acqurers gan when acqurng a prvate target and lose when acqurng a publc target. We note that most of the prevous studes on acquston were conducted n the developed markets, partcularly the U.S. and U.K. markets, and very lttle research has been done on developng markets. Hence, t begs the queston of whether ther fndngs are equally relevant to a developng market. In ths study, we focus on the returns to the acqurng frms durng acquston announcement n the Malaysan stock market for the perod 2000 to We frst examne the market reacton to acqurng frm returns n acquston around the announcement date. We then Publshed by Canadan Center of Scence and Educaton 177

2 Internatonal Journal of Economcs and Fnance Vol. 3, No.3; August 2011 analyze the market reacton n relaton to target status and methods of payment. We hope that by studyng ths event n a developng country, such as Malaysa, we would be able to assess whether related fnance theores that have been found relevant n developed markets are equally applcable n a developng market. The rest of the paper s organzed as follows. The next secton revews the relevant lterature, and Secton 3 descrbes the data and methodology used n the study. Secton 4 dscusses the results and Secton 5 presents the concluson. 2. Lterature Revew 2.1. Performance of Acqurng Frms Emprcal evdence n the U.S. largely supports that acqustons create value for shareholders, although some evdence to the contrary exsts. Value creaton evdence s qute clear for target frm shareholders, but less certan for acqurng frm shareholders. Some U.S. studes report that acqurng frms earn a statstcally sgnfcant postve abnormal return. For example, Asquth, Burner and Mullns (1983) and Asquth and Km (1982) fnd that acqurng frms earn sgnfcant postve abnormal returns of 0.2% to 1% on the announcement day. Later studes, such as Fuller, Netter and Stegemoller (2002), Moeller, Schlngemann and Stulz (2004), Draper and Paudyal (2006) and Ismal (2008) also fnd smlar evdence. In contrast, there are studes that report negatve returns for the acqurng frms, for example, Amhud, Lev and Travlos (1990), Healy, Palepu and Ruback (1992), and Moeller et al. (2005). Furthermore, there are studes, for example, Bradley, Desa and Km (1983), Draper and Paudyal (1999), Facco, McConnell and Stoln (2006) that fnd that abnormal returns to acqurng frms are not sgnfcantly dfferent from zero. In developng markets, Koh and Lee (1988), and Cheung and Shum (1993) fnd no sgnfcant abnormal return to the acqurng frms n the Sngapore and Hong Kong markets, respectvely. In fact, Cheung and Sum (1993) document that acqurng frms suffer a sgnfcant % abnormal returns over the 61-day event wndow n ther study. In Malaysa, Isa and Lm (1993), Mat-Nor (1993) and Isa (1994) report that around the announcement days both the acqurng and target frms earn postve abnormal returns Method of Payment Prevous studes (e.g., Travlos and Papaoannou, 1991; Draper and Paudyal, 1999; and Facco and Masuls, 2005) ndcate that dfferent methods of payment n acquston yeld dfferent mpacts on acqurng frm s share prces around the announcement perod. There seems to be a general consensus among the emprcal studes that the acqurng frms have hgher excess returns n cash acqustons compared to n share-exchange acqustons. For acqurng frms, Martynova and Renneboog (2008) and Ismal (2008) fnd that all-cash acqustons generate hgher returns to acqurers than share-exchange acqustons. The explanaton s that the announcement of a cash acquston sgnals to the market that acqurng managers beleve that ther frm s shares are underprced. However, the evdence on the effect of share-exchange acqustons on acqurng frms s mxed; Travlos (1987) reports sgnfcantly negatve abnormal returns of 1.47% over the perod (day -1 to 0), whle Wansley, Lane and Yang (1987) report that acqurers earn nsgnfcant negatve abnormal returns. Moeller, Schlngemann and Stulz (2004) report sgnfcant postve abnormal returns over the three-day announcement perod (day -1 to day 1) n both cash and share acqustons. In the Malaysan market, Isa (1994) reports that acqurng frms n cash acqustons earn a postve return of 0.12% and share acqustons earn a negatve return of 0.65% over the perod of day -1 to day 0. The sgnalng model of Leland and Pyle (1977) and Myers and Majluf (1984) suggests that due to nformaton asymmetry, managers prefer cash offers f they beleve that the acqurng frm s share s undervalued. As such, the method of payment acts as a sgnalng devce about the acqurng frm s stock value; nvestors nterpret cash offers as good news and share offers as bad news. Consequently, proposals of cash offers n acquston are expected to have a postve mpact, whereas the share offers are expected to have a negatve mpact on the acqurng frm s share prce. Ths suggests that cash-offer acqustons should gan more than share-offer acqustons Target Frm Status In an acquston, a lsted frm may acqure another publc lsted company or a prvate and unlsted company. There are several basc dfferences n the valuaton and prcng of a prvate vs-à-vs a publc frm. Normally, prvate frms are closely held by a lmted number of shareholders; often tmes they are famly owned and contan a hgh proporton of owner-managers. Ther shares are not publcly traded and; therefore, valuaton can be dffcult. Consequently, acqurng a prvate company may be subject to a dfferent expectaton from the target shareholders. The works of Chang (1998), Ang and Kohers (2001), Fuller et al. (2002), and Draper and Paudyal (2006) report that acqurers of prvate targets experence postve returns whle acqurers of publc targets suffer losses n acqustons. We dscuss below several hypotheses offered by Chang (1998), and Draper and Paudyal (2006) to explan the reasons for the dfferences n returns to acqurers n relaton to the status of the targets. 178 ISSN X E-ISSN

3 The manageral motve hypothess The manageral hypothess says that acqurers are motvated by prvate benefts such as ncreased prestge when acqurng another publc company. Hence, they are wllng to pay extra so long as the acquston s successful. Ths hypothess s suggested by Draper and Paudyal (2006), and Agrawal and Samwck (2003). However, when acqurng a smaller and less known prvate frm, the prvate motve no longer apples, nstead acqurers are more motvated by the potental synerges from the acquston and a desre to maxmze ther shareholders wealth. Ths hypothess predcts that acqurers tend to overpay for publc targets and not overpay the prvate targets. Thus, announcements of publc target acqustons should brng less beneft to the shareholders of the acqurng frms compared to the announcement of prvate target acqustons The lqudty hypothess If the acquston market s compettve, the acquston tself wll be a zero net present value transacton, and the acqurng frm should have no abnormal returns around the announcement date (Chang, 1998). However, Draper and Paudyal (2006) argue that ths outcome depends on the avalablty of nformaton to generate competton among potental acqurers. Informaton on a publc target s normally readly avalable for the purpose of valuaton and shares are publcly traded. Assumng an effcent market, share prces of a publc target are reasonable estmates of frm value, whereas a prvate frm s not publcly traded and nformaton s scarce and hard to obtan. Because of the llqud market for prvate targets and the lack of competton theren, acqurng frms have greater barganng power n negotaton. Ths suggests that acqurers of prvate targets should gan more than acqurers of publc targets The barganng power hypothess Normally, prvate frms are closely held by a lmted number of shareholders; hence, managers of prvate frms are more concerned to ensure a postve mpact on frm value n acquston negotaton. Draper and Paudyal (2006) suggest that closely controlled frms may have sgnfcant barganng strength, whch allows the owners to receve a better prce for ther frm. Therefore, acqurers of a prvate target may have to agree to a hgher prce, thereby reducng potental benefts to ther shareholders. Ths argument contradcts prevous arguments n predctng lesser gans to acqurers of a prvate target compared to acqurng a publc target. Addtonally, f t s a share-exchange offer, target shareholders may end up as a sgnfcant block holder of the acqurng frm n the post-acquston perod. Consequently, the target s shareholders may play an effectve montorng role n ensurng the post-acquston actvtes are n the shareholders nterest. Ths hypothess predcts greater returns to acqurers of a prvate target compared to acqurers of a publc target. 3. Methodology 3.1. Data Our sample covers all non-fnancal frms that were lsted on the Malaysan stock exchange from 1 January 2000 through 31 December The nformaton on corporate acquston and the announcement date were obtaned from dfferent sources. Pror to July 2003, the source of data was the Record of Issues secton contaned n the monthly publcaton of the exchange, called the Investor Dgest. The data was hand-collected from ths publcaton. Acquston data for July 2003 onwards was obtaned from the exchange webste. Each announcement date obtaned from these sources was crosschecked wth the Busness Tmes daly newspaper. The event date taken s effectvely the frst announcement that appeared n the press on the acquston. Acquston of companes may be of dfferent amounts and szes, rangng from a small fracton of the target to 100% take over. Asquth, Burner and Mullns (1983), and Rahman and Lmmack (2004) suggested that f the nvestment amount n the target s small relatve to the value of the acqurng frm, the change n value from the acquston may not cause much change n the acqurng frm s share prce. To be ncluded n ths study the acquston sze must be more than 33% of the votng rghts of the target frm wth a purchase value no less than RM10 mllon. The ntal sample conssts of 188 observatons. Screenng for the avalablty of nformaton on frm share prce, acqurer s sze and acquston value reduces the sample to 139 observatons for acqurng frms. Share prces and ndex values were obtaned from the Bloomberg database Method of Analyss Ths study uses standard event study methodology to examne the market reacton to an acquston announcement. The analyss starts wth the estmaton of the market model parameters for frm on day t n the estmaton perods, before and after the event: where, R, t Rm, t, t (1) Publshed by Canadan Center of Scence and Educaton 179

4 Internatonal Journal of Economcs and Fnance Vol. 3, No.3; August 2011 R,t = the daly stock return of frm on day t, R m,t = the daly market return (Kuala Lumpur Composte Index) on day t, α = the ntercept measurng the mean return over the perod that s not explaned by the market, β = the slope coeffcent measurng the senstvty of frm s stock to the market, ε,t = s an uncorrelated error term wth mean zero and constant varance. To account for thn tradng n the Malaysan market, we use the Sholes-Wllams (1977) method to adjust the OLS parameters ofˆ and ˆ. These parameters are estmated separately before and after the announcement. Before the announcement, the estmaton perod s for 100 days, from day -130 to day -31 relatve to the announcement day (day 0). The estmaton perod after the announcement s from day 31 to day 130. These estmates obtaned from the pre-event estmaton perod are used to obtan the pre-announcement (days -30 to -1) abnormal returns and those obtaned n the post-event estmaton perod are used to obtan the abnormal return for the announcement and post-announcement days (days 0 to 30). The abnormal returns are calculated as follows: AR R, t Rˆ (2) where, AR,t = abnormal returns for frm on day t, R,t = actual returns for stock on day t, Rˆ = expected return calculated from the market followng model equaton: R ˆ ˆ ˆ R (3) m, t where ˆ and ˆ are the Sholes-Wllams estmates of the OLS parameters. For each day t wthn the event study, the cross-secton average abnormal return ( ARt ) s calculated as follows: N 1 (4) AR t AR N 1 where AR t s average abnormal return for day t, and N s the number of stocks n the portfolo. The cumulatve abnormal return (CAR t ) s obtaned by accumulatng the daly average abnormal returns from the begnnng of the event perod. T 2 (5) CAR T1, T 2 AR t t T1 The CAR reveals the average total effect of an acquston announcement over the specfc perod. To test for the T 1, T 2 sgnfcance of abnormal returns, AR,t s dvded by ts estmated standard devaton, S(AR,t ) to yeld a standardzed abnormal return, AR,t. where, AR ' S ( AR AR * S ) AR ( AR ) T 2 T 1 AR T2 AR, t T1 AR * 2 and where T 1 and T 2 are the begnnng and endng days of the estmaton perods. The test statstc for any gven day s gven by: T t N t AR 1 t 3.3 Sample Descrpton ' 1 / 2 * ( N ) Of the 139 acquston announcements, 32 (23%) are for acqurng publc (lsted) targets and 107 (77%) are for acqurng prvate (unlsted) frms. Tables 1(a) and (b) report the dstrbuton of our data by year, the yearly mean and medan of the acqurers and the target sze (value of purchase) over the study perod. The sze of the acqurng frms equals the share prce one month pror to the announcement multpled by the number of shares ssued. Whle for the target, the sze equals the value of the deal when announced. The table shows that not only are publc targets larger than prvate targets, but also the acqurers of publc targets are larger than acqurers of prvate targets. There s also a great varaton n the mean sze of acqurers from year to year, 180 ISSN X E-ISSN

5 partcularly for publc acqustons. Another pont worth notng s that the prvate targets are ndeed very small relatve to ther acqurers wth a mean relatve sze of 8.0%, whle the relatve sze for publc targets s 31.8%. Insert Table 1(a) and 1(b) 4. Emprcal Results 4.1. Returns to Acqurng Frms Table 2(a) presents our analyss of abnormal returns around the announcement day to the acqurng frms. The table shows the daly average abnormal returns (AR) and the respectve t-statstcs, and also the cumulatve abnormal returns (CAR) that are accumulated begnnng from day -30 to day -30. Table 2(a) shows that sgnfcantly postve abnormal returns exst for day 0 and day 1 for the acqurng frms. Returns on other days are practcally non-sgnfcant. The behavor of the cumulatve abnormal returns (CAR) s shown n Fgure 1. The Fgure shows that the CAR takes a slght upward trend durng the event wndow wth a small hump that lasts for a few days begnnng from day 0. These results clearly show the exstence of a sgnfcant postve mpact of the announcement to the acqurng frms. Table 2(b) presents the analyss on the cumulatve abnormal returns (CARs) and the correspondng t-statstc for varous sub-wndows wthn the event wndow. Snce there may be uncertanty as to the exact date that the announcement nformaton becomes publc, t s safer to take 3 days around day 0, that s day -1 to day 1 as the announcement perod. We note frst that CAR for the announcement days (day -1 to 1) s 1.48%, sgnfcant at the 1% level. Second, for the entre event wndow, day -30 to day 30, the acqurng frm s CAR of 2.64% s nsgnfcant at the 10% level. The pre- and post-announcement perods CAR are also nsgnfcant. Past evdence on the performance of the acqurng frms are mxed; the majorty of the studes show non-postve CAR, whle some studes show small postve CAR. In ther wdely cted paper, Jensen and Ruback (1983) conclude that acqurng frms do not lose. In a more recent paper, Martynova and Renneboog (2008) revew the academc lterature on the mergers and acqustons studes and fnd that acqurng frms earn announcement abnormal returns that are statstcally ndstngushable from zero. Our results, whch show that acqurng frm overall returns are not sgnfcant over the event perod, are consstent wth most of the prevous lterature, for example Bradley, Desa and Km (1983), Moelller, Stegemoller and Stulz (2004), and Draper and Pudyal (2006). Insert Table 2(a) Insert Fgure 1 Insert Table 2(b) 4.2. Acqurng Frm Returns and Target Status Table 3 shows abnormal return analyss for acqurng frms based on the target status. The table shows that the purchase of ether prvate or publc targets would yeld sgnfcant abnormal returns to the acqurng frms. The announcement perod CAR s 1.24% (sgnfcant at the 1% level) and 2.20% (sgnfcant at the 5% level) for acqurng prvate targets and publc targets, respectvely. In addton, we note that the announcement mpact for publc acquston s greater than for prvate acquston. For the entre event wndow (day -30 to day 30), the publc acquston CAR s 4.53% compared to the prvate acquston CAR of 0.50%, gvng a dfference of 4.03%, whch s sgnfcant at the 10% level (sgnfcance test not shown). On the whole, our results ndcate that publc acqustons generate greater abnormal returns than prvate acqustons to the acqurng frms. Insert Table 3 Our results n ths secton contradct the fndngs of most of the prevous studes that show that prvate acqustons generate greater abnormal returns than publc acqustons. These nclude the studes by Chang (1998), Fuller et al. (2002), Conn et al. (2005) and Facco et al. (2006). For example, Fuller et al. (2002) fnd that acqurng publc targets yeld an abnormal return of 0.57% (nsgnfcant) for the acqurng frms durng the event perod of days -2 to 2, whle acqurng prvate targets yeld an abnormal return of 1.90% (sgnfcant at the 5% level). Draper and Paudyal (2006) fnd that acqurers for publc target experence a sgnfcant loss of 0.41% (sgnfcant at the 5% level) whle acqurers for prvate target experence a sgnfcant gan of 0.81% (sgnfcant at the 5% level) around the event perod (day -1 to 1). Two explanatons have been proposed by Chang (1998), and Draper and Paudyal (2006) to explan why prvate acqustons should generate greater returns than publc acqustons. One s the manageral motve and the other s the lqudty hypothess. The manageral motve states that acqustons are motvated by shareholder wealth maxmzaton. In other words, acqurng a prvate target s consdered as a postve net present value nvestment for the acqurng frm; hence, an upward movement n share prces s expected. However, when acqurng a publc Publshed by Canadan Center of Scence and Educaton 181

6 Internatonal Journal of Economcs and Fnance Vol. 3, No.3; August 2011 target, managers of acqurng frms may be more nfluenced by prvate benefts rather than maxmzng shareholder value. Hence, publc acqustons tend to generate less or no abnormal returns, or even negatve abnormal returns. The lqudty hypothess s based on the fact that a publc target has lqudty n the sense that the shares are publcly traded on an exchange and prces are determned by the market. Conversely, a prvate target does not have such lqudty and the deal value s arrved at through prvate negotaton. The lqudty hypothess says that acqurng frms pay a lqudty premum to the publc lsted target and no lqudty premum to the prvate targets. Hence, acquston benefts would be less when acqurng a publc frm compared to acqurng a prvate frm. Our results are not consstent wth ether of these hypotheses. However, that does not mean that the hypothess s nvald n Malaysa. Both the manageral motve and lqudty hypotheses may be consdered as unversal behavors and should be applcable n all markets. In addton, there may be other forces that are present n our data that overcome the value maxmzaton and lqudty motves. One such possblty s the barganng power hypothess. Ths hypothess states that the closely held nature of the prvate target leads to greater barganng power of the target s management, whch ends up wth the acqurng frm havng to pay more than when acqurng a smlar publc target Interacton between Target Status and Method of Payment Acqurng frms may make an offer to buy shares of target frms ether n the form of cash settlement or share-exchange settlement or a combnaton of both. Prevous studes ndcate an overwhelmng preference for cash. We address ths ssue n our study by dvdng the sample accordng to the method of payment. The majorty of acqustons are full cash offers (99 out of 139 cases or 71%), whle 26 cases (19%) are full share offers and the remanng 14 cases (10%) are mxed offers. Followng the work of Draper and Paudyal (2006), ths study only looks at full cash and share offers. Table 4 shows abnormal return analyss n relaton to the method of payment and target status. The last two columns of the table show that cash-settlement offers are clearly preferred by shareholders to share-settlement offers. The table shows that there s a postve abnormal return of 1.43% (sgnfcant at 1% level) on the announcement days of -1 to 1, for cash settlement and an nsgnfcant abnormal return for share settlement. The overall gan for the entre wndow also shows that the CAR for the cash offer (2.85%) s greater than for the share offer (-0.52%). Insert Table 4 The sgnalng hypothess of Myers and Majluf (1984) stpulates that the market beleves that acqurng managers use cash payments when ther shares are undervalued (hence, postve market reacton) and use share payment when ther shares are overvalued (hence, negatve market reacton). The acqurer s shareholders preference for cash-offers, as shown by our evdence, s consstent wth the sgnalng theory. In terms of target status, our fndngs suggest that publc acqustons generate greater returns to acqurng frms than prvate acqustons, for both methods of payment. For cash settlement, the CARs for the announcement perod are 1.02% (sgnfcant at the 5% level) and 2.70% (sgnfcant at the 5% level) for prvate and publc acqustons, respectvely. It s also observed that for all the sub-wndows n the cash settlement subsample, the publc CAR s greater than the prvate CAR. For share-offer settlement, we fnd the results to be less determnstc compared to the cash-offer settlement, however, there s a tendency for the publc acqustons to have a greater abnormal return compared to prvate acqustons. Comparng methods of payment for the dfferent target status, we fnd that cash settlement s clearly preferred to share settlement for both publc and prvate targets. Our results are not consstent wth the corporate montorng hypothess proposed by Chang (1998), and Fuller et al. (2002), whch says that prvate acquston wth share offers should have greater returns than cash offers because the prvate targets are usually owned by a famly or a lmted number of partners and, thus, could effectvely montor post-acquston actvty. Our results are consstent wth Da Slva Rosa et al. (2004) who fnd that share settlement for prvate acqustons s not assocated wth hgher abnormal returns to acqurers. We ratonalze our results by the fact that the sze of the prvate targets s very small relatve to the sze of the acqurng companes. Hence, the target shareholders would not become a sgnfcant block holder that could effectvely montor post-acquston actvty. 4.4 Relatve Sze of Acquston Eckbo, Maksmovc and Wllams (1990) mentoned that the relatve sze of the target to acqurng frms may be a factor n determnng the sze of abnormal returns to the acqurng frms. Acqurng a target that s relatvely very small compared to the acqurer s sze would result n lttle value added to the acqurng frm, whch may not be statstcally sgnfcant. Therefore, t makes sense to expect the sze of the acquston to be related to the amount of benefts the acqurer wll obtan. To explore ths ssue, we calculate the relatve sze of the acquston by dvdng 182 ISSN X E-ISSN

7 the asset purchase value by the market value of the acqurng frm. We then rank the sample by the relatve sze and dvde t nto two equal sub-samples of hgh and low relatve szes, respectvely. The acqurer market value s calculated usng the market prce one month pror to the announcement multpled by the number of shares outstandng. The target value for acquston s taken as the deal value as announced. Table 5(a) reports the abnormal returns analyss for the acqurng frms based on the relatve sze for the prvate acquston. The results show that large acqustons tend to yeld greater abnormal returns to acqurers compared to small acqustons. Ths s true for all sub-wndows. Smlar observatons may be made for the publc acqustons as shown n Table 5 (b). These results are consstent wth those of Asquth, Burner and Mullns (1983), Jensen and Ruback (1983), and Fuller et al. (2002) who fnd that acqurer abnormal returns ncrease wth the relatve sze of acquston. Insert Table 5(a) and 5(b) Comparng the prvate versus publc acqustons by relatve sze (.e. Table 5(a) versus Table 5(b)) we fnd that the publc acqustons yeld greater returns than the prvate acqustons for all sub-wndows and for both sze categores. Because the relatve sze of the publc acqustons s generally larger than the prvate acqustons, there may be two effects drvng these results; the relatve sze effect and the target status effect. We further analyze the two effects by selectng matched-sample acqustons by ther relatve szes, and rerun the abnormal returns analyss. The results are shown n Table 5(c). The table reveals that publc acqustons yeld greater returns compared to the prvate acqustons for all sub-wndows. Therefore, we conclude that our results ndcate the ndependent exstence of both the relatve sze and target status effects. Insert Table 5(c) 4.5. Regresson Analyss Our analyses n the prevous sectons reveal the exstence of several factors that nfluence the returns to the acqurng frms; these nclude method of payment, relatve sze and target status. In order to complement the unvarate analyses, ths secton presents our results on multvarate regresson as an alternatve test f those factors are ndeed sgnfcant n explanng the abnormal returns of the acqurng frms. In addton, the multvarate analyss allows nteracton between varables. The acqurng frm cumulatve abnormal return over the announcement days, CAR(-1,1) s used as the dependent varable. The explanatory varables nclude three dummes (for cash settlement, multple bdders and publc targets), three nteracton varables and the log of deal value relatve sze. The data for the OLS regresson conssts of all-cash and all-share methods of payment only, excludng the mxed offer sample, hence the sample sze s reduced from 139 to 125. The regresson equaton s as follows: CAR(-1,1) = 0 + β 1 (CashDum)+ β 2 (MultDum)+ β 3 LnRS + β 4 (PubDum) + β 5 LnRS*CashDum + β 6 LnRS*PubDum+ β 7 CashDum*PubDum + where, CashDum = cash settlement dummy, where CashDum = 1 f the acqurer uses cash settlement and 0 otherwse, MultDum = multple acqustons dummy, where MultDum =1 f acqurer have multple acqustons and 0 otherwse, LnRS = natural log of relatve sze. The relatve sze s measured by the deal value dvded by the sze of the acqurer. PubDum = publc acquston dummy, where =1 for publc acquston and 0 otherwse, and LnRS*CashDum, LnRS*PubDum and CashDum*PubDum are nteracton varables. We run three regressons: Frst, usng all the data, then usng the subsample data on prvate and publc acqustons, respectvely. The results are presented n Table 6. The F-statstc confrms the overall sgnfcance of the regressons, whle the adjusted R-squared ranges from 11.49% to 24.85%. The postve ntercept for all three regressons confrms that, on average, acqurng frms gan durng the 3-days surroundng the announcement perod regardless of the target status. The estmates of the coeffcents confrm the nfluence of methods of payment, relatve deal sze and the target status to acqurers gans. The coeffcent for cash payment s postve for both prvate and publc acqustons suggestng the overall preference for cash settlement regardless of the target status. The results also show that the coeffcent for the relatve deal sze s sgnfcantly postve, whch means that the larger the relatve deal sze the greater s the mpact to the acqurer s wealth. Ths s consstent wth the sze effect of acquston where the larger the acqured value compared to the acqurer, the larger would be the value to be added to the acqurng Publshed by Canadan Center of Scence and Educaton 183

8 Internatonal Journal of Economcs and Fnance Vol. 3, No.3; August 2011 frm. The nteracton varable representng relatve sze and cash payment has a postve mpact on acqurer s excess returns, suggestng that the nfluence of cash settlement remans postve after controllng for relatve sze. Overall, the regresson results concur wth our unvarate analyss. Insert Table 6 5. Concluson In ths study, we examne stock market reactons to the acqurng frms n corporate acquston announcements n Malaysa for the perod 2000 to We use the market model wth Scholes-Wllams (1977) adjustments for thn tradng to obtan the abnormal returns. Our evdence suggests that acqurng frms obtan statstcally sgnfcant postve abnormal returns on the announcement day (day 0) and around the announcement perod (days -1 to +1). Ths s consstent wth the fndngs of the prevous studes n the Malaysan market, notably Isa and Lm (1993) and Mat-Nor (1993). Prevous studes on the target status reveal that acqurng frms obtan a sgnfcant premum when acqurng prvate (unlsted), as opposed to publc (lsted) targets. Prevous lterature proposes that ths may be explaned by the lqudty hypothess and manageral motve hypothess. Our evdence shows the opposte n the Malaysan context: acqurng frms that acqure publc targets obtan greater abnormal returns than those that acqure prvate targets. Our results are consstent wth the barganng power hypothess. The management of the prvate target has greater barganng power by vrtue of the frm beng closely held by a fewer number of shareholders. Further analyss on the relatve sze of the deal value confrms that sze does matter. We fnd that for both prvate and publc targets, acqustons wth large relatve sze outperform the small relatve sze n terms of provdng abnormal returns to acqurers. Ths fndng s consstent wth Asquth, Burner and Mullns (1983), Jensen and Ruback (1983) and Fuller et al. (2002) who fnd that acqurer abnormal returns ncrease wth the relatve sze of target to acqurng frm. Our matched-par sample test reveals that publc acqustons provde greater abnormal returns to acqurers than prvate acqustons. Hence, there exsts the target status effect beyond the relatve sze. Abnormal returns to the acqurng frms are also found to be nfluenced by the methods of payment of the acquston, ether cash or share settlement. Our evdence shows a clear preference of the market for cash offers as opposed to share offers. Cash offers generate a substantal premum over share offers for acqurng frms. We argue that ths can be explaned by the sgnalng hypothess n that acqurers managers wll use cash offers f they beleve ther shares are undervalued. The evdence that acqurng frms share prces ncrease when cash offers are proposed s consstent wth ths hypothess. We also fnd that both the method of payment effect and the targets status effect are present n our sample; both effects seem to exst ndependently. References Agrawal, R. K. and Samwck, A. A. (2003). Why do managers dversfy ther frms? Agency reconsdered. Journal of Fnance, 58(1), Amhud, Y. Lev, B. and Travlos, N. G. (1990). Corporate control and the choce of nvestment fnancng: The case of corporate acqustons. Journal of Fnance, 45(2), Andrade, G., Mtchell, M. and Stafford, E. (2001). New evdence and perspectve on mergers. Journal of Economc Perspectve, 15(2), Ang, J. and Kohers, N. (2001). The take-over market for prvately held companes: The US experence. Cambrdge Journal of Economcs, 23(26), Asquth, P. and Km, E. H. (1982). The mpact of merger bds on the partcpatng frms' securty holders. Journal of Fnance, 37(5), Asquth, P., Burner, R. F. and Mullns, Jr.D. (1983). The gans to bddng frms from merger. Journal of Fnancal Economcs, 11(1), Bradley, M., Desa, A. and Km, E. H. (1983). The ratonale behnd nterfrm tender offers: nformaton or synergy? Journal of Fnancal Economcs, 11(1-4), Chang, S. Y. (1998). Takeovers of prvately held targets, methods of payment, and bdder returns. Journal of Fnance, 53(2), Cheung, Y. L. and Shum, C. K. (1993). Corporate takeover and shareholders wealth n Hong Kong. Brtsh Accountng Revew, 25(3), Conn, R. L., Cosh, A., Guest, P. M. and Hughes, A. (2005). The mpact on UK acqurers of domestc, cross-border, publc and prvate acqustons. Journal of Busness Fnance and Accountng, 32(5&6), ISSN X E-ISSN

9 Da Slva Rosa, R., Izan, H.Y., Stenback, A. and Walter, T. (2000). The method of payment decson n Australa takeovers: An nvestgaton of causes and effects. Australa Journal of Management, 25(1), Draper, P. and Paudyal, K. (1999). Corporate takeovers: Mode of payment, returns and tradng actvty. Journal of Busness Fnance and Accountng, 26(5&6), Draper, P. and Paudyal, K. (2006). Acqustons: Prvate versus publc. European Fnancal Management, 12(1), Eckbo, B. E., Maksmovc, Y. and Wllams, J. (1990). Consstent estmaton of cross-sectonal models n event studes. Revew of Fnancal Studes, 3(3), Facco, M. and Masuls, R. W. (2005). The choce of payment method n European mergers and acqustons. Journal of Fnance, 60(3), Facco, M., McConnell, J. J. and Stoln, D. (2006). Returns to acqurers of lsted and unlsted targets. Journal of Fnancal and Quanttatve Analyss, 4(1), Fuller, K., Netter, J. and Stegemoller, M. (2002). What do returns to acqurng frms tell us? Evdence from frms that make many acqustons. Journal of Fnance, 57(4), Healy, P.M., Palepu, K.G. and Ruback, R.S. (1992). Does corporate performance mprove after mergers? Journal of Fnancal Economcs, 31(20, Isa, Mansor. and Lm, L. L. (1993). Share prce behavor around acquston announcement n Malaysa. Captal Market Revew, 1(2), Isa, Mansor. (1994). The effects of acquston announcement and method of payment on shareholder returns n the Malaysan stock market. Securty Industry Revew, 20(1), Ismal, Ahmad. (2008). Whch acqurers gan more, sngle or multple? Recent evdence from the USA market. Global Fnance Journal, 19(1), Jensen, M. C. and Ruback, R. S. (1983). The market for corporate control: The scentfc evdence. Journal of Fnancal Economcs, 11(1-4), Koh, Francs. and Lee, S. H. (1988). Rsks and returns of acqurng and acqured frms n Sngapore: An emprcal analyss. Asa Pacfc Journal of Management, 5(3), Leland, H. and Pyle, D. (1977). Informatonal asymmetres, fnancal structure, and fnancal ntermedaton. Journal of Fnance, 32(2), Martynova, M. and Renneboog, L. (2006). The performance of the European market for corporate control: Evdence from 5 th takeover wave. ECGI-Fnance Workng Paper No. 135/2006. Tlburg Unversty, the Netherlands Martynova, M. and Renneboog, L. (2008). A century of corporate takeovers: What have we learned and where do we stand? Journal of Bankng and Fnance, 32(10), Mat-Nor, Fauzas. (1993). The effects of acquston announcement on the securty prces of bddng and target frms: The Malaysan experence. Irsh Busness and Admnstratve Research, 14(2), Moeller, S. B., Schlngemann, F. P. and Stulz, R. M. (2004). Frm sze and the gans from acqustons. Journal of Fnancal Economc, 73(2), Moeller, S. B., Schlngemann, F. P. and Stulz, R.M. (2005). Wealth destructon on a massve scale? A study of acqurng-frms returns n the recent merger wave. Journal of Fnance, 60(2), Myers, S. C. and Majluf, N. J. (1984). Corporate fnancng and nvestment decsons when frms have nformaton that nvestors do not have. Journal of Fnancal Economcs, 13(2), Rahman, R. A. and Lmmack, R. J. (2004). Corporate acqustons and the operatng performance of Malaysan companes. Journal of Busness Fnance and Accountng, 31(3-4), Scholes, M. and Wllam, J. T. (1977). Estmatng betas from nonsynchronous data. Journal of Fnancal Economcs, 5(3), Travlos, N. (1987). Corporate takeover bds, methods of payment, and bddng frms stock returns. Journal of Fnance, 42(4), Travlos, N. and Papaonnou, G. (1991). Corporate acqustons: Method of payment effects, captal structure effects, and bddngs frms stock returns. Quarterly Journal of Busness and Economcs, 30(4), Publshed by Canadan Center of Scence and Educaton 185

10 Internatonal Journal of Economcs and Fnance Vol. 3, No.3; August 2011 Wansley, J. W, Lane, W. R. and Yang, H. C. (1987). Gans to bdder frms n cash and securtes transactons. Fnancal Revew, 22(4), Table 1(a). Sample dstrbuton by year and sze of acqurers and target value for prvate acqustons Acqurng frms (N=107) Target value (N=107) Year N Mean Sze (RM 000) Medan Sze (RM 000) Mean Sze (RM 000) Medan Sze (RM 000) , ,100 78,954 35, , ,435 47, ,375, ,867 57,412 32, , ,932 90,213 35, , ,635 61,800 23,000 Average 888, ,000 70,873 35,396 Note: Durng all the years the exchange rate s fxed at RM3.80 = USD1.00. Table 1(b). Sample dstrbuton by year and sze of acqurers and target value for publc acqustons Acqurng frms (N=32) Target value (N=32) Year N Mean Sze (RM 000) Medan Sze (RM 000) Mean Sze (RM 000) Medan Sze (RM 000) ,289,211 2,202, , , , , , , ,014, , , , , , , , , , , ,481 Average 1,098, , , ,088 Note: Durng all the years the exchange rate s fxed at RM3.80 = USD1.00. Table 2(a). Abnormal returns analyss for the entre sample of acqurng frms around the acquston announcement day. Day AR (%) t-statstc CAR (%) *** ** * Note: Returns are n percentage. H 0 : mean daly abnormal returns AR T = 0. ***, ** and * ndcate sgnfcance at the 1% level, 5% level and 10% level. 186 ISSN X E-ISSN

11 Table 2(b). Cumulatve abnormal returns (CAR) for acqurng frms: all sample Acqurng Frms (N=139) Holdng Returns CAR Range CAR (%) t-statstc Day -30 to Day Day -1 to Day ** Day 2 to Day Day -5 to Day * Day -20 to Day Note: Returns are n percentage. H 0 : cumulatve mean daly abnormal returns n the nterval, CAR T1,T2 = 0. ***, ** and * ndcate sgnfcance at the 1% level, 5% level and 10% level, respectvely. Table 3. Cumulatve abnormal returns (CAR) of acqurng frm by target status: all sample Holdng Returns CAR Range Prvate Acqustons (N=107) Publc Acqustons (N=32) CAR (%) t-statstc CAR (%) t-statstc Day -30 to Day Day -1 to Day *** ** Day 2 to Day Day -5 to Day * Day -30 to Day Note: Returns are n percentages. Publc acqustons refer to acqurng a lsted target and prvate acqustons refer to acqurng an unlsted target. H0: cumulatve mean daly abnormal returns n the nterval, CAR T1,T2 = 0. ***, ** and * ndcate sgnfcance at the 1% level, 5% level and 10% level, respectvely. Table 4. Cumulatve abnormal returns (CAR) of acqurng frm by method of payment and target status. Panel A: Cash Acqustons Holdng Returns Prvate Acquston Publc Acquston All Acqustons CAR Range CAR (%) t-statstc CAR (%) t-statstc CAR (%) t-statstc Day -30 to Day Day -1 to Day ** ** *** Day +2 to Day Day -5 to Day * Day -30 to Day Panel B: Share Acqustons Day -30 to Day Day -1 to Day Day +2 to Day Day -5 to Day Day -30 to Day Note: Returns are n percentage. H 0 : cumulatve mean daly abnormal returns n the nterval, CAR T1,T2 = 0. ***, ** and * ndcate sgnfcance at the 1% level, 5% level and 10% level, respectvely. Publshed by Canadan Center of Scence and Educaton 187

12 Internatonal Journal of Economcs and Fnance Vol. 3, No.3; August 2011 Table 5(a). Cumulatve abnormal returns (CAR) of acqurng frms by relatve sze (RS) of the acquston value for prvate target Holdng Returns CAR Range Small RS (N=54) Large RS (N=53) CAR (%) t-statstc CAR (%) t-statstc Day -30 to Day Day -1 to Day * *** Day 2 to Day Day -5 to Day Day -30 to Day Note: Returns are n percentage. H 0 : cumulatve mean daly abnormal returns n the nterval, CAR T1,T2 = 0. ***, ** and * ndcate sgnfcance at the 1% level, 5% level and 10% level, respectvely. Table 5(b). Cumulatve abnormal returns (CAR) of acqurng frms by relatve sze (RS) of the acquston value for publc target Holdng Returns CAR Range Small RS (N=16) Large RS (N=16) CAR (%) t-statstc CAR (%) t-statstc Day -30 to Day Day -1 to Day ** Day 2 to Day Day -5 to Day Day -30 to Day Note: Returns are n percentage. H 0 : cumulatve mean daly abnormal returns n the nterval, CAR T1,T2 = 0. ***, ** and * ndcate sgnfcance at the 1% level, 5% level and 10% level, respectvely. Table 5(c). Cumulatve abnormal returns (CAR) of acqurng frm by target status: matched sample (by relatve sze) Holdng Returns CAR Range Prvate Target (N=27) Publc Target (N=27) CAR (%) t-statstc CAR (%) t-statstc Day -30 to Day Day -1 to Day * *** Day 2 to Day Day -5 to Day * Day -30 to Day Note: Returns are n percentage. H 0 : cumulatve mean daly abnormal returns n the nterval, CAR T1,T2 = 0. ***, ** and * ndcate sgnfcance at the 1% level, 5% level and 10% level, respectvely. Table 6. OLS Regresson analyses. Independent Varable All sample Prvate acquston Publc acquston Constant ** ** ** Cash Dummy, 1=cash ** ** *** Multple acqustons dummy, 1=multple acqustons Log of deal value relatve sze (RS) *** *** ** Publc acquston dummy, 1=publc ** Interacton varable (RS*CashDum) *** *** *** Interacton varable (RS*PubDum) Interacton varable (CashDum*PubDum) F-statstc *** *** *** Sample sze, N Adjusted R-Squared Note: Dependent varable s CAR(-1,+1). The total sample sze for the OLS regresson conssts of all-cash and all-share methods of payment only. 188 ISSN X E-ISSN

13 CAR (%) Days Fgure 1. Cumulatve Abnormal Returns (CAR) for acqurng frms. Publshed by Canadan Center of Scence and Educaton 189