UNIVERSITY OF ECONOMICS INSTITUTE OF SOCIAL STUDIES HO CHI MINH CITY THE HAGUE VIETNAM THE NETHERLANDS VIETNAM NETHERLANDS PROGRAMME FOR M A IN DEVELOPMENT ECONOMICS ACQUIRER ABNORMAL RETURNS IN M & A[.]
UNIVERSITY OF ECONOMICS INSTITUTE OF SOCIAL STUDIES HO CHI MINH CITY THE HAGUE VIETNAM THE NETHERLANDS VIETNAM - NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS ACQUIRER ABNORMAL RETURNS IN M & A WITHIN BANKS: EVIDENCE FROM SELECTED ASEAN COUNTRIES A thesis submitted in partial fulfilment of the requirements for the degree of MASTER OF ARTS IN DEVELOPMENT ECONOMICS By NGUYEN THI NGOC DUNG Academic Supervisor: CAO HAO THI HO CHI MINH CITY, November, 2013 Contents List of tables iii List of figures iii Abbreviation iv Abstract v Chapter : Introduction 1.1 Problem statement 1.2 Research objectives .2 1.3 Research questions 1.4 Research scope 1.5 Research contribution 1.6 Research structure Chapter : Literature review 2.1 Definition of M & A 2.2 Types of M & A 2.3 Motivation for M & A 2.4 Method for calculating event’s effect 2.4.1 Important time frames .6 2.4.2 Calculating Method 2.5 Testing significant of abnormal return 2.6 Factors affect M & A 11 2.7 Overview the using of event study in banking M&A 15 2.8 Overview about bank M & A in ASEAN countries 15 2.9 Conceptual framework 16 Chapter : Methodology 18 Page: i 3.1 Research process 18 3.1.1 Estimation of abnormal returns 18 3.1.2 Analyzing factors effect CAR 21 3.2 Data collection .21 Chapter : Data analysis .26 4.1 Descriptive statistics 26 4.1.1 Abnormal return and cumulative abnormal return to acquirers 26 4.1.2 Cross-section regression analysis 34 4.2 Inferential statistics 38 Chapter : Conclusions and recommendations 40 5.1 Conclusion 40 5.2 Recommendations .41 5.3 Limitation and further research 42 Reference 43 Appendice .50 Page: ii List of tables Table 3-1: Descriptive statistics of sample characteristic 22 Table 3-2: M&A by country .23 Table 3-3: M & A by year 24 Table 4-1: Average abnormal return, and number of positive and negative observations for 20 days before through 20 days after announcement date 28 Table 4-2: Average abnormal returns between groups and test of difference 30 Table 4-3: Cumulative average abnormal returns and their t statistics .33 Table 4-4: Cumulative average abnormal returns between groups and test of difference .33 Table 4-5: CAAR(-18, 10) categorized by characteristics 35 Table 4-6: Correlation testing 37 Table 4-7: Cross-section regression result 38 List of figures Figure 2-1: Process for calculating abnormal return 17 Figure 2-2: Acquirer's cumulative average abnormal return (CAR) from merger and potential effect factors 17 Figure 4-1: Average abnormal returns from -20 days to +20 days around announcement date 28 Figure 4-2: Cumulative abnormal returns from -20 days to +20 days around announcement date 32 Page: iii Abbreviation AAR(s): Average Abnormal Return(s) APT: Arbitrage Pricing Theory AR: Abnormal return ASEAN: Association of Southest Asian Nations B&B: Bank and Bank B&O: Bank and Other institute CAAR(s): Cumulative Average Abnormal Return(s) CAPM: Capital Asset Pricing Theory CAR(s): Cumulative Abnormal Return(s) M & A: Mergers and Acquistions Page: iv Abstract This paper examines acquirer’s abnormal return from merger and acquisition (M & A) between two banks, and between bank and non-bank institute By using event study and market model, M & A cases announcing from Jan 2005 to Dec 2012 of ASEAN are investigated Besides that, determinants affect abnormal return like acquirer size, listing target status, payment method, learning by doing, bidder leverage, type of M & A, Tobin q ratio, target nation, etc are analyzed in detail Key words: ASEAN mergers, acquisition, M & A, abnormal return, event study, market model Page: v Chapter : Introduction 1.1 Problem statement The problem of merger and acquisition (M & A) has been mentioned much these days in Vietnam This can be checked through a famous searching tool – Google by typing key word “sáp nhập” – mean merger, and you will find at least 426,000 related results with that This trend can be understood that when the economic, or especially financial situation become hard, firms think more about M & A as a resolution for rescuing To measure the effectiveness of a M & A deal, we need even years to know, but it does not take long time to see the reaction of the market Besides that, according to Andrade, Mitchell and Stafford (2001), market response around acquisition announcement is the best way to examine the success of the deal Problem is how to measure the response of market in front of M & A news? One of the common methods is comparing the return from the company’s stock around the time of announcement of M & A which called abnormal return and average return of it in normal time (expected return) By using market model and event study, we can calculate abnormal return to specific object like acquirers (who offer merger), target firms, or even their industry rivals, etc Applying this method, many studies have done With specific country, there are some researches like Bae, Kang and Kim (2002) about Korean Chaebol, Filbien et al (2011) about Canada, Brown and Fung (2009) about Japan Keiretsu, etc With specific industry, many researches about merging between banks, such as: James and Weir (1987), Hannan & Wolken (1989), Houston and Ryngaert (1994), DeLong (2001), Cornett et al (2003), Anderson et al (2004), Beitel et al (2004), Lepetit et al (2004), Karceski et al (2005), DeLong and DeYoung (2007) etc, or about telecommunication firms (Akdogu, 2009) One notice point here is most of above studies based on data and situations of developed market like US, or European Limited studies about M & A in emerging market: William & Liao (2008) search Page: for value created between international banks and targets banks in emerging market, Crouzille et al (2008) measure the reaction of ASEAN stock market to bank mergers after the 1997 financial crisis, Goddard et al (2012) research about emerging market including Asia and Latin America To date, as what we know, there is no studies really calculate abnormal return in merger cases within ASEAN countries 1.2 Research objectives By doing this research, we try to find out reaction of market in front of M&A acquisition announcement especially in bank mergers cases of ASEAN In details, this paper will measure abnormal returns to acquirers in bank mergers It also compares the return of acquirer come from acquisition among banks, and between bank and other institute Moreover, some common factors like acquirer size, listing target status, payment method, studying by doing, bidder leverage, types of merger, Tobin q ratio, and target nation will be analyzed to estimate the impact of potential factors on abnormal return 1.3 Research questions From research objectives, following questions are tried to answer: - Are abnormal returns to acquirers in bank mergers positive? - Are abnormal returns to acquirers in bank mergers higher than those of mergers between bank and other institution? - What is the impact of common factors on abnormal return? 1.4 Research scope With the objective of researching about ASEAN situation, the paper will be conducted base on bank merger within six ASEAN countries: Indonesia, Malaysia, Philippines, Singapore, Thailand, and Vietnam Moreover, updated deals announced and completed from 01/01/2005 to 31/12/2012 are selected The information about deals will be collected from Zephyz – Bureau van Dijk – one of the best databases about M&A around the world Page: 1.5 Research contribution Answering three research questions, the usefulness of this paper concentrate on two main points First, it is an overview about bank mergers in ASEAN countries within the period from 2005 to 2012 Second, from the lesson of ASEAN countries, this study is expected to be a reliable source for further researches related to this field 1.6 Research structure Continue to this introduction chapter, some literature concepts about merger and acquisition (M & A) as well as related empirical studies will be presented in chapter Methodology and data collection will be stated clearly in chapter Data analysis will be indicated in chapter Chapter will end with conclusion and recommendations for further researches Page: Chapter : Literature review In this chapter, some important points related to M & A will be explained First, definition, types, and motivation of M & A will be stated clearly in Section 2.1, 2.2, and 2.3 Then, Section 2.4, 2.5 will show an overview about recent methods for calculating event’s effect as well as testing significance Next, some main factors effect M & A will be discussed Section 2.7, 2.8 will present an overview about the using of event study in banking M & A, and situation of M & A in banks in ASEAN countries Last section will end with conceptual framework 2.1 Definition of M & A Merger and acquisition (M & A) are usually seen going with each other However, these two concepts have a little bit difference in meaning While merger is defined as the act of joining two or more organizations or businesses into one (Oxford dictionary), acquisition is understood as the act of buying something to add to what they already own Acquisition is sometimes called “takeover” In this paper, we not distinguish these two words and use it with same meaning 2.2 Types of M & A According to Megginson & Smart (2008), there are three main kinds of merger include: horizontal, vertical and conglomerate M & A In which, horizontal merger is a combination of competitors within the same geographic market (p.854) This kind of merger is considered as a market extension method and the greatest potential for wealth creation It is explained that by combining of their resources, merged firm will have advantage in scale and scope of economy, as well as saving cost thanks to decreasing, or eliminating overlapping resources Moreover, the combination creates a market power for merged firms compare to other weaker competitors Vertical merger occurs when companies with current or potential buyer-seller relationships combine to create a more integrated company (p.855) Main advantage of vertical merger is decrease the risk of acquirer/target’s input/output which leads to higher efficiency However, this kind of merger must faces with a serious problem of entering a new line of business which they maybe have not enough knowledge to manage With last kind of merger, pure Page: 2.7 Overview the using of event study in banking M&A As MacKinlay (1997), event study was first published by James Dolley in 1933, and many studies use this method after that By reviewing data from five leading journals include Journal of Business, Journal of Finance, Journal of Financial Economics, Journal of Financial and Quantitative Analysis, Review of Financial Studies, Kothari and Warner (2006) found that there were about 565 studies which applied event study within the period of 1974 to 2000 Although it has been long process of development, basic format of event study was not much change, and still focused on calculating securities’ mean as well as cumulative mean abnormal return around event time One of the most important changes in methodology of event study is the using of daily stock data rather than monthly one This is considered as an advance step which helps to give more precise calculation and better explanation (Kothari and Warner, 2006) Measuring wealth creation of M&A in banking field, empirical studies gave out different results While Trifts and Scalon (1987), James and Weir (1987) found significant gain for target firms, and no abnormal return to acquirer, Millon Cornett and De (1991) reported positive return to acquirer on the day of announcement Examining combine wealth of the bidder and the target, Hannan and Wolken (1989) discover that wealth transfer from acquirer to its target, and no overall gain It is clearly that in different situation and market, we have not same earning from M & A What we should is analyzing the detail of situation so can find out right lesson for self 2.8 Overview about bank M & A in ASEAN countries Merger in bank is more special than merger in other industries Beside the role of facilitating, bank provide liquidity to the economic Because of this important function, banks are usually controlled by central bank and must follow strictly the regulation of government This make merger in bank fields become more complicated compare to other fields Page: 15 Moreover, with the historical dependence on the banking sector (Eichengreen and Luengnaruemitchai, 2004), ASEAN countries’ government often keep a close watch over their banking system In some bank acquisitions, government may provide support, or even as a middleman Acquirer chooses to take over a target not only base on profit but sometimes base on arrangement of Government Besides that, in the period of 2005 to 2012, there was a wave of M&A which bank tried to restructure after crisis Among 305 cases of merger which have at least one party of bank from Zephyz and among 95 cases indicated information, we find out 62 cases of restructuring purpose This gives us an overview about situation of merger in Asian countries About abnormal return to bank mergers in emerging markets including ASEAN countries, limited studies had done and gave mixed results William and Liao (2008) when research the cross countries merger cases between acquirer from developed countries and target in emerging markets, reported a positive abnormal return to both acquirers and targets In contrast, Crouzille et al (2008) record an evidence of negative abnormal return to both bidder and targets when examining domestic bank mergers in eight Southeast Asian countries within the period of 1997 to 2000 Most recent study of Goddard et al (2012) based on 132 bank mergers in Asia and Latin America between 1998 to 2009 conclude that, on average, M&A tend to create value for targets while not lose value for shareholder of acquirer 2.9 Conceptual framework According to event study theory, announcement of unexpected event may create an impact on firms’ stock prices Under ordinary conditions, stock price of firm is expected to follow the trend of market However, new information may drive stock price increase or decrease “unexpectedly” which lead to abnormal return (Cowan, 1992) Figure 2-1 summarizes that process Page: 16 Figure 2-1: Process for calculating abnormal return Normal return of market porfolio in pre-event period Normal/Expected return in event window Normal return of the bidder in pre-event period Abnormal return Actual return in event window (stock price) To have an overview about bank merger and factors affect merger’s result, a framework is shown in Figure 2-2 It is expected that there will be a positive effect on stocks price which leads to positive abnormal return to shareholder of acquirer Moreover, eight factors which have impact on abnormal return around announcement of merger are shown without sign Positive or negative effect of each factor will be analyzed more with data in empirical result part Figure 2-2: Acquirer's cumulative average abnormal return (CAR) from merger and potential effect factors Listing status of target Bank M&A Bank + Type of merger Method of payment Tobin’s ratio CAR of ACQUIRER Target nation Learning by doing Bidder’s leverage Acqurier size Literature chapter have reviewed briefly some concepts related to merger, event study method, and ASEAN situation about merger Next chapter will present the method, and data which used in our thesis Page: 17 Chapter : Methodology This chapter will present the process of research as well as the way of collecting data In research process, two main parts will be explained: How to estimate abnormal return is in Section 3.1.1; How to analyze effect of drivers on CAR is in Section 3.1.2 After describing method, the way to collect data, selection criteria and characteristics of sample will be indicated in Section 3.2 3.1 Research process 3.1.1 Estimation of abnormal returns In this study, event study method similar to what described in Brown and Warner (1985), and MacKinlay (1997) is employed to test our hypothesis In general, five major steps will be processed as follow: (1) Define event date and event window; (2) Calculate normal return of each firm in the absence of event within event window; (3) Measure abnormal return (the difference between actual observed returns and above normal return); (4) Aggregate abnormal returns across firms and across time; (5) Test abnormal returns if they are significant or not This procedure is followed study of Henderson (1990) In order to set up the length of event window and estimated window for our study, we follow the advice of MacKinlay (1997) The reason for doing this is event study and its application in economics as well as finance was reviewed and summarized well in MacKinlay’s study In details, event window within [-20 days; +20 days], and estimation window of [-121 days; -21 days] around the day of merger announcement date are chosen This choice is also consistent with recent empirical studies: Hannan & Wolken (1989), Millon Cornett & De (1991) chose window of [15 days, + 15 days], Baradwaj et al (1990) chose [-60 days, +60 days], Delong (2001) select [-10 days, +10 days] Moreover, within [-20 days; +20 days] event window, some other flexible intervals will be chosen based on the significant value of abnormal return Second, normal returns (expected returns) are calculated base on market model: Page: 18 Rit = αi + βiRmt + εit (5) In which, Rit is the daily return of bidder i at time t, Rmt is the daily return of the market portfolio - FTSE/ASEAN index at time t, and εit is error term which assumed to be normally distributed Two parameter αi and βi are estimated by market model regression over a period of 121 days to 21 days before announcement date of merger Third, abnormal return of each day for each firm (ARit) will be obtained by deducting expected return from actual return: ARit = R it − α ̂i − β̂R mt (6) Forth, with a sample of N firms, average abnormal return of each day within event window (cross time abnormal return) will be computed as follow: AARt = 𝑁 ∑𝑁 𝑖=1 𝐴𝑅 it (7) Cumulative abnormal returns (CAR) from event day t1 to event day t2 of each firm (cross firm abnormal return) can be calculated under following function: CARi (t1,t2) = ∑𝑡2 𝑡=𝑡1 𝐴𝑅 it (8) Then, Cumulative average abnormal returns (CAAR) from all firms within event window t1 to t2 will be estimated by: CAAR (t1,t2) = ∑𝑡2 𝑡=𝑡1 𝐴𝐴𝑅 t (9) or ∑𝑡2 𝑡=𝑡1 𝐴𝑅 it (10) ∑𝑁 𝑖=1 𝐶𝐴𝑅(t1,t2) (11) CAAR (t1,t2) = N or CAAR (t1,t2) = N Page: 19 Last, to test the significant of abnormal return different from zero, for each day t, we employ t stat in (2) In which, variance of average abnormal return at day t will be calculated according to three methods of MacKinlay, Brown and Warner, and Kothari and Warner as follow: As MacKinlay (1997): 𝑣𝑎𝑟(𝐴𝐴𝑅𝑡 ) = 1932 ∑193 𝑖=1 𝑣𝑎𝑟(𝐴𝑅𝑖𝑡 ) (12) In which, variance of abnormal return of each firm at each time is independent through time, and equal to the disturbance variance within estimation window (-121 days,-21 days): 𝑣𝑎𝑟(𝐴𝑅𝑖𝑡 ) = 99 ∑−21 ̂i − β̂R mt )2 −121(R it − α (13) As Brown and Warner (1985): var(AARt) = ̅̅̅̅̅̅ ∑−21 𝑡−121 (𝐴𝐴𝑅𝑡 − 𝐴𝐴𝑅) ̅̅̅̅̅̅ 𝐴𝐴𝑅 = AAR t = 100 101 𝑁 (14) ∑−21 −121 𝐴𝐴𝑅𝑡 (14.1) ∑𝑁 𝑖=1 𝐴𝑅𝑖𝑡 (14.2) (t vary from (-121 days, -21 days)) As Kothari and Warner (2006): var(AARt) = ̅̅̅̅̅̅ ∑+20 𝑡−121 (𝐴𝐴𝑅𝑡 − 𝐴𝐴𝑅) 141 Page: 20 (15) ̅̅̅̅̅̅ 𝐴𝐴𝑅 = 142 AAR t = ∑+20 −121 𝐴𝐴𝑅𝑡 𝑁 ∑𝑁 𝑖=1 𝐴𝑅𝑖𝑡 (15.1) (15.2) (t vary from (-121 days, +20 days) For CAAR within interval (t1,t2): t-stat (CAAR(t1-t2)) = 𝐶𝐴𝐴𝑅(t1 ,t2 ) √var(CAAR(t1 ,t2 ) (16) Where variance of cumulative average abnormal return equal to total variance of average abnormal return from t1 to t2: t var(CAAR(t1 , t )) = ∑t21 var(AAR t ) (16.1) To test whether the differences in abnormal return between two groups of mergers are significant or not, we employ two-sample t-test (3) where 𝑥̅1 , 𝑥̅1 , v1, v2, N1, N2 are mean value of AAR/CAAR, variance and size of two groups 3.1.2 Analyzing factors effect CAR Cross-sectional regression method is employed to investigate the effect of potential factors – characteristic of acquirers, targets, and mergers on CAR In first stages, how different the CARs corresponding with each characteristic are identified In second stage, an OLS model is run to examined the association effect of factors on abnormal return 3.2 Data collection This study examines merger cases involving banks in ASEAN countries between 2005 and 2012 The sample is identified through the search on Zephyr – Bureau van Dijk - one of the best databases about M & A cases around the world The deals will be selected base on following criteria: (1) Acquirer and target are within ASEAN countries, and at least one of them is bank as Zephyz’s classification; (2) Acquirer Page: 21 is listed to make sure their stock prices are available; (3) Deals announced on or after 01/01/2005 up to 31/12/2012, and in completed status; (4) Type of deal is acquisition, or merger; (5) Method of payment is cash, or shares Above criteria resulted in a sample of 209 cases Some merger proposals were further eliminated because of following reasons: bidder announces a second merger within 20 days prior to the initial one; there are not enough stock prices and other financial data of the merger deal Final sample contain 193 merger cases which have characteristic described in Table 3-1 below Table 3-1: Descriptive statistics of sample characteristic No Characteristic of merger cases Number of cases Kind of M & A Between Bank and Bank 42 21.76% 151 78.24% 170 88.08% 23 11.92% 163 84.46% 30 15.54% Acquirer experienced prior merger 72 37.31% Acquirer not have experience self 121 62.69% 13 6.74% 181 93.78% Between Bank and other institute Method of payment Paid by cash Paid by non – cash method Cross countries element Acquirer and target in same country Acquirer and target in different country Percentage Experience Listing status of target Target list Target non-list Through descriptive table, a general comment is unbalance of distribution of sample First, there are only 42 cases of merger between bank and bank while 151 cases of merger are between bank and other institute Second, nearly 90% of mergers are performed by cash Third, 163/193 mergers happened between acquirer and target in same country Forth, 94% of target is non-list Only the ratio between acquirers who had experienced prior merger and ones who have no experience is rather balanced (ratio of 4:6) The large different in scale of each kind is not good for comparing the effect of each group with each other Page: 22 In order to collect acquisitions related to bank, Zephyz’s classification is followed with item “bank” Although they classify as “bank”, it includes monetary intermediation, trusts, funds and similar financial entities, and other financial service activities except insurance and pension funding (code 641, 643, 649 of UK SIC (2007)) In this study, a deal is considered as merger between bank and bank when both have code of 641, 643, or 649, and so on With experience factor, an experience firm is distinguished with non-experience firm by investigate their history of merger A firm is considered as having experience if it occurred at least one merger deal prior to the initial one, and within the period of 2005 to 2012 Moreover, It is not care how many merger deals the firm had done before because just only the difference between a freshman and an experience one is concentrated Table 3-2: M&A by country Country Indonesia Indonesia 12 Malaysia Malaysia Singapore 72 6 Thailand 87 52 65 19 19 Vietnam 21 Total 12 Thailand Total Vietnam 12 Philippines Singapore Philippines 74 65 24 3 193 Table 3-2 gives us an overview about number of mergers related to bank within six ASEAN countries Malaysia is rank as first one with 87 deals, and Singapore is second position with 65 cases These two countries leave behind other four countries about number of merger deals Besides that, Malaysia and Singapore are also two countries have most cross-border mergers – 12 deals Indonesia, Thailand, Vietnam have only domestic mergers within years Page: 23 Table 3-3: M & A by year Year Number of deals Bank & Bank Bank and other Total 2005 10 14 24 2006 11 13 2007 23 27 2008 21 28 2009 17 23 2010 20 22 2011 26 32 2012 19 24 Total 42 151 193 Distribution of mergers by year is stated clearly in Table 3-3 Year 2011 is marked with highest number 32 deals while maximum deals in other years were only 28 Among 42 deals merger between bank and bank, about 25% was happened in 2005, and only deals occurred in each year of 2006 and 2010 After getting deals of merger from Zephyz, firms’ stock prices will be collected through DataStream – global financial and macroeconomic database in the world Besides the data of daily stock prices, the benchmark which used to calculate normal return is FTSE/ASEAN index this index is chosen because it includes market index of five countries: Malaysia, Indonesia, Singapore, Philippines, and Thailand Instead of collecting indexes of six countries related to our sample, FTSE/ASEAN index seem a good option when it represent for index of five among six related countries With financial data needed for regression cross-section model, it will be collected from Thomson One Banker – a famous financial database of public companies The way that value of the factor to be defined is as follow: First, logarithm of market value of acquirer’s equity will represent for acquirer size factor Logarithm value is chosen to show relative scale of acquirer Second, leverage ratio is calculated by dividing acquirer’s total equity by its debt Third, Tobin’s q ratio is the ratio of bidder’s market capital and its total asset value Other factor like listing status of target, method of payment, type of merger, target nation, learning by doing will be Page: 24 dummy variables Specifically, if target is a list firm then receiving the value of 1, otherwise is 0; if 100% payment by cash then value is 1, otherwise 0; merger between bank and bank: 1, otherwise: 0; acquirer and target in same country: 1, otherwise: With dependent variable CAR of each firm, we choose CAR in the interval within the period of (-20 days, +20 days) which CAARs reach highest value Methodology chapter have already defined the data used, method to estimate, as well as the way to test statistically significant of abnormal return In next chapter, empirical result will be presented and analyzed in detail After that, inferential statistics section will explain more about some conclusions withdrawn from our data Page: 25 Chapter : Data analysis 4.1 Descriptive statistics Empirical result of this thesis will be presented in this part First, Section 4.1.1 will analyze average abnormal returns and cumulative abnormal returns to acquirers in mergers Second, result from cross section regression on characteristics of acquirer, target, and merger to cumulative abnormal returns will be presented and analyzed in Section 4.1.2 4.1.1 Abnormal return and cumulative abnormal return to acquirers Abnormal return of each firm in samples is calculated based on market model same as the method of Brown and Warner (1985), and MacKinlay (1997) In order to examine abnormal return to merger between bank and bank (B&B), and merger between bank and other institute (B&O), all samples of 193 mergers are split into two groups respectively B&B group contain 42 mergers, in which both sides are banks, and the rest of mergers (151) belong to B&O group To test the statistically significant of average abnormal return (AARs) and cumulative average abnormal return (CAARs), three methods of statistic testing of MacKinlay (1997), Brown and Warner (1985), and Kothari and Warner (2006) are applied The testing results from these methods gives almost same conclusion, so in some tables, only one t-stat result are shown for representation Figure 4-1 plots the behavior of average abnormal returns (AARs) 20 days before through 20 days after announcement date for all samples, B&O group, and B&B group With all samples, AARs are not much fluctuated in the period before announcement time although recording a bottom of - 0.788% in day -19 On the date of announcement of merger and four days after that AARs are negative numbers However, from day +5 to +10, it shows a trend of increasing, and to the day of +10, the top is hit with 1.086% The trend of changing in B&O group is rather consistent with the trend of all samples This can be explained partly by the reason that the merger deals of B&O group make up more than 78% of total deals However, the amplitude of fluctuation in this group is higher than all samples one Page: 26 The trough is reached on day -19 with -1.026%, and the top is hit on day +10 with +1.358% Not like the behavior of above two groups, the changing in B&B group is big different In this group, lowest return is -0.997% on day -12, and highest return is 0.667% on day -1 In addition, abnormal returns between two consecutive days are rather big Representative example is the change of average abnormal return from 0.997% on day -12 to 0.395% on day -11 The fluctuation in this group show an unstable expect of shareholder about success of mergers Besides that, average abnormal return in period before announcement date is higher than those of period after official information of merger The above situation implies a leakage of information before official announcement, and merger announcement seems to receive negative reaction from market at first This comment is based on two reasons First, lowest daily abnormal return is recorded on day -19 with all samples and B&O group, while with B&B group lowest level is hit on day -12 In addition, AARs of B&O group and B&B group in this period show an unusual and high fluctuation On the other hand, within three days after publishing merger information, there is a trend of decreasing in AARs With B&B group, AARs start at 0.565% on announcement date, then go down and record negative value of -0.536% on the third day With all samples and B&O group, although there is a slightly increasing in AARs in day after announcement, a falling trend happen next two days Tải FULL (63 trang): https://bit.ly/3DGQVpw Dự phòng: fb.com/TaiHo123doc.net Details of AARs in each day for all sample, two groups of mergers, their significant, and number of positive and negative observations are showed in Table 4-1 The first point we see in this table is the tops, and troughs of each group are highly significant Moreover, with all samples, daily abnormal returns are statistically significant on day -19, -16,-11, -3, 7, 8, 10 With B&O group, beside significant abnormal return on above days, day 18 and 19 are also included With B&B group, number of days which have significant abnormal returns even more with 12 days among 41 days of event window This confirms an actual occur of Page: 27 abnormal returns around event date The second point to be seen is number of positive observations on almost days is smaller than number of negative ones both in all samples and two groups Number of firms benefit from mergers is smaller than those lost from mergers Tải FULL (63 trang): https://bit.ly/3DGQVpw Dự phòng: fb.com/TaiHo123doc.net Figure 4-1: Average abnormal returns from -20 days to +20 days around announcement date 1,500% 1,000% 0,500% All -25 -20 -15 B&O 0,000% -5 -10 10 15 20 25 B&B -0,500% -1,000% -1,500% Table 4-1: Average abnormal return, and number of positive and negative observations for 20 days before through 20 days after announcement date All (n = 193) Day AAR (%) B&O (n= 151) CAR (%) Positive: Negative AAR (%) B&B (n=42) CAR (%) Positive: Negative AAR (%) CAR (%) Positive: Negative -20 -0.190 -0.190 87:106 -0.154 -0.154 71:80 -0.317 -0.317 16:26 -19 -0.788*** -0.978 77:116 -1.026*** -1.181 57:94 0.068 -0.249 20:22 -18 0.173 -0.805 85:108 0.213 -0.968 69:82 0.028 -0.221 16:26 -17 0.296 -0.509 87:106 0.372 -0.595 69:82 0.020 -0.201 18:24 -16 0.514** 0.005 108:85 0.538* -0.057 82:69 0.427* 0.226 26:16 -15 -0.083 -0.078 94:99 -0.045 -0.102 74:77 -0.218 0.008 20:22 -14 -0.159 -0.238 84:109 -0.237 -0.339 62:89 0.119 0.127 22:20 -13 0.078 -0.160 89:104 0.045 -0.294 67:84 0.193 0.320 22:20 -12 -0.303 -0.464 81:112 -0.111 -0.404 71:80 -0.997*** -0.677 10:32 -11 0.471* 0.008 85:108 0.492* 0.088 67:84 0.395 -0.281 18:24 Page: 28 All (n = 193) Day AAR (%) B&O (n= 151) CAR (%) Positive: Negative AAR (%) CAR (%) B&B (n=42) Positive: Negative AAR (%) CAR (%) Positive: Negative -10 -0.087 -0.079 93:100 0.012 0.101 77:74 -0.444* -0.726 16:26 -9 -0.028 -0.107 89:104 0.057 0.157 73:78 -0.332 -1.058 16:26 -8 -0.044 -0.151 93:100 -0.202 -0.045 66:85 0.527** -0.531 27:15 -7 -0.041 -0.191 89:104 0.118 0.073 73:78 -0.611** -1.142 16:26 -6 0.223 0.031 84:109 0.219 0.292 66:85 0.236 -0.906 18:24 -5 -0.081 -0.050 84:109 -0.006 0.286 66:85 -0.351 -1.257 18:24 -4 -0.272 -0.322 88:105 -0.302 -0.016 71:80 -0.165 -1.422 17:25 -3 0.536** 0.214 90:103 0.524* 0.508 66:85 0.577** -0.845 24:18 -2 -0.169 0.045 78:115 -0.268 0.240 59:92 0.187 -0.658 19:23 -1 0.059 0.103 93:100 -0.111 0.130 73:78 0.667*** 0.009 20:22 -0.001 0.102 86:107 -0.159 -0.029 70:81 0.565** 0.573 16:26 0.040 0.142 88:105 -0.069 -0.098 67:84 0.433* 1.007 21:21 -0.178 -0.036 92:101 -0.213 -0.311 71:80 -0.055 0.952 21:21 -0.333 -0.369 83:110 -0.277 -0.588 66:85 -0.536** 0.416 17:25 -0.079 -0.448 86:107 0.002 -0.585 71:80 -0.369 0.047 15:27 0.206 -0.242 96:97 0.328 -0.258 78:73 -0.232 -0.185 18:24 -0.122 -0.363 95:98 -0.093 -0.351 73:78 -0.224 -0.409 22:20 0.503** 0.139 88:105 0.720*** 0.369 72:79 -0.278 -0.687 16:26 -0.469* -0.329 81:112 -0.472 -0.103 65:86 -0.456* -1.143 16:26 -0.245 -0.574 89:104 -0.384 -0.487 66:85 0.253 -0.890 23:19 10 1.086*** 0.511 95:98 1.358*** 0.871 73:78 0.108 -0.782 22:20 11 -0.292 0.219 81:112 -0.387 0.484 64:87 0.048 -0.734 17:25 12 -0.104 0.115 83:110 0.057 0.541 70:81 -0.683*** -1.417 13:29 13 0.016 0.131 82:111 0.001 0.543 60:91 0.066 -1.351 22:20 14 -0.116 0.015 85:108 -0.079 0.464 63:88 -0.249 -1.600 22:20 15 0.241 0.256 96:97 0.316 0.780 74:77 -0.028 -1.628 22:20 16 0.067 0.323 90:103 0.099 0.878 69:82 -0.047 -1.675 21:21 17 -0.149 0.173 93:100 -0.095 0.784 77:74 -0.345 -2.020 16:26 18 -0.377 -0.203 86:107 -0.557* 0.227 63:88 0.270 -1.750 23:19 19 -0.109 -0.312 89:104 -0.192** 0.035 69:82 0.191 -1.559 20:22 20 0.016 -0.296 * Significant at 10% level ** Significant at 10% level *** Significant at 10% level 102:91 0.027 0.061 79:72 -0.023 -1.582 23:19 Result of testing the difference in daily abnormal returns between groups is shown in Table 4-2 Once again, the testing results performed according to MacKinlay, Brown and Warner, Kothari and Warner seem give same conclusion Within 41 days around announcement date, average abnormal return between B&B group and Page: 29 6677038 ... Negative -2 0 -0 .190 -0 .190 87:106 -0 .154 -0 .154 71:80 -0 .317 -0 .317 16:26 -1 9 -0 .788*** -0 .978 77:116 -1 .026*** -1 .181 57:94 0.068 -0 .249 20:22 -1 8 0.173 -0 .805 85:108 0.213 -0 .968 69:82 0.028 -0 .221... -0 .444* -0 .726 16:26 -9 -0 .028 -0 .107 89:104 0.057 0.157 73:78 -0 .332 -1 .058 16:26 -8 -0 .044 -0 .151 93:100 -0 .202 -0 .045 66:85 0.527** -0 .531 27:15 -7 -0 .041 -0 .191 89:104 0.118 0.073 73:78 -0 .611**... -0 .611** -1 .142 16:26 -6 0.223 0.031 84:109 0.219 0.292 66:85 0.236 -0 .906 18:24 -5 -0 .081 -0 .050 84:109 -0 .006 0.286 66:85 -0 .351 -1 .257 18:24 -4 -0 .272 -0 .322 88:105 -0 .302 -0 .016 71:80 -0 .165 -1 .422