An event study analysis of Chinese listed companies' short-term M&A performance

(1)

An Event Study Analysis of Chinese Listed Companies’

Short-term M&A Performance

Amsterdam Business School

Msc Finance, Quantitative Finance Track

Master Thesis

Student: Mengjie Zhang, 11721898

June, 2018

(2)

Statement of Originality

This document is written by Mengjie, Zhang, who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document are original and that no

sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

Acknowledgements

I feel much indebted to many people who have instructed and favored me in the course of writing this thesis.

First of all, I would like to express my heartfelt gratitude to my supervisor, Dr. Terovitis, for his warm-heart encouragement and most valuable advice, especially for his insightful comments and suggestions on the draft of this thesis. Without his help, encouragement and guidance, I could not have completed this thesis.

Then, I would also like to express my thanks to my family for their valuable encouragement and spiritual support during my study.

Meanwhile, I would like to express my thanks to my friends who prepared for CFA exam together during May and June, and encouraged each other to finally complete this thesis.

(4)

Abstract

Whether or not M&A creates shareholders’ wealth of either targets or acquirers, and how different factors can influence the potential effect are both important and meaningful. Unlike plenty of literatures focused on the United States and European countries, there have been very few studies that comprehensively analyze the performance of M&A.

In this analysis, empirical research is conducted on the performance of M&A in China. Acquiring and target listed companies of China are selected as samples during the period from 2013 to 2017. I measure the wealth effects of both acquiring and target firms by calculating the cumulative average abnormal returns (CAARs) in an event study, and then investigate the value drivers of acquiring and target abnormal returns.

The overall performance of M&A found in the analysis is not much different from foreign markets, with positive announcement effects for target firms as well less but still significant effects for acquiring firms. Moreover, the results suggest that different listing exchanges, payment methods, industries and targets’ attitudes can all influence M&A performance to different extents under different circumstances.

(5)

I. Introduction

Merger & acquisition (M&A) is a common external strategy to achieve a firm’s rapid growth, which can be especially effective when a firm encounters some bottlenecks in intrinsic

growth or the internal environments can no longer boost the firm’s internal growth. However, whether the widely adopted strategy could bring a positive influence on the acquiring or target firm and how the potential powerful effect could be achieved attract both scholars and practitioners. Ideally, M&A has an immeasurable influence on a `firm’s resource allocation perfection, industrial structure adjustment and operating efficiency enhancement.

Nevertheless, M&A is a double-edged sword — obviously not all takeovers can be successfully implemented, and there can be various material uncertainties during the takeovers or at the post-merger stage, which usually deviate from the acquiring and target firms’ original intentions. To identify whether both acquirers and target firms can benefit from M&A transactions and how different corporate characteristics and transaction features can affect the potential influence is a significant implication for the decision-makers. Regulators could lay down effective operating guidance as well as set out laws and regulations so that the original aim of the financial instrument could be realized and the financial market would be more efficient. Managers in listed companies could draw a more comprehensive blueprint before putting the takeover into practice which could directly obtain a significant cost reduction during both the purchasing process as well as the post-merger stage. Additionally, the analysis offers a reasonable and reliable methodology for investors to evaluate listed company’s purchases, supporting their investing and asset-allocation

decisions.

M&A has developed for many years in the United States and European. As a contrast, the first takeover occurring in China was in 1993. After that, the practice of M&A began to

(6)

widely spread over the Mainland China. Consequently, the previous studies of M&A in China suffered from the scarcity of data as well as the imperfect methodology. As a frequent case in developed countries, whether M&A could create value for the shareholders of either target or acquirer firms is still a very important question to answer. Furthermore, although there have been some Chinese scholars conducting several researches in the performance of M&A. Some of them focused only on the one party of the M&A. Some of them studied the M&A with financial accounting method. However, they hardly involved the event analysis and discussed about the factors which could affect the M&A performance. Due to the scarcity of the data, there were few scholars in China including the empirical researches about the motivation of the M&A. However, as the security market in China becomes more complete as well as mature, M&A activities occur more frequently than ever before. The increasing amounts of takeover events in China contributes to a more comprehensive and all-round analysis of M&A. Base on the analysis above, the researches on the M&A performance of Chinese listed companies have their realistic meaning and practical value. Thus, how to effectively analyze and measure the performance of a listed company’s M&A? What can contribute to the announcement effect? They both require theoretical and empirical explanations.

In this analysis, empirical research is conducted on the performance of acquisition

transactions in China. Acquiring and target listed companies of China are selected as samples during the period from 2013 to 2017. I measure the wealth effects of both acquiring and target firms by calculating the cumulative average abnormal returns (CAARs) in an event study, and then investigate the value drivers of acquiring and target abnormal returns.

(7)

depends on financial and accounting information disclosed by listed companies. Obviously, only taking the internal information into consideration while ignoring the market reaction may result in an upward bias and suffer from the management manipulations or even frauds. Thus, this research adopts the more strict and rigorous methodology of ‘event study’ by focusing on the price impact as it is captured by CAR. Secondly, the research studies the performances of both target and acquiring firms and digs into multiple potential factors. The methods of payment and the different attitudes the target firm shows to the bidders have been studied before in the United States and European markets. Nevertheless, what the difference in investors’ opinions and reactions for transactions between companies listed in different stock exchanges and for companies specialized in different industries still remain a question. This analysis can offer a rough implication for their corresponding firms when making a strategic purchasing decision and when implementing the takeover transactions.

After performing an overall analysis between the M&A short-term performance of listed acquirers and targets based on the sample during the period from 2013 to 2017, it shows that target firms can obtain an instant increase in their stock price by 3.33% in a very short window (2 trading days). When time passes, a relatively longer period, 20 trading days, can add extra wealth up to a level of 6.54% to the existing shareholders. However, the effect for acquiring firms is a bit weaker, only 2.23% instantly and 2.93% abnormal return generated in the follow month. In the univariate analysis, the results are shown as below: firstly,

companies listed in Shenzhen normally get a stronger market faith than Shanghai-listed companies, statistically, 2.89% for target firms and 3.51% for bidding firms and suffer less from confidential information leakage or insider trading. Secondly, the choice of cash payment can release a positive signal of targets to the market so that it benefits target firms much more than other payment methods with an abnormal return of 4.52%. However, this is

(8)

opposite for acquiring firms, all-cash payment decreases bidding firms’ CAAR by 2.79%. Thirdly, for target firms, non-industrial companies can earn around 3.01% abnormal return than industrial companies. Nevertheless, it is totally different for bidding firms, where

industry has no effect on the perceived market reaction. Last but not the least, the attitude that target firms hold to bidding firms shows its relatively long-term influence in the event

window of 20 trading days. A friendly takeover can in average earn an extra CAAR of 3.85% for target firms and 5.90% for bidding firms than neutral M&A.

The empirical results suggest that, for legislators and regulators, they should notice the fact that there is significant leakage of material nonpublic information and corresponding insider trading, and to realize their duties to perfect the financing mechanism and to ensure a fair market. For managers and directors in both listed targets and acquirers, they should take their characteristics of firms and transactions into account when planning a takeover. For example, an all-cash payment implied an overvaluation of bidders’ stock price, but is a positive news for target firms, which can strength investors’ belief in targets’ enterprise value, and in the relatively long-run (one-month), a friendly takeover can show its potential advantages in cost-reduction and post-merger integration, so a wise manager should try to reach the consensus between the two parties. For investors, the trends revealed from this analysis provide some extra supporting which can help investors to maximize their investment profits or capital gains, no matter he or she is an event-driven arbitrageur or a value investor.

This paper proceeds as below: Section I: Introduction. Section II: Literature review.

(9)

Section V: Performing empirical analysis of M&A short-term performance. Section VI: Conclusions and further discussion.

(10)

II. Literature review

Neoclassical theory [Manne (1965)] takes M&A as an activity that managers engage in to fulfill the maximization of shareholders’ wealth and then increase the company’s value. The reduction of cost, the reinforce of the monopoly as well as seeking alternates for low

efficiency management are the ways to increase the shareholders’ wealth.

Since 1960s, the amount of M&A activities is increasing dramatically. Whether M&A is able to create wealth for shareholders of both target and acquirer companies or not becomes a significantly focused top in the related area. There have already been many literatures

analyzing and researching the performance of M&A activities in the short-term as well as the long-term periods, such as the United States, Canada, Japan and some European countries.

In the analysis of the short-term performance of M&A, most scholars agree that M&A can get a relatively stable positive premium to the target (Franks and Harris 1989). However, in the analysis of the acquirers, the results are more sophisticated and vary quite differently. The empirical results of some scholars (Franks and Harris 1989, and Mulherin, Boone 2000) show that M&A cannot bring benefits to the acquirers which means that the returns are not obvious or even negative, while other scholars generated the opposite results (B. Espen Eckbo, Karin S. Thorburn 2000).

In the empirical research of Jensen and Ruback (1983), they found that the tender offer could bring 30% profit to shareholders, while there was only 20% profit in mergers. Franks and Harris (1989) analyzed over 1800 British companies from 1955 to 1985. They stated that there was 25%-30% CAR for targets while none for acquirers around the announcement date

(11)

targets from Canadian listed companies. Their research showed that, local acquirers gained significantly positive CAR in event window of 1 month and 2 days. On the opposite, the American acquirers didn’t profit in the M&A activities. Additionally, the profits of American acquirers are obviously lower than the profits of Canada acquirers. Mulherin and Boone (2000) selected 1305 sample companies in 59 industries announced M&A during 1990 – 1999 widely and used stock price changes to evaluate returns during the announcement period. The result showed that M&A announcements brought significantly positive returns and target firms gained 20.2% CAR during event window. However, the returns of the acquirers are slightly negative based on a insignificant confidence level. Furthermore, they concluded that the size of a firm had a direct effect on the M&A performance, which was consistent with synergies theory. Marc Georgen and Luc Renneboog (2004) studied the shareholders selected all the large-scale M&A in Europe markets from 1993 to 2000. After analyzing the short-term effect of the takeovers, they found that the announcements of the takeovers brought 9% profit to targets, and CAR reached 23% from 2 months before the announcements, while only 0.7% significant profit to acquirers. Additionally, they pointed out that M&A premium was related to the locations of targets and domestic M&A gained more short-term wealth effect than cross-border M&A.

In emerging Chinese markets, M&A activities emerged since 1990s while the scholars began paying attention to the M&A in recent years. Therefore, few scholars studied the performance of the M&A completely and systematically. Some of them only focused on the acquirers while ignored the targets (Lei, L. I., and Z. G. Song. 2009). Compared to event analysis, there are many scholars analyzing the long-term performance of M&A with financial accounting method (Lei, L. I., and Z. G. Song. 2009). However, some shcolars began to research the performance of M&A in Chinese listed companies base on the event analysis recently and

(12)

reached inconsistent results. For target companies, most scholars including Jian Gao, Xinwei Chen (2000), Lanyu Liang, Shanmin Li, Yugang Chen (2002) had a conclusion with

significantly positive cumulative abnormal returns. However, Zongxin Zhang, Lei Ji (2003), Yiwen Fei (2000) arrived at an opposite conclusion. They reach a conclusion with an

insignificant negative return. For acquirers, almost half of the literatures have positive returns. (Such as Xinghui Lei and Qi Zhang (2000); Wenzhang Zhang and Huihui Gu (2002)). However, there are also some researches which reached negative returns on the contrary.(such as Guang Yu and Rong Yang (2003); Xin Zhang (2003)). Xinyuan Chen and Tianyu Zhang (1999) studied 30 M&A events happening in 1997. They found that acquirers gained a increasing positive cumulative abnormal returns in the event window (-10, 20). However, it was insignificant when they tested the result. Guang Yu and Rong Yang (2000) collected M&A of 19 listed companies in Shanghai and Shenzhen security market from 1993 to 1995. The empirical research revealed that the shareholders of targets made positive cumulative abnormal returns (Listed companies had a 14.3% profit in Shanghai security market and a 5.71% profit in Shenzhen security market), compared to them, the acquirers hardly made positive cumulative abnormal returns. However, the result is affected by the limit of the sample amount and their studies were based on a small amount of samples.

(13)

III. Methodology

This analysis adopts the methodology of event study to analyze the short-term M&A

performance of China’s listed companies, by calculating and testing the cumulative average abnormal returns (𝐶𝐴𝐴𝑅s).

Event study is a very important financial analysis method usually exploited when analyzing the market response to an important event, in which the estimation window is determined to obtain the parameters of the stocks’ behavior and normal return (𝑁𝑅), event window is determined based on the announcement date of an M&A transaction, and abnormal return (𝐴𝑅) is calculated as the differences between the real returns and expected normal returns, cumulative abnormal return (𝐶𝐴𝑅) is used to measure the market reaction to the

announcement of the M&A transaction, and finally to evaluate M&A’s effect on the target firm or the acquiring firm’s performance.

Fama, Fisher, Jensen and Roll (1969) incorporated the methodology of event study in their study of the market behavior within the event window of the stocks’ event stripping. In their study, they compared the real return after stripping the event with the expected return assuming if without the occurrence of the event, and performed the hypothetical test to examine whether the abnormal return equals zero. Currently, the methodology of event study is widely adopted to analyze the behavior of stock prices in the event window of the M&A transactions’ announcements.

An event study analysis implied two basic assumptions: firstly, the capital market is efficient and its reaction to a certain event is unbiased; secondly, it can exclude the redundant effect from other factors in the related period.

(14)

Graph 1. Event Window and Estimation Window of an event study

Step 1: To begin with, the analysis determines the event window and estimation window for the whole announcement dates of M&A transactions. As shown in Graph 1, t = 0 is defined as the event date, the announcement date of China’s listed companies’ M&As.

The estimation window is [𝑇₁, 𝑇₂], which is deemed as a clean period free from multiple transactions’ interaction and usually located from 250 days to 30 days prior the

announcement date. During estimation window, the M&A announcement usually exerts no effect on the stock prices of both the target firms and acquiring firms, so that the clean period can be used to obtain the estimated return of the two parties without M&As involved. In this analysis, estimation window starts from 𝑇₁ = −250 (roughly one year) and ends at 𝑇₂ = −125 (roughly half a year), 6-month period ending 6 months prior to the event date.

The event window is [𝑡₁, 𝑡₂], which is usually centered around the corresponding event date and is used to calculate abnormal returns (𝐴𝑅_𝑖𝑡). In the following analysis, the shorter event window is [-2, 2], which represents the period from 2 trading days before the event date t = 0 to 2 trading days after the event date, totaling 5 trading days. The comparatively longer event window is [-20, 20], which represents approximately one month before/after the event date. In this analysis, different time horizons (event window) are considered: [-2, 0], [0, 2], [-2, 2], [-20, 0], [0, 20], and [-20, 20] (all in trading days).

(15)

Each M&A transaction has its own independent estimation window and event windows. It should be noticed that when a certain firm is involved in more than one M&A transaction in a short term, the estimation window and event windows of the two events will overlap, leading to a biased normal return estimation and abnormal return calculation. This analysis takes this into account when conducting sample screening.

Step 2: When estimating the normal returns (𝑁𝑅_𝑖𝑡) and abnormal returns (𝐴𝑅_𝑖𝑡), there are usually 3 applicable models: Mean-adjusted Model, Market Model and Capital Asset Pricing Model (CAPM). It is empirically proved (Stephen J. Brown and Jerold B. Warner, 1980) that the model requiring market information can get better results than those who do not.

In market model, the expect stock return is measured from the behaviors of the market index return.

𝑅_𝑖𝑡 = 𝛼_𝑖+ 𝛽_𝑖𝑅_𝑚𝑡+ 𝜀_𝑖𝑡 𝑁𝑅_𝑖𝑡 = 𝛼_𝑖 + 𝛽_𝑖𝑅_𝑚𝑡𝑧 𝐴𝑅_𝑖𝑡 = 𝑅_𝑖𝑡− (𝛼_𝑖+ 𝛽_𝑖𝑅_𝑚𝑡)

Step 3: For a certain company, this analysis defines abnormal return (𝐴𝑅_𝑖𝑡) is the difference between the actual daily returns (𝑅𝑖𝑡) and the expected returns (𝑁𝑅𝑖𝑡) obtained from market

models.

𝑅_𝑖𝑡 = ln(𝑃_𝑖𝑡+ 𝐷_𝑖𝑡) − ln⁡(𝑃𝑖𝑡)

𝐴𝑅_𝑖𝑡 = 𝑅_𝑖𝑡− 𝑁𝑅_𝑖𝑡

The last step is to test if market responses to listed companies’ M&A transactions and how it reacts. Assuming the sample size is N, the matrix of abnormal returns can be built as below:

(16)

𝐴𝑅_1,𝑡₁ … 𝐴𝑅_𝑁,𝑡₁ |⁡⁡⁡⁡ … ⁡⁡⁡⁡⁡| 𝐴𝑅_1,−1 … 𝐴𝑅_𝑁,−1 𝐴𝑅_1,0 … 𝐴𝑅_𝑁,0 𝐴𝑅_1,1 … 𝐴𝑅_𝑁,1 |⁡⁡⁡⁡ … ⁡⁡⁡⁡⁡| 𝐴𝑅_1,𝑡₂ … 𝐴𝑅_𝑁,𝑡₂

In this matrix, 𝐴𝑅𝑖𝑡 is defined as the abnormal return that company i obtained from the its

M&A transaction announced on date = t (𝑡 ∈ [𝑡1, 𝑡2]). Each row in the matrix represents the

abnormal returns that company i earns during the period, and each column represents the abnormal returns that N companies earn on a specific trading day.

To test whether the abnormal returns are significantly different from zero, cumulative

abnormal return (𝐶𝐴𝑅_𝑖) is built from accumulating all abnormal returns that company i earns during the event date [𝑡1, 𝑡2],

𝐶𝐴𝑅_𝑖 = ∑ 𝐴𝑅_𝑖𝑡

𝑡=𝑡2

𝑡=𝑡1

and used to measure the change in shareholders’ wealth, which is a most important indicator for a company’s performance.

The average of 𝐶𝐴𝑅𝑖 is

(17)

𝐻₀: 𝐸(𝐶𝐴𝑅_𝑖) = 0,

which implies that the M&A transactions has no impact on the listed companies’ stock return. The t-statistic constructed for 𝐶𝐴𝐴𝑅 is:

𝑇𝑆 = √𝑁 𝐶𝐴𝐴𝑅 𝑠(𝐶𝐴𝑅_𝑖)⁡

where cross-section sample variance of cumulative abnormal return (𝐶𝐴𝑅_𝑖) is:

𝑠(𝐶𝐴𝑅_𝑖) = √ 1

𝑁 − 1∑(𝐶𝐴𝑅𝑖− 𝐶𝐴𝐴𝑅)2

𝑁 𝐼=1

(18)

IV. Data and descriptive statistics

Sources of the sample data

In this analysis, all the data of the M&A transactions and the related information of the acquiring firms and target firms in each transaction are retrieved from Mergers & Acquisition Module in ThomsonONE database. Detailed information of the M&A transactions available in ThomonsONE includes the deal’s announcement data, status (completed or not), total value, payment method (the percentage of cash, stock, debt payment and mixture payment etc.) and the attitude of the acquiring firm towards the target firm (friendly or neutral), as well as the target or the acquiring firm’s company name, SEDOL ticker, industry, and their trading exchange (Shanghai or Shenzhen, excluding Hong Kong Exchange here).

The stocks’ daily prices of the firms’ can be directly retrieved from the Compustat Capital IQ Module in the Wharton Research Data Services (WRDS). This database provides all stocks’ daily prices all around the world, and in this analysis, the daily closing prices for related stocks in the M&A transactions are used to calculate their stock returns. However, what is worthwhile notice is that the dates of the stock prices available is not totally consistent with the trading days in Chinese market. Taking this into account, additional attention should be paid when processing the stock price data.

The data of Chinese market return required in the market model is retrieved from the CSMAR database, which is a professional domestic data service provider focusing on the Chinese capital market. In this analysis, the highly representative benchmark, CSI300 index (Shanghai-Shenzhen 300 index, which is composed of the selected 300 stocks currently

(19)

China, so the M&A transaction information retrieved from ThomsonONE can be cross-referenced with CSMAR.

In case of unanticipated cross-reference, some detailed information in Wind database is exploited to support the data creditability and to supplement the missing data is other databases. Wind is a widely-accepted and well-renowned domestic financial data service provider in China.

Sample screening procedures

In the process of data screening and sample selection, there is a set of criteria in this analysis that the samples should meet, listed as below:

Firstly, to have an overall analysis of the short-term performances of listed companies’ M&As in recent years, only the M&A transactions occurred during the period from January 1, 2013 to December, 31 2017, signaled with the announcement date of each transaction. In the following event study, announcement date is the event date.

Secondly, to test the abnormal returns, it is important that the data of stock prices and the financial conditions of both the target and acquiring firms in a M&A transaction is available. Given the searching mechanism in ThomsonONE when retrieving the M&A transactions’ data, this analysis only selects the samples of transaction in which at least one party of either the target firm or the acquiring firm is listed in Shanghai or Shenzhen exchange.

Thirdly, to strengthen the representativeness of the M&A transaction samples, this analysis tends to select only deals with a comparatively large transaction value. Considering all the transaction values retrieved from ThomsonONE are measured in U.S. dollar, $1 million is chosen as the starting point to exclude the comparatively small transactions.

(20)

To expel the potentially opposite effect of M&A transactions that didn’t eventually succeed on the firms’ short-term performance, only samples of completed deals are chosen.

Financial firms are usually subject to special laws and regulations, and usually tend to take different strategies from normal industries, due to their special characteristics and roles in the society. To avoid financial industry’s special performance, target firms or acquiring firms which are classifies as financial industrial are excluded from the samples.

Finally, it is common that a specific listed company involved in more than one M&A transaction. If the announcement dates of two M&A transactions in which a specific target firm or acquiring firm is involved in are too close, it is likely that they interact with each other, or the former transaction will cause a deviation in the estimation results in the estimation window of the latter one, so that both will bias the result of a single M&A transaction’s effect on short-term performance. To avoid both the two kinds of overlapping effect in the event study analysis, the latter sample should be deleted if the interval is less than 250 trading days (approximately one year), and both the two samples of transaction should be deleted if the interval is less than 60 trading days.

Data Descriptions

According the screening criteria listed above, the total qualifies sample can be obtained.

Part 1: An overview of acquiring firms and target firms

Table 1a roughly describes the results of sample screening. In this analysis, the sample of acquiring firms consists of 1,021 transactions and the sample of target firms consists of 500 transactions during the period from January 1, 2013 to December, 31 2017. To acquiring

(21)

Table 1a. Year distributions of acquiring and target firms

This table shows the year distributions of acquiring and target firms, including the numbers and corresponding percentage of each year from 2013 to 2017.

Acquirers Targets Year N percentage (%) N percentage (%) 2013 184 18.02 50 10.00 2014 193 18.90 100 20.00 2015 288 28.21 162 32.40 2016 186 17.83 109 21.80 2017 174 17.04 79 15.80 Observations 1,021 500

Part 2: Samples of acquiring firms and target firms classified by listing exchanges

The acquiring and target firms are classified based on their different stock exchanges: Shanghai Stock Exchange and Shenzhen Stock Exchange. The information of exchanges where listed companies are traded is provided from ThomsonONE. From Table 1b., it is shown that a two-third of the sample companies are listed in Shanghai stock exchange.

Table 1b. Trading exchange of acquirer and target firms

This table shows the trading exchange distributions of acquiring and target firms, including the numbers and corresponding percentage of Shanghai or Shenzhen stock exchange.

Acquirers Targets

Exchange N percentage (%) N percentage (%)

Shanghai (SH) 677 66.31 326 65.20

Shenzhen (SZ) 344 33.69 174 34.80

Observations 1,021 500

Part 3: Samples of acquiring firms and target firms classified by different industries

The acquiring and target firms are classified into twelve different industries: Consumer Products and Services, Energy and Power, Healthcare, High Technology, Industrials,

(22)

Materials, Media and Entertainment, Real Estate, Retail, Consumer Staples,

Telecommunications and Finance. As financial firms are usually subject to special laws and regulations, and usually tend to take different strategies from normal industries, due to their special characteristics and roles in the society, target firms or acquiring firms which are classifies as financial industrial are excluded from the samples. From Table 1c., it is shown that the top three industries that account for the sample of acquiring and target firms are Materials, Industrials and Real Estate.

Table 1c. Industry distribution of acquirer and target firms

This table shows the industry distribution of acquiring and target firms, including the numbers and corresponding percentage of their categories. The eleven industries are: Consumer Products and Services, Energy and Power, Healthcare, High Technology, Industrials, Materials, Media and

Entertainment, Real Estate, Retail, Consumer Staples and Telecommunications. When performing the univariate factor analysis of the M&A’s effect on a firm’s short-term performance, firms of Energy and Power, High Technology, Industrials and Materials are grouped together into the “Industrial Group” and others into the “Nonindustrial Group”

Acquirers Targets

Industry N percentage (%) N percentage

(%) CPS, Consumer Products and

Services

45 4.41 21 4.20

ENERGY, Energy and Power 95 9.30 47 9.40

HEALTH, Healthcare 75 7.35 45 9.00

HT, High Technology 90 8.81 43 8.60

IND, Industrials 234 22.92 75 15.00

MATERLS, Materials 207 20.27 104 20.80

MEDIA, Media and Entertainment

35 3.43 16 3.20

REALEST, Real Estate 98 9.60 57 11.40

RETAIL, Retail 44 4.31 31 6.20

(23)

TELECOM, Telecommunications 19 1.86 6 1.20 — Industrial Subgroup 626 61.30 269 53.80 — Nonindustrial Subgroup 395 38.70 231 46.20 Observations 1,021 500

When performing the univariate analysis of different industries, firms of Energy and Power, High Technology, Industrials and Materials are grouped together into the “Industrial Group” and others into the “Nonindustrial Group”.

Part 4: Samples of acquiring firms and target firms classified by different transaction payment methods

For an M&A transaction, the decision of payment method usually requires taking many financial, economic and legal factors into accounts and are subjected to strict regulations. The availability of funds, costs of capital, time matching and professional knowledge all limits a firm’s M&A payment method. According to the ThomsonONE database, this analysis classifies those transactions in which cash payment accounts for more than 90% of total payment into “cash payment” group, and leaves the others into “others” group. From Table 1d., it is shown that there is a significant imbalance between the cash payment method and other methods. About 65% of the M&As that the target firm is listed in Chinese stock exchange use cash to finance the acquisition, while only about 20% of the listed acquiring firms fund with almost 100% cash. The condition might reflect the situation that financing tools have been more diversified than before in China.

Table 1d. Payment methods of acquirer and target firms

This table shows the payment method distribution of acquiring and target firms, including the numbers and corresponding percentage of their categories. The transactions in which cash payment accounts for more than 90% are considered to be paid by cash.

(24)

Payment methods N percentage (%) N percentage (%) Cash 208 20.37 325 65.00 Others 813 79.63 175 35.00 Observations 1,021 500

Part 5: Samples of acquiring firms and target firms classified by different transaction attitudes

Due to the special and limited financial regulations, hostile acquisitions rarely occur in Chinese capital market. From Table 1e., it is obvious that for the listed acquiring firms, almost all M&As are friendly, however, for the listed target firms, only about 70% transactions are friendly.

Table 1e. Transaction attitudes of acquirer and target firms

This table shows the transaction attitude distribution of M&As, including the numbers and corresponding percentage of their categories. A friendly M&A represents that the acquiring firm holds a friendly attitude towards the target firms.

Acquirers Targets

Attitude N percentage (%) N percentage (%)

Friendly 986 96.57 347 69.40

Neutral 29 2.84 148 29.60

Not applicable 6 0.59 5 1.00

(25)

V. Short-term Performance Empirical Analysis

In this section, I examine the responsiveness of stock abnormal returns to acquisition

announcement and compare that for acquiring and target firms at various time horizons. Then I perform the univariate analysis of the effect of one specific corporate characteristic or transaction nature on firms’ abnormal returns at a time.

Part 1: Target firms versus acquiring firms

Panel A of Table 2 shows that for the target firms, the event of M&A announcement brings 3.53% cumulative average abnormal return (CAAR) for shareholders during the event window [-2, 2]. To be specific, the target firm gains 3.33% CAR in average just in two trading days following the announcements, and from two trading days before the event, the estimated CAAR is 1.66%, which is just a half of 3.33%. Abnormal returns gained in the period [-2, 2] tend to be slightly larger than that of period [0, 2], which might result from a potential information leakage or rumors. All of the CAARs mentioned above are statistically significant at a 99% confidence level. It is obvious that the increase in stockholders’ wealth is anticipated at least two trading days before the official announcement, however it seems to be mitigated in terms of longer horizon [-20, 0] (20 trading days) with an abnormal return of 1.38% (significant at 1% level). For the 2-month window centered around the event date, there is a substantial positive abnormal return of 6.46%, doubling that of the 5-trading-day window. Comparing the results for the event window [-20, 0], [0. 20] and [-20, 20], it can be inferred that transaction information leakage and insider trading contribute little to abnormal returns.

Panel B of Table 2 focuses on acquiring firms. The situation of acquiring firms is

(26)

Table 2. Cumulative Average Abnormal Returns (CAARs) for Target and Acquiring Firms

This table shows cumulative abnormal average returns (CAARs) measured over several event windows for target and bidder firms. *, **, and *** indicate significance at 10%, 5%, and 1%, respectively.

Panel A: Target firms Panel B: Acquiring firms

Event window CAAR (%) t-value N CAAR (%) t-value N

[-2, 0] 1.6592*** (0.3714) 4.4673 477 3.7822* (0.2083) 1.8161 933 [0, 2] 3.3333*** (0.5285) 6.3066 475 2.2326*** (0.3134) 7.1244 933 [-2, 2] 3.5323*** (0.1265) 6.0863 475 1.9534*** (0.3293) 5.9318 933 [-20, 0] 1.3840* (0.1619) 1.8669 477 -0.6256* (0.4552) -1.3742 933 [0, 20] 6.5448*** (0.2155) 6.5714 468 2.9258*** (0.8623) 3.3932 919 [-20, 20] 6.4605*** (0.2493) 5.6068 468 1.6205** (0.9217) 1.7582 919

(27)

window [-2, 0], 2.23% on [0, 2] to 1.95% on [-2, 2]. Though the impact of M&A

announcement on market reaction to acquiring firms before announcement date is much larger than that to targets by 2.12% increase, it weakens quickly, which poses an opposite trend. In the longer horizon, acquiring firms normally suffer from a loss in abnormal returns of 0.63%, but still gain 2.93% and 1.62% separately from period [0, 20] and [-20, 20].

According to the significant levels above, it is likely that abnormal returns in [-20, 0] is much more volatile than those in [0, 20] and decrease abnormal returns as well as predictability in the full period [-20, 20].

The empirical result in this analysis shows that the announcements of M&As can bring significantly positive abnormal returns to both target firms and acquiring firms instantly and can substantially lift the abnormal returns up to 6.50% to targets over a longer horizon. However, the implication for acquiring firms is ambiguous. Compared with American or European capital market, M&As in China seem to function much weaker to stir up investors’ faith in the corporate strategies.

Part 2: Univariate analysis of short-term performance, by relating the cumulative abnormal returns (CARs) to one specific bid/corporate characteristics at a time.

Factor 1: Based on different stock exchanges

This section aims to analyze the effect of M&A on firms listed in Shanghai or Shenzhen Stock Exchange. As Hong Kong Stock Exchange is operated under a more mature and advanced financial framework, this analysis excludes companies listed in Hong Kong Stock Exchange. Generally speaking, there are no obvious differences between the operation and regulation system between Shanghai Stock Exchange and Shenzhen Stock Exchange.

(28)

Table 3a: Cumulative Average Abnormal Returns (CAARs) for Target and Acquiring Firms Based on Different Stock Exchanges

(1) Shanghai Exchange (2) Shenzhen Exchange (3) Difference

CAAR (%) t-value N CAAR (%) t-value N CAAR (%) t-value

Panel A: Target firms

[-2,0] 1.6859*** 3.4188 310 1.6096*** 2.9912 167 0.0763 -0.0469 [0,2] 2.9404*** 4.3940 309 4.0645*** 4.7401 166 -1.1241 -1.0728 [-2,2] 3.2776*** 4.4197 309 4.0064*** 4.3310 166 -0.7288 -0.7007 [-20,0] 1.3889* 1.4641 310 1.3750 1.1654 167 0.0139 0.0120 [0,20] 5.5321*** 4.5158 304 8.4218*** 4.9412 164 -2.8897* -1.3865 [-20,20] 5.5254*** 3.9676 304 8.1940*** 4.0252 164 -2.6686 -1.1049

Panel B: Bidding firms

[-2,0] 1.7961 0.7041 603 0.7411** 2.0618 330 1.0550* -1.3132

[0,2] 1.5372*** 4.1858 603 3.5034*** 6.1170 330 -1.9662*** -2.9722

[-2,2] 1.3605*** 3.5207 603 3.0368*** 5.0350 330 -1.6763*** -2.3997

[-20,0] -0.6657 -1.2359 603 -0.5523 -0.6650 330 -0.1134 -0.1165

(29)

However, on account of the government’s historical preferences, corporates of larger size tend to go public in Shanghai Stock Exchange and those who are relatively smaller prefer to choose Shenzhen Stock Exchange.

Panel A of Table 3a shows that for target firms listed in Shanghai or Shenzhen Stock Exchange, market reactions to M&A transactions are very similar. In the periods deviating from the event date by 2 trading days, targets’ shareholders benefit a lot from the M&A announcement with an estimated CAAR of more than 1% in window [-2, 0], around 3% (Shanghai) and 4% (Shenzhen) separately in the other two event windows. Again, the investors react less strongly from twenty trading days prior to announcements than that over [-2, 0], and abnormal returns gained over [0, 20] and [-20, 20] is almost twice the gains over [0, 2] and [-2, 2].

For acquiring firms listed in Shanghai, the abnormal returns pattern is very similar to that of full sample with a decreasing but positive impact on shareholders’ wealth over the shorter event windows (± 2 trading days) and a negative abnormal return of -0.67% over [-20, 0]. It can be inferred from Table 3 that bidding firms listed in Shenzhen tend to get larger abnormal return after the events and lose less before them.

When looking at column (1) and (2), it is obvious that companies listed in Shenzhen normally perform better than Shanghai listed companies and suffer less from confidential information leakage or insider trading. What puts forward a potential explanation is firm size. It is the fact that small- and medium-size companies are more inclined to go public in Shenzhen, of which a further study on this can be investigated.

(30)

Table 3b: Cumulative average abnormal returns (CAARs) for target and bidding firms based on different payment methods

This table shows cumulative abnormal average returns (CAARs) measured over several event windows for target and bidder firms by payment methods. *, **, and *** indicate significance at 10%, 5%, and 1%, respectively.

(1) Cash Payment (2) Other Payment (3) Difference

[-2,0] 2.2758*** 5.3744 310 0.5146 0.7291 167 1.7612** 2.3340 [0,2] 4.3799*** 7.2875 308 1.4029* 1.4014 167 2.9770*** 2.7334 [-2,2] 4.7118*** 6.8519 308 1.3569* 1.3065 167 3.3549*** 2.8229 [-20,0] 1.4773* 1.6316 310 1.2109 0.9374 167 0.2664 0.1695 [0,20] 8.1384*** 6.4802 303 3.6183** 2.2453 165 4.5201** 2.1775 [-20,20] 7.6584*** 5.1291 303 4.2608*** 2.4049 165 3.3976* 1.4100

[-2,0] 0.2933 0.5964 193 0.4004** 1.7461 740 -0.1071 -0.1905

[0,2] 1.3326** 1.9667 193 2.4674*** 6.9876 740 -1.1348* -1.4825

[-2,2] 1.2308** 1.7426 193 2.1419*** 5.7573 740 -0.9111 -1.1390

(31)

Whether adopting different payment methods can affect shareholders’ wealth is another important research field on M&A. In this section, the effect of acquiring firms’ decision of payment method on short-term performance is examined.

Strong evidence is observed in Table 3b that the stock price reaction for target firms is more sensitive to the purchase payment method. Cash offers often strength investors’ confidence in target firms and can trigger higher abnormal returns (4.38% in two trading days and 8.14% in twenty trading days) than other offers. However, Panel B of Table 3b shows that over both shorter and longer event windows, means of other payments encourages investors more effectively than cash payment. Column (3) displays that investors greet target firms in all-cash-funded acquisitions more favorably than those in all-debt-funded, all-equity-funded or mixed-paid transactions, with a gap of 2.98% CAAR in two trading days and of 4.52% in twenty.

The underlying explanatory theory of investors’ perspectives is signal theory. Several years ago, constrained by limited financing vehicles and strict regulation by government, most Chinese companies can only use cash to fund their takeovers. With the gradual progress of financial environment, more diversified instruments are available to qualified companies to raise capital. If the acquiring firm’s management regards the current market value of their own shares is overvalued, it will have a preference towards all-cash financing, so that future changes in their share prices will only benefit bidding firm’s shareholders. Conversely, if the managers in an acquiring firm believe their shares to be undervalued, equity offer is

preferred. Thus, the signal theory illustrates how information asymmetry between public investors and bidder’s management when evaluating the bidder’s corporate value may have a

(32)

potential relationship with the choice between cash or debt or equity payments in an offer. Meanwhile, we cannot ignore the prohibitive function of laws and regulations.

Factor 3: Based on different industries

This section studies how corporates in different kinds of industries gain or loss from M&As, and further discuss whether industry is an important driving factor of either target firms or acquiring firms’ short-term performance.

According to the industry distribution of total sample firms, this section divides eleven industries into 2 categories: industrial group, including Energy and Power, High Technology, Industrials and Materials, and nonindustrial group, including Consumer Products and

Services, Healthcare, Media and Entertainment, Real Estate, Retail, Consumer Staples and Telecommunications.

In Panel A of Table 4c, it poses a normal pattern as that of the total sample in Table 2. Target firms achieved positive abnormal returns during the longer scenario twice as much as the benefits within 2 trading days after the event. Announcement effect is stronger for nonindustrial targets at 1% significant level in the shorter period.

Nevertheless, it seems not the same picture for acquiring firms. All differences shown in Column (3) in Panel B of Table 4c are not statistically different from zero, which indicates industry factor does not play a crucial role in determining the wealth of shareholders in bidders.

The possible reason for the results might be that in recent years, China has experienced a great progress in information technology development and innovations, especially in

(33)

Internet-plus industry and electronic commerce. Many emerging entrepreneurs are taking advantage of these waves and earning incredibly high expectation from investors.

Factor 4: Based on different types of M&A

A friendly takeover refers to a transaction in which both the target and the acquiring firm reach an agreement. There are usually fewer confrontational situations and less frictional costs during the acquisition progress.

In the Panel A of Table 3d, it is striking that neutral takeovers seem to bring quite less benefit to shareholders of target firms. Obviously, there is no surprise in terms of friendly M&As, but over a longer scenario (about one month), targets in neutral deals underperform by a CAAR of approximately 4% (at 1% significant level).

As for bidding firms, the implication is ambiguous. Except for event window [-20, 0] (with a negative CAAR of -0.72% at 10% significant level), all CAARs achieved from friendly transactions to acquiring firms are positive with high confident levels. Surprisingly, for the first time, acquiring firms suffer from losses within 20 trading days after announcements. The differences between friendly and neutral takeovers are 5.90% over window [0, 20] and 3.58% over window [-20, 20] accordingly.

The facts lead us to the conclusion that how target companies are treated (friendly or neutrally) might have a substantial impact in a relatively longer run (one month), and shareholders of acquiring firms lose more than those of targets. The negative power of takeover friction in post-merger integration cannot be neglect, if the management and shareholders hope to realize the supposed synergy effect from M&As.

(34)

Table 3c: Cumulative average abnormal returns (CAARs) for target and bidding firms based on different payment methods

This table shows cumulative abnormal average returns (CAARs) measured over several event windows for target and bidder firms by sub-groups of industries. *, **, and *** indicate significance at 10%, 5%, and 1%, respectively.

(1) Industrial Companies (2) Nonindustrial Companies

(3) Difference

[-2,0] 0.9326*** 1.7057 259 2.5225*** 5.2199 218 -1.5899** -2.2416 [0,2] 2.1243*** 2.7834 258 4.7706*** 6.7450 217 -2.6463*** -2.5508 [-2,2] 2.0215*** 2.4656 258 5.3285*** 6.6633 217 -3.3070*** -2.9309 [-20,0] 1.5862* 1.4650 259 1.1437 1.1546 218 0.4425 0.2995 [0,20] 5.1541*** 3.6647 255 8.2096*** 5.8947 213 -3.0555* -1.5306 [-20,20] 5.7528*** 3.5634 255 7.3078*** 4.4627 213 -1.5550 -0.6714

[-2,0] 0.3604* 1.2534 571 0.4063* 1.4133 362 -0.0459 -0.0643

[0,2] 2.2578*** 5.5812 571 2.1931*** 4.4225 362 0.0647 0.0505

[-2,2] 1.9484*** 4.5742 571 1.9613*** 3.7765 362 -0.0129 -0.0730

(35)

Table 3d: Cumulative average abnormal returns (CAARs) for target and bidding firms based on different types of M&A

This table shows cumulative abnormal average returns (CAARs) measured over several event windows for target and bidder firms by different attitudes that the bidding firms hold towards the target firms. *, **, and *** indicate significance at 10%, 5%, and 1%, respectively.

(1) Friendly (2) Neutral (3) Difference

[-2,0] 1.5075*** 3.2201 334 1.7886*** 3.0749 137 -0.2811 -0.3091 [0,2] 3.1212*** 4.6396 334 3.7781*** 4.6151 136 -0.6569 -0.6072 [-2,2] 3.3554*** 4.5747 334 3.7078*** 4.0654 136 -0.3524 -0.3269 [-20,0] 1.2576 1.4078 334 1.4065 1.0486 138 -0.1489 -0.0923 [0,20] 7.7573*** 6.4870 328 3.9032** 2.1444 135 3.8541** 1.7512 [-20,20] 7.7538*** 5.7339 328 3.3909* 1.5102 135 4.3629** 1.7078

[-2,0] 0.3801** 1.7768 900 -0.2821 -0.3128 28 0.6622 0.5346 [0,2] 2.2570*** 7.0169 900 0.9820 0.6953 28 1.2750 0.6971 [-2,2] 1.9738*** 5.8420 900 0.5787 0.3757 28 1.3951 0.7261 [-20,0] -0.7174* -1.5448 900 1.2051 0.5255 28 -1.9225 -0.7209 [0,20] 3.0613*** 3.4399 887 -2.8341 -1.2755 27 5.8954 1.1523 [-20,20] 1.6637** 1.7558 887 -1.9170 -0.5361 27 3.5807* 1.6552

(36)

Part 3: Robustness Check

The analysis in Section V is derived from the market model based on an estimation window of [-250 trading days, -125 trading days], approximately a 6-month window half a year prior to the announcement date. Thus, in the robustness check section, I change the estimation window from [-250 trading days, -125 trading days] to [-150 trading days, -25 trading days], which is still a 6-month window, while only 1-month prior to the announcement date. The results for robustness check are shown in the Appendix, from which we can see that the conclusion remains unchanged, but a bit stronger in the degree. We can roughly take the conclusion drawn from above for an effective and robust one.

(37)

VI. Conclusions and further discussions

When performing an overall analysis between the M&A short-term performance of listed acquirers and targets, significant announcement effect can be observed. In another word, based on the sample during the period from 2013 to 2017, target firms can obtain an instant increase in their stock price by 3.33% in a very short window (2 trading days). When time passes, a relatively longer period, 20 trading days, can add extra wealth up to a level of 6.54% to the existing shareholders. However, the effect for acquiring firms is a bit weaker, only 2.23% instantly and 2.93% abnormal return generated in the follow month. Obviously, more investors’ sentiment is incorporated into the market information, and listed target firms perceive much more market faith in their operating and strategies.

Compared with the previous cross-sectional studies in foreign countries, e.g., 30% tender offer brought to shareholders (Jensen and Ruback, 1983) and 20.2% CAR for target firm in Mulherin, Boone’s study, the M&A’s premium effect is quietly weak with a level of 6.54% for target firms. Since different scholars have not arrived at an agreement on the

announcement effect for acquiring firms, here I skip the comparison between Chinese market and foreign markets.

In the univariate analysis part of Section V, there are some interesting phenomena. Firstly, the listing stock exchange (Shanghai or Shenzhen) does have an influence on the listed companies’ M&A performance. Companies listed in Shenzhen normally perform better than Shanghai listed companies and suffer less from confidential information leakage or insider trading.

Secondly, the choice of cash payment can release a positive signal of targets to the market so that it benefits target firms much more than other payment methods. However, this is

(38)

opposite for acquiring firms. The market reaction corresponds to the ‘signal theory’ in M&A, in which if the management in an acquiring firm believe its outstanding public shares to be overvalued, they would prefer an all-cash financing.

Thirdly, for target firms, non-industrial companies can earn around 3% abnormal return than industrial companies. Nevertheless, it is totally different for bidding firms, where industry has no effect on the perceived market reaction.

Last but not the least, the attitude that target firms hold to bidding firms shows its relatively long-term influence in the event window of 20 trading days. It is consistent with people’s common sense that a friendly takeover can realize much more synergy, probably resulting from the lower post-merger cost and better operating integration.

When investigating the potential factors that might influence the accomplishment of takeovers, it is hard to identify how deep we should dig into it. Stock exchange is a

characteristic of either the acquiring firm or target firm, however it represents far more than just where the listed firms’ stocks are traded. Due to some historical and regulatory

considerations in China, the listing exchange might imply a firm’s size, specialization or governmental preferences. However, it would be extremely complicated if further digging into the very intrinsic and fundamental forces, probably out of the scope of finance. Moreover, the explanations for these results obtained in this analysis must rely on the soundness of some theories or hypothesis, for example, the very basic ‘Efficiency Market Hypothesis’ and ‘Signal Theory’, etc. The related theories can hold in some mature and developed markets, however, they might fail in Chinese markets. This would directly lead to a misunderstanding of the results above, thus further testing and study should be performed to

(39)

A take way of this thesis is that future research should investigate the very intrinsic and fundamental power that can influence how public investors perceive a firm’s M&A

transaction and should widen and explore other various potential domestic-specific factors so that a comprehensive model can be built.

(40)

Appendix: Robustness check

Table 4. Cumulative Average Abnormal Returns (CAARs) for Target and Acquiring Firms

Panel A: Target firms Panel B: Acquiring firms

Event window CAAR (%) t-value N CAAR (%) t-value N

[-2, 0] 1.7000*** (0.0800) 4.6535 480 0.5410*** (0.0634) 2.6469 961 [0, 2] 3.4685*** (0.1164) 6.5167 478 2.3191*** (0.0031) 7.4811 961 [-2, 2] 3.7548*** (0.1274) 6.4422 478 2.1519*** (0.1032) 6.4660 961 [-20, 0] 2.3418*** (0.1664) 3.0837 480 -0.3973 (0.1508) -0.8169 961 [0, 20] 7.2790*** (0.2204) 7.1691 471 3.0660*** (0.2673) 3.5281 946 [-20, 20] 8.2423*** (0.2618) 6.8333 471 1.9872** (0.0099) 2.0050 946

(41)

Table 5a: Cumulative Average Abnormal Returns (CAARs) for Target and Acquiring Firms Based on Different Stock Exchanges

(4) Shanghai Exchange (5) Shenzhen Exchange (6) Difference

[-2,0] 1.7290*** 3.5865 312 1.6462*** 3.0574 168 0.0828 -0.0523 [0,2] 3.0355*** 4.5046 311 4.2748*** 4.9537 167 -1.2393** -2.0219 [-2,2] 3.4596*** 4.6667 311 4.4044*** 4.5888 167 -0.9448 * -1.7628 [-20,0] 2.3097*** 2.3424 312 2.4014** 2.0566 168 -0.0917 -0.2479 [0,20] 6.0492*** 4.8600 306 9.5598*** 5.4818 165 -3.5106*** -7.5893 [-20,20] 7.0326*** 4.8038 306 10.4857*** 4.9636 165 -3.4531*** -8.7648

[-2,0] 0.2412 0.9739 629 1.1092*** 3.0902 332 -0.8680*** -3.5427 [0,2] 1.6232*** 4.4844 629 3.6373*** 6.3537 332 -2.0141*** -5.3124 [-2,2] 1.4633*** 3.7367 629 3.4564*** 5.6768 332 -1.9931*** -6.3107 [-20,0] -1.0399** -1.7108 629 0.8202 1.0168 332 -1.8601*** -8.3453 [0,20] 1.6377* 1.5733 616 5.7322*** 3.6967 330 -4.0945*** -10.3135 [-20,20] 0.2139 0.1751 616 5.2973*** 3.1487 330 -5.0834*** -15.6944

(42)

Table 5b: Cumulative average abnormal returns (CAARs) for target and bidding firms based on different payment methods

This table shows cumulative abnormal average returns (CAARs) measured over several event windows for target and bidder firms by payment methods. *, **, and *** indicate significance at 10%, 5%, and 1%, respectively.

(4) Cash Payment (5) Other Payment (6) Difference

[-2,0] 2.3648*** 5.7571 312 0.4654 0.6607 168 1.8994*** 4.4285 [0,2] 3.5915*** 7.5454 310 1.3963 1.3975 168 2.1952*** 5.0511 [-2,2] 5.0373*** 7.2792 310 1.3881 1.3395 168 3.6492*** 6.8424 [-20,0] 2.8181*** 3.0913 312 1.4573 1.0730 168 1.3608*** 3.9141 [0,20] 9.2750*** 7.2644 305 3.6117** 2.2031 166 5.6633*** 12.3166 [-20,20] 10.2163*** 6.6170 305 4.6153*** 2.4441 166 5.6010*** 14.2890

[-2,0] 0.3631 0.7509 197 0.5869*** 2.6086 764 -0.2238 -0.7365 [0,2] 1.3008** 1.9602 197 2.5816*** 7.3770 764 -1.2808*** -2.9177 [-2,2] 1.2820* 1.8424 197 2.3762*** 6.2860 764 -1.0942*** -3.0088 [-20,0] 0.9139 0.9538 197 -0.7354 -1.3149 764 1.6493*** 6.2776 [0,20] 0.8249 0.5658 196 3.6517*** 3.5552 750 -2.8268*** -6.0549 [-20,20] 1.3574 0.8080 196 2.1518* 1.8376 750 -0.7944*** -2.0900

(43)

Table 5c: Cumulative average abnormal returns (CAARs) for target and bidding firms based on different payment methods

This table shows cumulative abnormal average returns (CAARs) measured over several event windows for target and bidder firms by sub-groups of industries. *, **, and *** indicate significance at 10%, 5%, and 1%, respectively.

(4) Industrial Companies (5) Nonindustrial Companies

(6) Difference

[-2,0] 1.0048* 1.8533 259 2.5147*** 5.3428 221 -1.5099*** -3.7549 [0,2] 2.2849*** 2.9933 258 4.8564*** 6.7193 220 -2.5715*** -4.2678 [-2,2] 2.2262*** 2.7017 258 5.5473*** 6.9019 220 -3.3211*** -6.5760 [-20,0] 2.6075*** 2.3760 259 2.0304** 1.9618 221 0.5771* 1.7505 [0,20] 5.8711*** 4.0421 255 8.9412*** 6.4064 216 -3.0701*** -6.9287 [-20,20] 7.5208*** 4.4867 255 9.0940*** 5.2417 216 -1.5732*** -4.1619

[-2,0] 0.5392* 1.9346 586 0.5440* 1.8660 375 -0.0048 0.0502 [0,2] 2.3256*** 5.8417 586 2.3089*** 4.6673 375 0.0167 -0.0372 [-2,2] 2.1410*** 5.0745 586 2.1690*** 4.0039 375 -0.0280 -0.2053 [-20,0] -0.3573 -0.6047 586 -0.4597 -0.5485 375 0.1024 0.4583 [0,20] 2.9980*** 2.7963 579 3.1733** 2.1580 367 -0.1753 -0.4652 [-20,20] 1.9607* 1.6171 579 2.0290 1.1966 367 -0.0683 -0.2376

(44)

Table 5d: Cumulative average abnormal returns (CAARs) for target and bidding firms based on different types of M&A

This table shows cumulative abnormal average returns (CAARs) measured over several event windows for target and bidder firms by different attitudes that the bidding firms hold towards the target firms. *, **, and *** indicate significance at 10%, 5%, and 1%, respectively.

(4) Friendly (5) Neutral (6) Difference

[-2,0] 1.5032*** 3.2445 334 1.9037*** 3.4360 141 -0.4005 -0.8196 [0,2] 3.0956*** 4.5850 334 4.2612*** 5.1062 139 -1.1656* -1.8016 [-2,2] 3.3594*** 4.5617 334 4.4072*** 4.8456 139 -1.0478* -1.9476 [-20,0] 2.0335** 2.1887 334 2.5384* 1.9486 141 -0.5049 -1.4094 [0,20] 8.1215*** 6.5867 328 5.3217*** 3.0100 138 2.7998*** 5.7818 [-20,20] 8.9572*** 6.1456 328 6.1160*** 2.8589 138 2.8412*** 6.9485

[-2,0] 0.5533*** 2.6353 927 -0.3419 -0.3879 29 0.8952 1.2842

[0,2] 2.3356*** 7.3374 927 1.3265 0.9712 29 1.0091 0.9714

[-2,2] 2.1711*** 6.3507 927 0.8694 0.5992 29 1.3017 1.5053

[-20,0] -0.4988 -0.9998 927 1.7634 0.9289 29 -2.2622*** -3.6498

(45)

Appendix: References

Andrade, G., M. Mitchell and E. Stafford. New Evidence and Perspectives on Acquisitions[J]. Journal of Economic Perspectives, 2001, 15, pp. 103-20

B. Epson Eckbo and Karin S. Thorburn. Gains to Bidder Firms Revisited: Domestic and Foreign Acquisitions in Canada[J]. Journal of Financial and Quantitative Analysis. 2000, 35(1), pp. 1-25

Brown, S. and J. Warner. Using Daily Stock Returns: The case of Event Studies[J]. Journal

of Financial Economics, 1985, 14, pp. 3-13

Duggal, R., & Millar, J. Institutional Ownership and Firm Performance: The Case of Bidder Returns[J]. Journal of Corporate Finance, 1999, (5), pp. 103-107

Franks, Julian R., Robert S. Harris, and Sheridan Titman, The post-merger shareprice performance of acquiring firms[J]. Journal of Financial Economics, 1991, 29, pp.81-96

Grinblatt Titman. Financial Markets and Corporate Strategy[M], McGraw-Hill Higher

Education Second Edition, 2001, Chapter.20 Mergers and Acquisitions

Jensen M C, W H Meckling. Theory of the Firm, Managerial Behavior, Agency Costs and Ownership Structure[J]. Journal of Financial Economics, 1976, 13, pp.305-360

Lei, L. I., and Z. G. Song. "Empirical Analysis of M&A of Chinese Listed Companies Based on Factor Analysis Method." Technoeconomics & Management Research, 2009

Marc, Georgen and Luc, Renneboog. Shareholder Wealth Effects of European Domestic and Cross-border Takerover Bids[J]. European Financial Management, 2004, 1(1), pp. 9-45

(46)

J. Harold Mulherin, Audrea L. Boone. Comparing acquisitions and divestures[J]. Journal of

Corporate Finance, 2000, 6, pp. 117-139

Roll, R. The hubris hypothesis of corporate takeovers [J]. Journal of Business, 1986, 59, pp. 197-216

Sara B. Moeller, Frederik P. Schlingemann and Rene M. Stulz. Firm size and the gains from acquisitions[J]. Journal of Financial Econometrics, 2004, 73, 00. 201-228

Travlos, N.G. Corporate Takeover Bids, Methods of payment, and Bidding Firms’ Stock Returns[J]. Journal of Finance, 1987, 42(4), pp. 943-963

Wansley, J.W., Lane, W.R., and Yang H.C. Abnormal Returns to Acquired Firms by Type of Acquisition and Method of Payment[J]. Financial Management, 1983, (12), pp. 16-22

(47)

Appendix: Stata code //0.for univariate analysis keep if strmatch(excg,"SH") keep if strmatch(excg,"SZ") keep if strmatch(pmt,"cash") keep if strmatch(pmt,"others") keep if inlist(industry,"ENERGY","HT","IND","MATERLS") drop if inlist(industry,"ENERGY","HT","IND","MATERLS") keep if strmatch(attd,"Friendly") keep if strmatch(attd,"Neutral") //1. calculate 'dif' for 'id'

*trading days:

by sedol evtdat: gen datnum=_n

by sedol evtdat: gen target=datnum if date==evtdat by sedol evtdat: egen td=min(target)

drop target

gen dif=datnum-td

//2. calculating the estimation windows

by sedol evtdat: gen estimation_window=1 if dif>=-150 & dif<=-25 by sedol evtdat: egen count_est_obs=count(estimation_window) summarize count_est_obs

replace estimation_window=0 if estimation_window==. //3. calculating the event windows

*************************************************************************** by sedol evtdat: gen event_window=1 if dif>=-2 & dif<=0

by sedol evtdat: egen count_evt_obs=count(event_window) summarize count_evt_obs

replace event_window=0 if event_window==. //4. drop insufficient obs

tab sedol evtdat if count_est_obs<100

*************************************************************************** tab sedol evtdat if count_evt_obs<3

drop if count_est_obs<100

*************************************************************************** drop if count_evt_obs<3

//5. create 'id' for combi(sedol,evtdat) egen id=group(sedol evtdat)

egen n=max(id) display n

//6. estimating normal performance gen predicted_return=.

(48)

*************************************************************************** forvalues i=1(1)n {

list id sedol if id==`i' & dif==0

reg stkret mktret if id==`i' & estimation_window==1 predict p if id==`i'

replace predicted_ret = p if id==`i' & event_window==1 drop p

}

//7. calculate abnormal returns as the diff between actual returns and normal returns sort id date

gen ar=stkret-predicted_ret if event_window==1

//8. keep the date and company&transaction identifiers as well as the abnormal returns for the event window only

keep if event_window==1 keep date id ar

//9. calculate CAAR and TS preserve

bysort id: egen cari=sum(ar) collapse (mean) cari, by(id) summarize cari

*************************************************************************** scalar TS=sqrt(n)*r(mean)/r(sd)

display TS restore

//10. ttest different for univariate analysis bys excg id: egen caar=mean(ar)

drop ar

duplicates drop ttest caar, by(excg)