An announcement effects of share repurchase based on the prior consecutive events

35 Figure 11 ARs and CARs of the repurchase announcement with a prior Corporate Bond of the Market Index Adjustment Model .... 37 Figure 12 ARs and CARs of the announcement repurchase wi

逢 甲 大 學

經 營 管 理 碩 士 在 職 專 班

碩士論文

連續事件基礎之庫藏股買回宣告效果

The Announcement Effects of Share Repurchase

based on the Prior Consecutive Events

指導教授: 王若愚

邱安安

中 華 民 國 一 百 零 七 年 二 月

I would also like in a special way to acknowledge crucial role played by my supervisors, Dr Wang and Dr Chiu, for their invaluable support and guidance at FengChia University Further, my gratitude and appreciations go to all those individuals and organizations whose contributions facilitated the completion of this thesis work I am sincerely grateful for those who shared their truthful and illuminating views on a number of issues related to this thesis Their invaluable and constructive criticism added value to this document Finally, I would like to appreciate the guidance given by the panelist especially my presentation at all stages that has since improved my presentation skills and then quality of

my document

ii

摘要

本研究測試台灣 2008 年至 2016 年股票市場上各種事件後的庫藏股買回宣告效果。研 究上將庫藏股宣告前之特定事件區分為現金增資及發行公司債券兩種，並比較這兩種事件後 宣告買回庫藏股之效果。本研究之資料來源為台灣經濟新報資料庫，共蒐集 2858 件庫藏股買 回事件，其中有 105 次庫藏股宣告之前一年亦宣告進行現金增資，而有 299 家次庫藏股宣告 前一年發行公司債。

本研究主要發現庫藏股宣告前企業之異常報酬為負值，而宣告後之異常報酬為正，顯 示庫藏股宣告對股價有正面的助益，尤其在宣告後的第一天更為明顯。而前一年先進行現金 增資，再宣告買回庫藏股的效果雖然有向上的趨勢，但是累積宣告效果始終為負值，亦即對 股價不會有正面的助益。而企業在前一年有發行公司債券者，則在第二年進行宣布買回庫藏 股之效果為正，顯示這樣的財務操作模式對於股價有正面的幫助。

關鍵字: 庫藏股，買回庫藏股 ，連續事

In my research, I examine the share repurchase announcement effect on various consecutive events in Taiwan from August 2008 to 2016 This study separates the dataset into different groups based on some events, and the announcement effect difference between the consecutive I choose the consecutive events of the seasoned equity offering, Bonds and stock distribution to comparing announcement share repurchases The paper consists of 2858 share repurchase announcements by 2588 firms listed on TEJ, 105 events that announce SEO and then repurchase announcement next year and 299 events that issue bond and then repurchase announcement next year, 555 events that stock distribution and then repurchase announcement next year Event study methodology and significance test have been used to analyze

This study found that Taiwan market is under-reacting to the announcements of share repurchases After announcement day +1 abnormal return was reached on top The paper found that Taiwan market is weak-reacting efficient to SEO announcement while significant reaction efficient to open-market repurchase announcement And issuing bonds was the lowest signal than repurchases announcement effect to Taiwan market, but issuing bond will effect

to share price of announcement repurchase next year The last, the repurchase announcement has a significant effect on stock return after the stock distribution announcement

Keywords: repurchase stock, buyback stock, consecutive events

iv

Table of Contents

Acknowledgement i

Abstract iii

Chapter 1 Introduction 1

1.1 Background 1

1.1.1 Share repurchase 1

1.1.2 Share repurchase announcement in Taiwan 2

1.2 Research Problem 3

1.3 Research Object 6

1.4 The structure of the paper 6

Chapter 2 Literature Review 7

2.1 Share Repurchases Motivations 7

2.2 Effect of share repurchases 9

Chapter 3 Data and Methodology 12

3.1 Data 12

3.2 Event study 13

3.3 Methodology 14

4.1 Whole market impact of announcement repurchases 16

4.2 The consecutive effects of the event 24

4.2.1 The effect the repurchases with a prior SEO announcement 24

4.2.2 The effects of repurchase announcement with a prior Bond issuing 34

4.2.3 The effects of repurchase announcement with a prior Stock dividend distribution 43

Chapter 5 Conclusions 51

References 52

List of Figures

Figure 1 AR and CAR of repurchase announcement of The Market Index Adjustment model

19

Figure 2 ARs and CARs of repurchase announcement of The GARCH Risk Adjustment Model 19

Figure 3 Events window CARs of Market Index Adjustment model 22

Figure 4 Event window CARs of the GARCH Risk Adjustment Model 22

Figure 5 Development of the ARs around the SEO announcement 27

Figure 6 Development of the CARs around the SEO announcement 27

Figure 7 ARs and CARs of the repurchase announcement with a prior SEO event 31

Figure 8 ARs and CARs of the repurchase announcement with a prior SEO event 32

Figure 9 CARs around Corporate Bond issuing 35

Figure 10 CARs around Convertible Bond issuing 35

Figure 11 ARs and CARs of the repurchase announcement with a prior Corporate Bond of the Market Index Adjustment Model 37

Figure 12 ARs and CARs of the announcement repurchase with a prior Convertible Bond with Market Index Adjustment Model 38

Figure 13 ARs and CARs of repurchase announcement with a prior Corporate bond issuing of GARCH Risk Adjustment Model 40

Figure 14 ARs and CARs of repurchase announcement with a prior Convertible Bond with GARCH Risk Adjustment 40

Figure 15 ARs and CARs of stock dividend distribution of Market Index Adjustment Model 44

Figure 16 ARs and CARs of stock dividend distribution of the GARCH Risk Adjustment Model 46

Figure 17 ARs and CARs of repurchase announcement with a prior Stock dividend distribution 48

Figure 18 ARs and CARs of repurchase announcement with a prior stock dividend distribution of the GARCH Risk Adjustment Model 49

vi

List of Tables

Table 1 Summary Statistic 12 Table 2 ARs and CARs around announcement repurchase analysis by the Market Index Adjustment Model 18 Table 3 ARs and CARs around announcement repurchase analysis by the GARCH Risk Adjustment Model 20 Table 4 ARs and CARs around repurchase transactions 23 Table 5 Development of the AR and CAR around the SEO announcement of Market Index Adjustment 26 Table 6 Development of the ARs and CARs around the SEO announcement of GARCH Risk Adjustment Model 28 Table 7 ARs and CARs of the repurchase announcement with a prior SEO event of the Market Index Adjustment Model 30 Table 8 ARs and CARs of the repurchase announcement with a prior SEO event of the GARCH Risk Adjustment Model 33 Table 9 ARs and CARs around Bond issuing 36 Table 10 ARs and CARs of Repurchase Announcement with a prior Corporate Bond issuing 41 Table 11 ARs and CARs of the Repurchase Announcement with a prior Convertible Bonds issuing 42 Table 12 ARs and CARs around Stock dividend distribution of The Market Index

Adjustment Model on the period 2000-2016 45 Table 13 ARs and CARs around stock dividend distribution of the GARCH Risk

Adjustment Model on the period 2000-2016 47 Table 14 ARs and CARs around repurchase announcement with a prior stock dividend distribution on the period 2000 - 2016 50

1997 compared with 14.2 billion Euro in 1996 Previous researchers show that there are 1159 events of announcement repurchase from 1989 to 1997 in Canada(Ikenberry et al 2000)

264 events of buyback from 1985 to 1998 in British (Rau and Vermaelen 2002) Hatakeda and Isagawa (2004) sought that 452 events from 1995 to 1998 in Japan and 800 events from

1993 to 1997 in Hong Kong (Zhang 2005) The share repurchase will be completed mainly

through open market transactions Brav et al (2005) found that share repurchases are now a

more important form of payout compared to the past Besides, Barclay and Smith (1988) assumed that repurchase stock is an important financial strategy of the firm in developed countries

Now that the economy is continuing its recovery, share buyback programs are quickly becoming all the rage once again with companies and investors When a company has excess cash at the end of the day, there are only a few things that it can do with it They can save it for a rainy day, invest in new property and equipment for the business, acquire another company, retire debt, issue a one-t special stock to shareholders, or buy back share

of their stock on the open market A “stock buyback program”, which can also be known as

a “share repurchase program”, is when a company buys its share back from current shareholders through the open stock market Buyback programs can be seen as a signal that

a company believes its shares are undervalued and is often viewed as an efficient way to put money back into its shareholders’ pockets According to Stephens and Weisbach (1998), there are three methods in which a firm can repurchase its own shares: tender-offer repurchases, Dutch auction repurchases and open-market repurchases Following McNally (1999) that an open market repurchase increases a firm’s stock price by increasing its return

on equity and its earnings per share So, the open-market share repurchase is the most known method The stock market reacted positively to the announcement of repurchase Many studies have investigated that when the firms announced repurchases, firms realize positive abnormal returns regardless of the time horizon Zhang (2005) found that the three-

well-year buy and hold abnormal return, which is measured against a portfolio of control firms

that are matched by size and book-to-market value ratios, is over 20% Chan et al (2004)

argued that after share repurchase announcements, announcing firms experience an average buy-and-hold yearly return of 26.2% following the announcement date In addition, a significant increase in the price of the repurchases stock in the next year, implying that the market reacts slowly with the information of buyback in short-term Further, the long-term performance of the stock repurchase is positive

About the reasons that why the firms buy back their stock A stock repurchase plan can be a good way for a business to reinvest in itself, by using any excess cash at its disposal

to buy back shares of its own stock Dittmar (2000) gives several reasons a firm may repurchase stock The first, Excess capital hypothesis Repurchase stock, like paying stocks because in open market repurchases, the firms do not have a commitment to repurchase (Dittmar 2000) and repurchase is a more flexible means of distributing capital since a penalty

is incurred if distributions are subsequently reduced (Denis et al 1994) The second,

Undervaluation hypothesis: Repurchase and Policy The repurchases stock is flexible in timing which is beneficial because firms can wait to repurchase until the stock price If the stock is undervalued, the firm may repurchase stock as a signal to the market or to invest in its own stock and acquire mispriced shares According to this hypothesis, the market interprets the action as an indication that the stock is undervalued (Dittmar 2000) The third, when the firms announced buy back their stock, there are no taxes paid by shareholders Open-markets repurchase stock, shareholders have the freedom to make their own decision when to pay taxes on their gain(Voss 2012) So, having several possible motives behind the companies announce buy back stock Voss (2012) gave 3 motives: undervaluation, managerial incentives, and the relation between stock repurchase and stocks Dittmar (2000) found that firms repurchase stock to take advantage of potential undervaluation and in many periods, to distribute excess capital

1.1.2 Share repurchase announcement in Taiwan

Share repurchases have a long history in the United States, and the research on that has been conducted in the early 1908s Since then, many research papers focused on US stock market There are also many papers focusing on other countries and districts like Canada and Europe These papers dig into the motives of share repurchase and show a large number of empirical results to prove the positive relationship between share repurchases and stock return or results to the positive of announcement repurchase stock In Taiwan, according to corporation law, share repurchase was regulated by articles 158, 163, and 167

of the Company Act However, changes started in 2000 due to the aftermath effect of the Asian financial crisis in 1997 and the burst of the Internet Bubbles in 2000 To reboot the securities market, in July 2000 Taiwan Legislative Yuan passed Article 28-2 of the Securities and Exchange Act and allowed corporations to repurchase their outstanding share for strategic purposes In the first year following the enactment of these regulations, about one-quarter of firms listed on either TWSE or the Over-the-Counter market (OTC) announced share repurchase programs within 5 months of that year Now firms could perform share repurchases for employee shares transferring, equity conversion for coordinating or maintaining the company’s credit, and shareholders’ equity

Listed companies have been permitted to repurchase their shares since August 200

in Taiwan According to the regulations, listed companies are allowed to make repurchase announcement for any of the following purposes: (1) providing shares as incentives, (2) converting bonds to shares, and (3) protecting company creditability and equity The regulations provide listed companies an option to execute them repurchase announcements without any buy-back obligations

A company repurchasing its own share at a centralized securities exchange market

or at the place of business a securities firm shall, within two days on which the resolution was made at a meeting of the board of directors, announce the repurchase The items should

be reported to the Securities and Futures Commission (SFC) including: (1) purpose of the repurchase, (2) type of shares to be repurchased, (3) ceiling on total monetary amount of the repurchase, (4) planned period for the repurchase and number of shares to be repurchased, (5) price range of the share to be repurchased, (6) method for the repurchase and (7) number

of shares held at the time of reporting and so on

Within two months of the day of expiration of the reporting period for the planned repurchase, the company may, through a majority vote at a meeting of the board of directors attended by at least a two-thirds quorum, amend the originally reported purpose of the repurchase by filing a report with the Securities and Futures Commission

1.2 Research Problem

There are several ways in which companies can manage the capital such as stock distributions, stock buybacks, corporate bonds and seasoned equity offering (SEO) All of them is the method of distributing cash to shareholders, but the effects on financial ratios and shareholders’ investment return are different Obviously, some critical information behind these methods of capital adjustment may be conveyed to the market

The firm's issue shares to raise equity capital but if there are no potential growth opportunities in sight, holding on to all the unused equity funding means sharing ownership for no good reason Buyback some outstanding share can be a simple way to pay off investors and to reduce the cost of capital Stocks can be undervalued for a number of reasons If the stock price is dramatically undervalued, the issuing company can repurchase some of its shares at this reduced price After the market price is corrected, the company then reissue the repurchase stocks to increase the equity capital without issuing any additional shares The rapid influx of investors artificially inflates the stock’s valuation and boosts the firm’s price to earnings ratio Share repurchase is an alternative to paying stocks in that it is another method of returning cash to investors A stock repurchase occurs when a company asks stockholders to tender their shares for repurchase by the company A company purchases its own share in the open market anytime to decrease the number of shares held by the public, thereby increasing the ownership stake of each remaining shareholders and hopefully the share price Because every share of stock is a partial share of a company, the fraction of the company that each remaining shareholder owns increases In the short-term, the stock price may rise because shareholders know that a buyback will immediately boost earnings per share Over the long-term, a buyback may or may not be beneficial to shareholders According to Hackethal and Zdantchouk (2006) the number of German buybacks depends

on market state; in Baisse (bear market) it is significantly larger than in Hausse (bull market) Moreover, the effect of buybacks on prices in the first phase is significantly strong than in the second So the first objective of this research is to test the share repurchase announcement effect of the whole market

The stock is a distribution of a portion of a company’s earnings, decided by the board

of directors, paid to a class of its shareholders Stocks can be issued as cash payments, as shares of stock, or other property When a company issues a stock to its shareholder, the value of the stock is deducted from its retained earnings Even if the stock is issued as additional shares of stock, the value of the stock is deducted Miller and Modigliani (1961) stated the irrelevance theory, thus that stocks do not affect the firm value under perfect capital market, under certain strong prerequisites as a certain market process, an efficient market and the absence of taxes bankruptcy, agency costs and asymmetric information Stock dividend distribution is, therefore, considered to be one of the most important financial decisions that corporate managers encounter (Bhattacharya 1979) Hence a firm ought to pay stocks to the shareholder if it cannot identify suitable investments which would bring higher

returns than those expected by the shareholders (Adefila et al 2004) Jagannathan et al

(2000) indicated that stock repurchases and stock are used at different times from one another, by different kinds of firms Stock repurchases are very pro-cyclical, while stocks increase steadily over time Firms repurchase stock following poor stock market performance and increase stocks following the good performance Following Dittmar (2000)

if stocks and repurchases are a substitute, this regulatory change should cause the volume of repurchases to decrease subsequent to the implementation of the change If repurchase and stocks are substitutes, then stock repurchases should be negatively related to a firm’s stock payout ratio With the many researches about the effect of repurchases stock and stock dividend distribution, in this paper, I want to test the firm announced share repurchases with/without stock dividend distribution in the previous year

SEO may signal to the market that the firm is overvalued, while share repurchase may signal the market that the firm is undervalued Indeed, SEO tends to cluster in times of high stock market valuations (Masulis and Korwar 1986, McLaughlin et al 1998 ,Hovakimian et al 2001), whereas share repurchase is more common in times of low market valuations (Ikenberry et al 1995) If managers are indeed timing the market in their

corporate finance decisions by issuing stocks, they would do SEO and repurchase when the stock is overvalued and undervalued respectively They should decrease their holding of company stock in years when there is an SEO, while they should increase their holdings in years when there is a share repurchase The distributions of net purchase and change in holding shift to the left when there is an SEO does not change when there is a cash acquisition and shift to the right when there is a share repurchase In this paper, I test the difference between when the firm announcement repurchases and when the firms give the SEO

In practice, some companies would like to raise funds by using bonds, which helps them obtain the funds supporting the capital budgeting and avoiding the earnings dilution The announcement of the new bond issue has been seen have a negative effect on stock prices Potential explanations of this negative effect- the price-pressure, wealth-redistribution, and information-release hypothesis- imply different share-price reactions to the announcement of bonds (Kalay and Shimrat 1987) Hotchkiss and Ronen (2002) gave that if the bond market is less efficient, stocks will reflect information about the value of underlying assets more quickly, stock returns have predictive power for future bond returns

In terms of informational efficiency, the behavior of these bonds is similar to that of the underlying stock This paper investigated the difference between the share-repurchase firms with corporate bond issues in the previous year

1.3 Research Object

This paper investigates the share-repurchase announcement effects on various consecutive events Most of the previous research focused on the announcement effects, however, none of them pay attention to the firms’ consecutive behaviors such as seasonal offering and corporate bonding issuing and then adopting the share-repurchase This research aims to test the share repurchase consecutive announcement effects

1.4 The structure of the paper

This research is divided into five sections including this one on Introduction Next section covers the past literature on the subject The third section lays down the research methodology employed in the study The empirical findings are reported in the fourth section

and finally, the last section concludes the paper with an outline of implications of the study

Chapter 2 Literature Review

A share repurchase refers to the process when a company re-acquires its own stock

or, in other words, the company buys shares back from its shareholders Share repurchase is

a tool for managers to buy back a company’s own shares on the stock market The company uses cash to buy back its own shares, decreasing the among of shares outstanding The in

1967 a Senate Committee wrote: “Corporate repurchases of their own securities may serve

a number of legitimate purposes For example, they may result from a desire to reduce outstanding capital stock following the cash sale of operating divisions or subsidiaries, or

to have shares available for options, acquisitions, employee or stock purchase plans, and the like, without increasing the total number of shares outstanding Repurchase programs, however, may also be utilized by management to preserve or strengthen their control by counteracting tender offers or other attempted takeovers or may be made in order to increase the market price of the company’s shares Whatever the motive behind the repurchase program, if the repurchases are substantial they will have a significant impact

on the market.” (Senate Report No.550, 90th Congress, 1967)

I divided this chapter into two sections In the first section, the motivations of share repurchase will be reviewed The second section which discusses the efficient market theory, which describes the effect of share repurchase announcements on the share price movements

2.1 Share Repurchases Motivations

This part describes the motives which explain why firms enter into a share repurchase program Share repurchases are similar to stocks because both are used to distribute cash back to stockholders and both are considered as a positive signaling In past literatures, there are several reasons for buybacks motives such as the signaling hypothesis or undervaluation

hypothesis (Ikenberry et al 1995; Baker et al 2003), the free cash flow hypothesis (Stephens

and Weisbach 1998); the preferential tax hypothesis, increase earnings per share and repurchases and stock options (Voss 2012) Dittmar (2000) investigated the relation between stock repurchase and distribution, investment, capital structure, corporate control, and compensation policies over the 1977-96 period He gave five reasons a firm may repurchase stock The first, excess capital hypothesis: Repurchases and Distribution The second, Undervaluation Hypothesis: Repurchases and Investment policy The third, Optimal leverage ratio Hypothesis: Repurchases and Capital structure policy The fourth, Management Incentive Hypothesis: Repurchases and Compensation policy The last, Takeover Deterrence Hypothesis: Repurchases and Corporate He tests the five hypothesis

discussed above with the following Tobit model estimated for each sample year using sectional data The sample consists of all firms listed on Compustat and by the Center for Research in Security Prices Dittmar stated that firms repurchase stock when they are potentially undervalued Firms also repurchase stock to distribute excess capital, increase their net leverage, fend off takeover attempts, and counter the dilution effects of stock options Firms repurchase stock to fend off takeover attempts in many of the year that coincide with peak merger periods The repurchasing to take advantage of undervaluation is the most consistently significant motive for repurchasing stock over the sample period however, it is only one of the significant motives for repurchasing A firm may repurchase for only one of these reasons, or it may repurchase only when multiple criteria are met So following the previous research, the main motives are constructed as follows: signaling hypothesis, free cash flow hypothesis, capital structure hypothesis, stock substitution hypothesis and tax efficiency hypothesis

cross-The signaling hypothesis is based on the belief that information asymmetries exist between management and outsiders The management team knows more than the shareholder does If managers believe the market price of their stocks does not provide a fair value for the discounted future cash flows, management can repurchase shares Therefore, managers can use share repurchase as a signal to the less informed outside investors if they disagree on the current share price or express their expectations on future earnings and firm performance Besides, The signaling hypothesis based on undervaluation (Vermaelen 1981) Undervaluation implies that based on the premise of information asymmetry between insiders and outsider, a firm may be misvalued (Dittmar 2000) Managers believe that the stock is undervalued, the firm may take action to repurchases shares as a signal to the outside market or to invest in its own misprices shares Then the market interprets the firm’s action

as a signal that the stock is undervalued Furthermore, the positive share price reaction to the announcement date should correct the misprices stock (Dann 1981; Vermaelen 1981)

However, firms do not have to actually take share repurchase action event if they make

an announcement (Rau and Vermaelen 2002) Chan et al (2004) showed that firms look at

changes in earnings and decide whether to repurchase after that They also find the negative relation between abnormal returns on announcement dates and abnormal returns after the announcement

Free cash flow hypothesis is a possible explanation for share repurchase According to

Wu (2012) the repurchases stock uses management ownership to measure the severity of firms’ agency problem The findings suggest that firms with a less severe agency problem

have more information in the repurchase announcements, buy back fewer shares, and perform better after the repurchase programs The firms will distribute excess cash flow to shareholders

so as to reduce manager power Stephens and Weisbach (1998) found a positive relation between excess cash flow and repurchase transactions, the more excess cash flow, and the larger quantities of share repurchase However, Grullon and Ikenberry (2000) gave that free cash flow hypothesis is inconsistent, and they found that firms which do not execute share repurchase after announcement earn a higher excess return than those who actually repurchase shares

The firms may use share repurchase to fine –tune their capital structure and respond to the potential dilutive effects from employee stock option plans Grullon and Ikenberry (2000) state that share repurchase is a popular way for firms to change the capital structure Dittmar (2000) found that when the firms distribute the capital, they reduce their equity and increases their leverage ratio However, Dann (1981) proposes that it is not the best option for a firm to initiate a share repurchase program if it tends to achieve the optimal leverage ratio since the issuance of new debt would be a better alternative

Stock substitution hypothesis and tax efficiency hypothesis are complementary to each other Stock substitution hypothesis implies that share repurchase is a substitute for the stock payout since it is more flexible Voss (2012) reported that the relative advantages of share repurchase over stocks because of the tax preference hypothesis, the type of shareholder hypothesis, and the financial flexibility hypothesis however limited, merit in explaining repurchases Zhang (2005) found the negative relation between stock cut and firm value, while share repurchases programs do not bear such kind of risk

Tax efficiency hypothesis implies that companies may prefer share repurchase if the tax burden is higher on stocks than on capital gains, since stocks are subject to the ordinary income tax, while in terms of share repurchase, investors only need to pay tax on the difference between the purchase price and the selling price, which is the capital gain tax (Grullon and Michaely 2002) In addition, share repurchase has the advantage of postponing the realization

of capital gains and therefore postpone the tax payment Bagwell and Shoven (1989) have indicated that firms substitute share repurchase for stock payout in order to get the tax benefit for shareholders

2.2 Effect of share repurchases

McNally and Smith (2007) found that companies utilize limit orders when repurchasing shares, suggesting that firms repurchase shares in an attempt to provide liquidity and buffer sell-side pressure in order to provide price support for falling stock prices McNally

and Smith (2007) found that, on average, share repurchases for TSX listed firms account for 12.25 percent of total trading volume during the repurchase program period

Repurchase stock reduces the number of shares in a company held by the public Because every share of stock is a partial share of a company, the fraction of that company that each remaining shareholder owns increase In the near term, the stock price may rise because shareholders know that a buyback will immediately boost earnings per share Over the long

term, a buyback may or may not be beneficial to shareholders Wang et al (2013) investigated

the term and the long-term price performance of repurchasing firm He found the term 4-day cumulate abnormal returns CAR (-1,2) is 1.9142 percent while the mean of CAR (-1,5) is 2.7086 percent It means the initial market reaction seems small and less than the daily standard deviation of returns for many stocks If manager buyback shares because of undervaluation for other reasons, it is likely to be in expectation of a larger price increase The long-term, the average 3-year buyback abnormal returns are 38.82% to the market index and 44.30% relative to the market model for all repurchasing firms The result indicates that the market adjusts slowly in the short run and the long-term price performance is positive Zhang (2005) investigates share price performance surrounding and following actual share repurchases The paper uses data from 135 firms that have made 3628 daily repurchases in Stock Exchange of Hong Kong from September 1993 to August 1997 The paper analysis basing on short-term (using 250 days of return data, from 270 to 21 days prior to the event study) and long-term (three years following the actual repurchase month) With short-term, the paper used the market model to calculate the cumulative abnormal returns The paper suggests the firms tended to repurchase share when their stock relatively underperformed the market with the CAR (-20, -1) is -1.84% After the announcement, the average CAR (0,2) for the three-day event period is 0.43% This means the market responded positively to the actual repurchases And the 21-day return CAR (0, 20) is 0.69% This mean, the average short-term market response to actual share repurchases is significantly positive, but the

short-magnitude of the market reaction is not very big In the long-term, the paper analyzes the

power and specification of test statistics for detecting long-run abnormal stock returns under three measurement benchmarks: reference portfolios, control firms, and the Fama-French three-factor model They suggested that the buy-and-hold returns following actual share repurchase events up to three years Buy-and-hold abnormal returns are calculated basing on firm size and book-to-market value The result of long-term that managers of value firms can deliver superior performance to long-term shareholders The three-year buy-and-hold

abnormal return, which is measured against a portfolio of control firms that are matched by size and book-to-market value ratios, is over 20%

According to Grullon and Michaely (2004) stated that the earnings-per-share impact

is an important factor in determining their repurchase decisions When conducting share repurchases, the EPS of a company tends to increase because the number of shares increases proportionately more than the earnings decrease as a result of a decline in interest income arising from the lower cash position In seasoned equity offerings, especially in cases where the share is used as a part of employee remuneration and not corporate investments, the EPS impact tends to be negative It is due to aforementioned EPS impacts that companies are much more likely to rely on share repurchase as a means of obtaining shares to cover employee stock option compensation rather than seasoned equity offerings

Maxwell and Stephens (2003) have identified a relationship between repurchases and wealth transfer between a firm’s equity and debtholders Maxwell and Stephens (2003) observed wealth expropriation through reaction in stock prices and bond markets to share repurchase Their studies found that on average bond returns fall by 18.5 basis point (at 1% significance level) around the time of the repurchase announcement, additionally, bond ratings, following a repurchase announcement, are more likely to be downgraded than upgraded

There are other papers that look at the effects of share repurchase announcement specific to the market In my paper, I will investigate the share repurchase announcement effect on various consecutive events in Taiwan The other point of my research is focused

on the firm’s consecutive behavior, compares before and after having an event

Chapter 3 Data and Methodology

This part of the paper contains a description of the underlying dataset that will be employed for the event study and the cross-sectional analysis This thesis is intended to examine the effect of open-market share repurchase announcements in Taiwan With the aim

of this paper, this thesis will use 2 methods to derive the excess returns, namely the market adjusted returns model and GARCH risk adjustment model As a result, this part divided into two parts: one is data using in my thesis and the other is event study methodology on abnormal returns

3.1 Data

For this research, a dataset of share repurchases focusing on the various consecutive events in Taiwan has been constructed The sample is constructed from TEJ database for the announcement of intention to repurchase share, stock distribution, SEO, and bond issuing The sample period covers 16 years between 2000 and 2016 The databases complement each other and other isolated announcements are adding from TWSE classification index and TSE Taiex is using as local market indexes for Taiwan stock market, respectively In order to conduct the event study and regression analysis in the later part, the companies selected should have enough information to investigate the share price effects Therefore, the share prices from 250 until 10 business days before the event date should be available as the estimation window

Table 1 Summary Statistic

This table describes the summary statistics for announcements repurchases, stock distribution, SEO

and Bonds during 2000-2016 for Taiwan

As a result, there are 2588 companies take part in TEJ from 2000 to 2016 into having

2858 share repurchase events, 610 SEO events, 4835 stock events, 809 corporate bonds, 629 convertible bonds The firms announce SEO at time t and then repurchase in the next year have 105 events The firm’s issue bond at time t and then repurchase in next year have 299 events (include 92 with the corporate bond and 207 with the convertible bond) The firms stock dividend distribution at time t and then repurchase in the next year have 555 events

3.2 Event study

This thesis began with the literature review of the theories and current research on the effect of open market repurchases announcements on stock prices and the consecutive effects on stock prices In these descriptions, the motives for firms to repurchase their stocks were discussed as well as the fluctuation Taiwan abnormal stock returns around the announcement day The main goal of this thesis is to estimate the price impact of open market share repurchases The price impact has been investigated in the short-term, around the announcement of the share buyback program Event study methodology has been performed in order to detect information about share price behavior on the event day and to define whether trends appear in the times surrounding of the event period

An event study is a statistical method to assess the impact of an event on the value

of a firm Event study analyses differentiate between the returns that would have been expected if the analyzed event would not have taken place and the returns that were caused

by the respective event An event study is conducted to measure the influence of a specific event on the value of a firm The purpose of event studies is to test for the existence of an information effect and to identify factors that explain changes in firm value on the event

date Campbell et al (1997) suggested that a typical event study is conducted by first

defining the event and establishing the event window This means to establish exactly what the event is and determining the period during which share price will be affected by this event The price impact has been investigated in the short-term, around the announcement

of the share buyback program Event study methodology has been performed in order to detect information about share price behavior on an even day and to define whether or not trends appear in the times surrounding of the event period Return are indexed according to the dates in the events window, with the event date at t The event date is equal to the announcement of the open market repurchase

Although other measures can be and have been used, an event study typically examines the stock price change for a sample of firms experiencing a common type of event(Kothari and Warner 2007) Stock market data are used here for expositional purposes Abnormal or unexpected return in a period surrounding the event is used to provide a measure of the unanticipated impact of the event on the wealth of the firms’ shareholders or other stakeholders The abnormal return (AR) is defined as the realized return minus the expected or normal return, with the latter determined by a variety of expected return models Although the focus has been on the mean (in cross-section or over time) AR around the occurrence of the event, other measures such as the change in returns variance and trading volume have also been used The basic steps for an event study are as follows: (1) Define the event of interest and identify the event date (2) Define the event window (the period the researcher wants to focus on when examining the effects of the event) and the estimation

window (the period during which the parameters of the expected return model are estimated) (3) Select the sample firms, eliminating those that have confounding effects or contaminating events (4) Calculate the Abnormal returns and Cumulative abnormal returns Where non-stock market measures are used the performance measure for the pre- and post-event periods, possibly after controlling for other relevant factors, needs to be computed (5) Aggregate the ARs in cross section and/or over time (6) Determine the statistical significance of the aggregated ARs

3.3 Methodology

To assess the market reaction to the share- repurchase announcement, I employ the Market Index Adjustment model and GRACH Risk Adjustment model to compute the abnormal return Figure 1 shows the timeline of an event study that using the thesis The event window to [-10, 10] With the GRACH, the estimation window spans from days -250

to day 0

 The Market Index Adjustment Model

The market model is used to estimate the coefficients The market adjusted model

assumes that alpha is set equal to zero and beta to one De Jong et al (2011)

The market adjusting model is described as follows a single factor market model:

In the Market Adjusted Model, the observed return of the reference market on day t

Rmt is subtracted from the return Rit of the observation i on day t

 The Generalized autoregressive heteroscedasticity (GARCH) model

The GARCH Risk Adjustment model has been found to provide a good description

of the variance of daily stock returns In this model, any large shock to a share price which causes an exceptionally high or low abnormal return on a particular day, also causes the variance of returns to be high on the following day, and to decay only slowly back to its long-run average “unconditional” value (Bollerslev 1986)

In the GARCH risk adjustment model with a single factor market model with GARCH (-10,10) errors is estimated, namely:

The abnormal returns are the crucial measure to assess the impact of an event The general idea of this measure is to isolate the effect of the event from other general market movements The abnormal return of firm i and event date is defined as the difference between the realized return and the expected return has given the absence of the event:

[ | ]

it it it it

AR R E R  (6) The expected return is unconditional on the event but conditional on a separate information set Dependent on the definition of the information set and the functional form there exist various models of the normal return Those models are extensively discussed in the following section

The next step is to calculate the cumulative abnormal returns (CAR) by summarizing the abnormal returns in the event window identified by each announcement CAR is the sum

of abnormal returns during the event window periods

Chapter 4 Data Analysis

This chapter presents the data analysis based on the daily share prices of the companies that have announced share-repurchase and are listed on TEJ The analysis uses the event study methodology and descriptive statistics to test the effect of the announced buy-back on the share prices of the companies Event study methodology determines whether there arises the positive or negative abnormal return around the announcement repurchase stock event by defining an event window covering a period of days before and after the actual event date

4.1 Whole market impact of announcement repurchases

My thesis carried out was to establish the effect of open announcement repurchase stock listed on TEJ The data collected to facilitate the study was the daily share prices of the companies for the period from 2000 to 2016 and the TSE Taiex This section presents the short-term price impact of share repurchase transaction In table 2, I present the abnormal share price performance surrounding the repurchase event day for the full sample

In Figure 1 shows ARs and CARs analysis by the Market Index Adjustment model

of the whole market around the announcement repurchase in event window [-10,10] Before the announcement, ARs was the gradual decrease The ARs touched the lowest in the day -

2 is -0.517 ARs was the increase from the day -1 and increased dramatically hitting a peak

is 1.8115 in the day +1 After ARs was decreasing but was positive

The empirical results show that the entire sample yields a significant negative CARs

on the following periods day -1 to day 0 This means that stock price is clearly undervalued

On the other hand, a significantly increase CARs is noted on day +1 and the day +2 This shows that stock repurchase announcements cause a significantly positive response from the market From -10 to -1, CARs is negative The similar fluctuate of ARs, CARs was sunk to

a bottom in the day -1 is -3.2553 On the announcement day, CARs was slightly increased from -3.4279 to -3.2533 After the announcement, CARs was increased In the day +1, CARs was suddenly surged (from -3.2533 on the announcement day to -1.4418 in the day +1)

The figure 2 and table 3 below illustrates the ARs and CARs of repurchase announcement of the GRACH Risk Adjustment Model in the event window [-10,10] ARs

of the GRACH Risk Adjustment model was negative from -10 days to -1 day On the announcement day, ARs was slightly decreasing (from -0.3417 to 0.2363) ARs reached the highest point in the day +1 From the announcement day, ARs was quickly increased in the day +1 (from 0.2363 to 1.8927) After the day +1, ARs was around 0.2 CARs of the GRACH

Risk Adjustment model was similarly fluctuation of CARs of the Market Index model Before the announcement day, CARs was decreasing and fallen to a low in the day -1 (-2.9174) and surged in the day +1 (from -2.6812 to -0.7885) Before the day +3, CARs was the positive and gradual increase

Both of model, CARs was waved The bottommost mark in the day -1 and reached the highest point in the day +1 The Market Index Adjustment model, after 2 days announcement, CARs was insignificant while CARs of the GARCH Risk Adjustment model was the slight rise

18

Table 2 ARs and CARs around announcement repurchase analysis by the Market Index Adjustment Model

I use standard event day methodology base on the Market Index Adjustment model The market Index in the TSE Taiex Event window is [-10,10] with the sample consists of

2588 event from 2000 to 2016

AR(%)

sectional statistics

Cross-Prob

Value

Statistics of sign test method

Prob

Value CAR(%)

Cross-sectional statistics (accumulation)

Prob

Value

Statistics of sign test method (accumulation)

Figure 1 AR and CAR of repurchase announcement of The Market Index

-10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10

AR (%) CAR (%)

20

Table 3 ARs and CARs around announcement repurchase analysis by the GARCH Risk Adjustment Model

I use standard event day methodology base on the GARCH Risk Adjustment model The market Index in the TSE Taiex Event window is [-10,10], the estimation is

[-250, 10] and day 0 is the announcement date, with the sample consists of 2588 events from 2000 to 2016

AR (%)

sectional statistics

Cross-Prob

Value

Statistics of generalized sign test method(p = 0.474)

Prob

Value

CAR (%)

Cross-sectional statistics (accumulation)

Prob

Value

Statistics of generalized sign test method (accumulation)

Prob Value

After I analyzed the fluctuation of the full sample, I focused on three periods to analyze In table 4 represents the ARs and CARs around repurchase transactions

In Table 4, ARs from -10 to -1 days was negative and then from the perspective of the signal effect, there were significant positive AR after share repurchase announcements; moreover, the highest ARs happened on day 1 of both of models (2.3861, 1.8661,1.3709) and (2.4518, 1.9024, 1.4912) This meant that the signal effect of the whole market existed, and the strongest effect was on day 1 However, ARs at day 0 (-0.1528 ad -0.1033) during the period 2007-2009 (the financial crisis) was still negative compared with the other periods 2000-2006 (0.2908, 0.3907) and 2010-2016 (0.2434 and 0.2582) thus, there was information delay during the financial crisis

In Figure 3 and 4, from the perspective of undervaluation, most CARs before share repurchase announcement was negative but turned from negative into positive after the announcement It means that the signal effect was the most powerful in the whole market into two periods (2000-2006), (2010-2016) However, during the financial crisis, CARs after the share repurchase announcement were negative but also the value grew up (Market Index Adjustment model from -5.1848 to 0.7817 and the GARCH Risk Adjustment model from -4.8597 to 0.0236), thus the undervaluation was serious especially during the financial crisis

It implied that the whole market faced a seriously undervalued situation before and during the financial crisis, and undervaluation degree was more obvious during the financial crisis However, the undervaluation degree was eased slowly after the Financial crisis With the GARCH Risk Adjustment model, CARs still 10 days after the announcement turned from negative into the positive These results suggest that Taiwan market is under-reacting to the announcements of share repurchases

22

This figure represents CARs of Market Index Adjustment model from day -10 to day +10 around share

repurchase transactions of periods (2000-2006, 2007-2009, 2010-2016) Figure 3 Events window CARs of Market Index Adjustment model

This figure represents CARs of GARCH Risk Adjustment model from day -10 to 10 around share repurchase

Table 4 ARs and CARs around repurchase transactions

I use standard event study methodology base on the Market index adjustment model and the GRACH Risk Adjustment model The market Index is the TSE Taiex The estimation period is from 250 to 10 days prior to the announcement and day 0 is the announcement date The sample consists of 2588 share repurchases trades for the period

2000 to 2016 I separate observations into three time periods: (1) 2000-2006, (2) 2007-2009, (3) 2010-2016

AR (%)

CAR (%)

AR (%)

CAR (%)

AR (%)

CAR (%)

AR (%)

CAR (%)

AR (%)

CAR (%)

AR (%)

CAR (%)

24

4.2 The consecutive effects of the event

4.2.1 The effect the repurchases with a prior SEO announcement

4.2.1.1 SEO announcement

To start the analysis of the event study results of additional equity offers impact to the whole market Figure 5 and Table 5 depict the development of the ARs around the SEO announcement day for the whole sample The red line represents the ARs of GARCH Risk Adjustment model and the blue line represents the ARs of market adjustment model The graphs can give the first impression of how the ARs move in the 20 trading-day periods surrounding the announcement and one can also get the first insight into which effects the announcement has on the different ARs

For firms that issue additional equity, Figure 5 and Table 5 show that ARs over a day window [-10, 10], where day 0 is the announcement date When interpreting the graph, the reader should be aware that each data point for AR corresponds to the adjusted closing stock price Therefore, the distance between day -1 and 0 constitutes the change on the SEO announcement day both of models AR at day -1 is -0.1862% and at day 0 is 0.0692%, these imply that the change in stock prices is caused by the announcement of SEOs that there is a negative impact of security price on SEO announcements Base on Figure 5, no significant mean or media AR occur for the five reported multi-day windows for the period of 2000-

21-2016 The lowest AR seem to occur on the after one date of the announcement Interesting

to note is that the days from -4 to -2 all report positive abnormal returns Based on Figure 3, significant positive mean and median ARs occur on the fourth, third and two days before the announcement date and highly significant and positive mean and median daily AR of 0.0983%, 0.1693% and 0.153% (Market Index Adjustment model) and 0.0246%,0.0815% and 0.2143% (GARCH Risk Adjustment model) ARs exhibit more negative average daily ARs around the announcement days and larger price reductions after the announcement days The mean ARs for the announcement 0 days and 1 day of 0.0692% and -0.3162% with the Market index adjustment model and 0.0132% and -0.3279% with the GARCH Risk Adjustment model Figure 5 and Table 5 indicate that the significant positive mean average abnormal returns after 5 date announce and with the largest ARs of 0.0986% 8 dates with the Market index adjustment model and -0.0267% on the day +7 with the GARCH Risk Adjustment model So, the findings of significant negative daily average ARs around seasoned equity offerings in Taiwan

Based on Figure 6 and Table 5, CARs during the event period (20 days prior and after the event date) decline gradually for the period of 2000-2016 The CARs found on

SEOs are all signed at the 1% level The results that are reported in the Figure 6 confirms that the significant positive abnormal returns over the pre-event from -6 date of 0.0395% to -2 date of 0.6378% with the Market Index Adjustment model and from -6 day of -0.3279%

to -2 day of 0.0762% with the GARCH Risk Adjustment model The highest CARs on the two days before of announcements And after the announcement, the CARs do drop substantially The blue line fluctuates around a CAR of Market Index Adjustment model of around -0.1% until the end of the observations at the day +10 The red line fluctuates around

a CAR of GARCH Risk Adjustment model trend decrease sharply until -1.7067% at the day +10 The evidence of the calculation in this section shows that additional equity offers have

a negative effect on the stock price of firms on average As on could have already guessed from Figure 5 and Figure 6, the negative reaction is more severe for the trading days directly surrounding the announcement days

26

AR

(%)

Cross-sectional statistics

Prob

Value

Statistics of sign test method

Prob

Value

CAR (%)

Cross-sectional statistics (accumulation)

Prob

Value

Statistics of sign test method (accumulation)

Prob Value

-10 0.0148 0.133 0.8942 -0.6614 0.5083 0.0148 0.133 0.8942 -0.6614 0.5083 -9 -0.0511 -0.4228 0.6725 -2.7402 0.0061 -0.0363 -0.205 0.8376 -1.7008 0.089 -8 0.1362 1.077 0.2815 -1.0394 0.2986 0.0999 0.4397 0.6602 -0.378 0.7055 -7 -0.0183 -0.138 0.8902 -0.7559 0.4497 0.0816 0.304 0.7611 -2.0788 0.0376 -6 -0.0421 -0.3338 0.7386 -1.6063 0.1082 0.0395 0.1263 0.8995 -2.2678 0.0233 -5 0.1777 1.3747 0.1692 -0.189 0.8501 0.2172 0.624 0.5326 -1.1339 0.2568 -4 0.0983 0.8009 0.4232 0.0945 0.9247 0.3155 0.8611 0.3892 -0.7559 0.4497 -3 0.1693 1.3608 0.1736 -1.1339 0.2568 0.4848 1.237 0.2161 -0.7559 0.4497 -2 0.153 1.2618 0.207 -1.4174 0.1564 0.6378 1.5475 0.1218 -1.4174 0.1564 -1 -0.1862 -1.429 0.153 -2.9292 0.0034 0.4516 1.0338 0.3012 -0.9449 0.3447

Table 5 Development of the AR and CAR around the SEO announcement of Market Index Adjustment

The table illustrates ARs of the firms around the SEO announcement of 610 events in the window [-10, 10] of Market Index Adjustment Model from 2000 to 2016

This figure represents ARs of two models from day -10 to day +10 around the SEO announcement from 2000

to 2016

Figure 5 Development of the ARs around the SEO announcement

The figure represents ARs of two models from day -10 to +10 around the SEO announcement from 2000 -2016

Figure 6 Development of the CARs around the SEO announcement

28

AR

(%)

Cross-sectional statistics

Prob

Value

Statistics of generalized sign test method (p = 0.467)

Prob

Value

CAR (%)

Cross-sectional statistics (accumulation)

Prob

Value

Statistics of generalized sign test method (accumulation)

Prob Value

-10 -0.0046 -0.0361 0.9712 0.0225 0.9821 -0.0046 -0.0361 0.9712 0.0225 0.9821 -9 -0.0708 -0.5541 0.5795 -0.5789 0.5627 -0.0754 -0.4175 0.6763 0.0225 0.9821 -8 0.0604 0.4702 0.6382 -0.0778 0.938 -0.0151 -0.0679 0.9459 0.4233 0.6721 -7 -0.1559 -1.2107 0.226 -0.6791 0.4971 -0.1709 -0.6655 0.5058 -0.7793 0.4358 -6 -0.157 -1.2164 0.2238 -0.7793 0.4358 -0.3279 -1.1397 0.2544 -1.7815 0.0748 -5 0.0837 0.6473 0.5174 1.1249 0.2606 -0.2442 -0.7738 0.439 -1.3806 0.1674 -4 0.0246 0.1903 0.8491 0.3231 0.7466 -0.2196 -0.6433 0.52 -0.8795 0.3791 -3 0.0815 0.6285 0.5297 -0.178 0.8587 -0.1381 -0.3781 0.7054 -0.6791 0.4971 -2 0.2143 1.6507 0.0988 0.3231 0.7466 0.0762 0.1964 0.8443 -0.6791 0.4971 -1 -0.2183 -1.6799 0.093 -2.0821 0.0373 -0.1421 -0.3472 0.7284 -1.4808 0.1387

Table 6 Development of the ARs and CARs around the SEO announcement of GARCH Risk Adjustment Model

The table illustrates AR and CARs of the firms around the SEO announcement of 610 events in the window [-10, 10] of the GARCH Risk Adjustment Model from 2000

to 2016

4.2.1.2 The effects of repurchase announcement with a prior SEO event

This section, I used two models to the analysis: Market Index Adjustment model and GARCH Risk Adjustment model

 The Market Index Adjustment model

Figure 7 and Table 7 present CARs and ARs of the Market Index Adjustment which

is the announcement SEO at the time t and repurchase next year The solid line shows that ARs for repurchase, whereas the dashed line shows the CARs for repurchase The graphs can give a first impression of how the ARs and CARs fluctuate in the 21 days surrounding the announcement

In the sample of firms that sign repurchase announcement, the graph of the AR starts

at about -0.1 on the day -10 and reach to the top on the day -3 the same trend as SEO announcement Before the announcement, the lowest of AR on the day -4 (-1.0607), go up -0.9471 on the day -3 and fall -1.0358 on the day -1 AR on event window [-1,1] upsurge On the announcement day, AR jumps from -0.8842 on the day -1 to 0.2231 AR peaks on top

on the day 1 is 1.8188

With CAR of repurchase, the mean CAR (-10, -1) value for all repurchase firms which issue SEO first and next year announcement buyback stocks from -0.1008 to -6.7036 From Figure 7, I observe a price decline in share repurchase before a repurchase transaction, which may indicate that managers on average time repurchases after a negative drift in the stock price The initial market reaction to the announcement is significantly positioned with

a CAR of -4.6617 in the day +1 and -4.0297 in the day +2 Although the positive response

is statistically significant under zero, the economic magnitude is relatively small After the event, the CAR for the (+2, +10) is insignificant These results suggest that the market is reacting to the announcement SEO and then repurchases next year but the effect is not significant

Cross-Prob

Value

Statistics of sign test method

Prob

Value

CAR (%)

Cross-sectional statistics (accumulation) Prob Value

Statistics of sign test method (accumulation)

Table 7 ARs and CARs of the repurchase announcement with a prior SEO event of the Market Index Adjustment Model

The table illustrates ARs and CARs of the firms who SEO announcement and then repurchase announcement next year of 610 observations in the window [-10, 10]

This figure represents ARs and CARs of the Market Index adjustment on the period days -10 to 10

from 2000- 2016

Figure 7 ARs and CARs of the repurchase announcement with a prior SEO event

 The GARCH Risk Adjustment model

The Table 8 shows that AR and CAR for both consecutive in event window [-10, 10] Before the announcement repurchase, ARs was a gradual decrease in event window [-

8, -1] with the sink to a trough in the day -2 (-0.8388) and then a substantial increase in event window [-1, 1] The highest of 2.0023 at the day +1 After the 2-day announcement, ARs plummeted

The Figure 8 below, from the undervaluation The dashed line indicated that the wave

of CAR of the repurchase after the firms announce SEO CARs in event window [-10,10] were negative From day -9 to day -1, CAR rapid decrease from -0.1284 to -5.1784, the lowest CARs on the day -1 However, CARs on the day 0 increased but insignificant compared with day -1 (-5.1784, -4.3977) After the day announcement, CARs was sequentially jumped, and reach the vertex on day 9 is -1.1919 It means that the signal effect

of announcement repurchases to the market

In sum, these results suggest that Taiwan market is weak-reacting efficient to SEO announcement while significant reaction efficient to open-market repurchase announcement