|USFDC Home | USF Electronic Theses and Dissertations||| RSS|
This item is only available as the following downloads:
Disentangling the Repurchase Announcement An Event Study Analysis to the Purpose of Repurchases by Robin S. Wilber A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Finance College of Business Administration University of South Florida Major Professor: Jianping Qi, Ph.D. Ninon Sutton, Ph.D. Christos Pantzalis, Ph.D. Gary Patterson, Ph.D. Date of Approval: March 4, 2005 Keywords: Agency, Executive Compensation, Stock Options, Acquisitions, Method of Payment Copyright 2005, Robin S. Wilber
i Table of Contents List of Tables iii Abstract iv Chapter 1: Introduction 1 Chapter 2: When is repurchase announcement not good news? 6 Literature Review 7 Repurchase or Dividend Decision 7 Free Cash Flows or Undervaluation 8 Type of Repurchase 12 Management Compensation and Options 16 Repurchases and Managerial Ownership 23 Prediction, Data and Methodology 28 Hypotheses 28 Sample 32 Methodology 34 Results 39 Conclusion 50 Chapter 3 Why do firms repurchase stoc k to acquire another firm? 79 Literature 80 Prediction, Data and Methodology 94 Hypotheses 94 Sample 100 Methodology 101 Results 101 Conclusion 106 References 118 Bibliography 124 Appendices 125 Appendix A: T-Test 126 About the Author End Page
ii List of Tables Table 2-1 Types of Repurchase Announcements 52 Table 2-2 Hypothesized Relationships 53 Table 2-3 Variable Definitions 54 Table 2-4 Returns to Repurchase Purpose 55 Table 2-5 Free Cash Flow and Executive Options 59 Table 2-6 Executive Options 61 Table 2-7 Executive Ownership and Options 62 Table 2-8 Type of Repurchase Abnormal Returns 64 Table 2-9 Market Reactions to Repurchase Announcements 69 Table 3-1 Hypothesized Relationships 107 Table 3-2 Variable Definitions 108 Table 3-3 Repurchase to Fund an Acquisitions 109 Table 3-4 Comparison of Acquisition with and without a Repurchase 112 Table 3-5 Acquiring Firm Characteristics 112 Table 3-6 Market Reaction to Type of Financed Acquisition Announcement 113
iii Disentangling the Repurchase Announcement An Event Study to the Purpose of Repurchases Robin Wilber ABSTRACT Researchers have consistently shown th at a firms repurchase announcement is met with positive abnormal stock price return reactions. Open-market repurchases are extremely flexible, non-committal and non-punitiv e; thus, it is puzzling that the mere announcement of an open-market repurchase will likely increase a firms stock price. I propose to disentangle a firms choice to re purchase its stock to determine when a repurchase announcement is good news for shar eholders and when the announcement is not. I find that the purpose of the repu rchase announcement matters. At the announcement date, managers intention of avoi ding dilution is significantly negative and enhancing shareholder value is significan tly positive, as expected. However, more interesting results are observed at two-y ears and three-years post announcement where I show that counteracting dilution is not a good reason to conduct a repurchase and, although not as strongly nega tive, enhancing shareholder value does not bear out its announcement promise. Furthermore, I find that firms that repurchase their shares to finance an acquisition are well compensated for th eir efforts, especially in the long run. I
iv attribute their success to higher cash flows re sulting from reducing their tax burden with their amortization deduction of the goodw ill created from the purchase accounting acquisition.
1 Chapter 1 Introduction It is well documented that when firms announce repurchase intentions, their stock price, on average, increases dur ing the repurchase announcement window1 and the increases are persistent.2 A common explanation for this is that firms repurchase their stock when the managers believe that it is undervalued (Dittmar (2000)); thus, the repurchase activity signals undervaluation to the market. Another frequently suggested explanation for the positive stock price reac tion to a firms repurchase announcement is that a repurchase is a good use of the firms free cash flows (Jensen (1986)). It is also suggested that repurchases may also provide for a better distribution of cash than dividends because of their temporary and fl exible commitment (Jagannathan, Stephens and Weisbach (2000)).3 This is likely one reason w hy companies choose to distribute cash to shareholders as repurch ases over dividend increases.4 1 Dann, 1981; Comment and Jarrell, 1991; and Ikenberry, Lakonishok, and Vermaelen, 1995. 2 Conrad and Kaul, 1993, and Lakonishok and Vermaelen, 1990. 3 See also Guay and Harford, forthcoming JFE. Lie, 2000, finds positive stock price reactions related to self-tender offers and also la rge special dividends and not for regular dividend increases. 4 Liang and Sharpe, 1999, report that in 1997 and 1998, share repurchases by S&P 500 companies exceeded dividend payments to common stockholde rs. Also, non-bank S&P 500 firms tripled their repurchases from 1994 to 1998 reach ing $150 billion. Over the same peri od dividends rose only 35% to $115 billion. In another study, Ikenberry, Lakonishok and Vermaelen, 2000, report that between 1996 and 1998, more than 4,000 open-market repurchases were announced, which if fully completed, would amount to approximately $550 billion. During this same period cash dividends totaled $490 billion. Weston and Siu, 2002, show repurchases as a percentage of dividends growing from 31.4 percent to 68.1 percent from 1994 to 1998. Despite different data sets, the empirical evidence establishes a high er growth in repurchases than dividends over the 1990s.
2 The increased use of open-market repurchases has coincided with an increasing reliance on stock options to compensate top managers. Stock options encourage managers to choose repurchases over conventional dividend payments because repurchases, unlike dividends, do not reduce the stock price (J olls (1998)). In a 1994-1998 sample of S&P 500 firms, gross repurchases reduced shares outstanding two percent annually; but, owing to the exercise of employee stock options, onl y about half of those shares were actually retired (Liang and Sharpe (1999)). Thus, it appears that repurchases are not only announced to signal undervaluation and as an appropriate use of fr ee cash flow, but may also be conducted to cover options prev iously committed by the firm. Kahle (2002) suggests that if firms are repurchasing shares to fund employee stock options, then in an efficient market the announcement period retu rn should not be as positive as if the repurchase were due to undervaluation or fr ee cash flow. Signaling undervaluation or an effective use of cash flow are well-docum ented viable hypotheses that support the positive stock price reactions observed with the repurchase announcement. Since it is well known that on average st ock prices increase after a repurchase announcement, it is possible that firm mana gers announce repurchases for opportunistic reasons. A firm managers options would increas e in value if the stock price increased at the mere mention of a repurchase. Furthermor e, an increase in stock price would support more favorable terms for an acquiring firm in a stock-financed acquisition. I propose that not all repu rchase announcements carry the same message. With this in mind, I propose to contribute to th is increasingly important payout choice by
3 disentangling the repurcha se announcement and distingu ishing between a good news repurchase and a no good news repurchase.5 In chapter 2, I use standard event-study analysis to investigate the stock price return reaction to firms announcing a repurch ase for possibly opportuni stic reasons such as to facilitate an acquisiti on, to counteract dilution e ffects and to cover options. Consistent with others, I find persistence in positive abnormal returns; however, the possibly opportunistic reasons are less positive. Furthermor e, the repurchase purpose of counteracting dilution shows signi ficant negative results at tw o-years and three-years post announcement and although not as strong, enhanc ing shareholder value does not bear out its announcement promise. The strongest posi tive reason to conduct a repurchase is to initiate or to fund an employee stock option plan. Since it is very likely that opportunistic behavior is motivated by the level of executive ownership in the firm, I investigat e the return reaction while controlling for current ownership levels and also controlling for option ow nership level of the firms chief executive. Consistent with agency theo ry, I find the best abnor mal three-years post announcement return performance is with firms in which CEOs own one to five percent of the stock and are compensated with a medi um level of options. Unexpectedly, I find the best two-year return performance is w ith firms in which the CEOs own no stock and receive no options. This group also has the dist inction of being the s econd best performer 5 Harford, 1997, recognizes that repurchases afford managers with the opportunity to behave opportunistically. In his investigation of Dutch-auction and fixed-price tender offer he argues that managers who are also shareholders can choose to participate or not in tendering their shares. If they choose to hold they are essentially putting their wea lth at risk (especially if the sign al is false). Thus, Harford argues managers could choose to participate in overpriced of fers and not participate in underpriced offers. Using
4 at three years post announcement. This sugge sts that option granting and CEO ownership do not influence performance, or that the co st of the options outweigh the benefit of improved CEO performance. In chapter 3, I focus on firms that choos e to conduct a repur chase of their own stock in order to facilitate an acquisition. This activity seems puzzling in that if a firm has the cash available to repurchase its stock and thus could use cash directly for an acquisition. Thus, it seems odd that a firm shoul d take an extra tran sactional step to acquire another firm, which might result in a loss of time and corporate value.6 Also puzzling is the research that shows that cas h-financed acquisitions perform better than stock-financed acquisitions. At first glance it would appe ar that firms are taking on additional transactions and on average might perform poorer. I find that this is not the situation. Firms that take on the extra fina ncing step are well compensated for their efforts, especially in the long run. These fi rms have cash available and positive earnings, but on average have negative abnormal return s prior to their repurchase announcements. Thus, these firms are likely to be underv alued and therefore c hoose this method of financing to signal undervaluation in the mark et place. Furthermore, the stock acquisition step allows these firms to share risk with th e target firms, counter act the negative effects of dilution by repurchasing their shares first, and enjoy a tax advantage for their efforts. Most research to date has exclusively focused on the open-market repurchase. The Securities Data Corporation (1994 date ) now tracks Dutch-au ction (2% of all probit analysis Harford finds that managers do not behave opportunistically, but rather set terms that offer to maximize shareholder wealth.
5 repurchases announced in 2000), fixed-pri ce tender-offers (3percent), negotiated (4percent), and open-market negotiated (58pe rcent), in addition to often-studied openmarket repurchases (33percent). I will incl ude open-market repurchases, Dutch-auction and fixed-price tender offers in this study and control for the announcement by the level of option granting and the mo tivation of the repurchase as indicated by management. Furthermore, most research has not had the advantage of the last few years of data. Original research on repurchases and options was carried out through a long-period of a bull market. Due to the market downturns of th e past four years, I have the advantage of studying repurchase and options during both an increasing retu rn market and a decreasing return market. In order to accomplish this, I use the Securities Data Corporation Platinum database to determine that my sample of firm s to be those that have chosen to repurchase their shares by the boards announcement date (s), the type of re purchase conducted, and the firms stated reason for conducting. Information on executive compensation and option variables are taken from S&P ExecuComp database. Finally, the stock price return data is obtained from CRSP and many of my control variables from Compustat. 6 Most research has shown that stock-financed acquis itions decrease the market value of the bidding firm. See Bradley, Desai, and Kim, 1988; Lang, Stulz, and Walkling; 1989; Servaes, 1991; and Dennis and McConnell, 1986.
6 Chapter 2 When is a repurchase announcement not good news? In recent years, firms have disbursed more cash to shar eholders in the form of repurchases than in the form of dividends,7 thus the rationale for repurchasing acquires added importance. In this chapter, I investig ate the reported purpose of such repurchases to see if the repurchases carry the positiv e stock price reactions documented by others and I find that the underlying purpose matters. Since 1995, firms reported the following purposes for their repurchase: to enhance shar eholder value; to c ounteract dilution; to fund a stock option plan; to indicate undervalua tion; to fund an acquisition; to support an employee benefit plan; and for general busin ess purposes. As examples of purposes mattering, I find that at the announcement date managers intention of avoiding dilution carries significantly negative returns and I also find that enhancing shareholder value is significantly positive. Moreover, when the results are observed at two-years and threeyears post announcement they are even more interesting. For example, counteracting dilution is negative and thus is not a good reason to conduct a repurchase and although not as negative, enhancing shareholder value does not bear out its announcement promise at the two-year and th ree-year periods. 7 Ikenberry, Lakonishok and Vermae len, 2000; Liang and Sharpe, 1999.
7 The organization of this chapter proceeds as follows. The first part discusses the choice between repurchases and dividends, th eoretical underpinnings of free cash flows and undervaluation and the research results, the repurchase choice a nd the influence of management options. The second section develops the hypotheses and methodology. The third section reports the empirical findings a nd the last section summarizes and concludes the chapter. Literature Review Repurchase or Dividend Decision During the 1990s, firms chose to disburse more cash to stockholders in the form of repurchases than in the traditional form of divide nd payments. In 1997 and 1998, in fact, share repurchases by S&P 500 companie s exceeded dividend payments to common stockholders. In fact, non-bank S&P 500 firm s tripled their repurchases from 1994 to 1998 to $150 billion. Over the same period divi dends rose only 35 percent to $115 billion (Liang and Sharpe (1999)). Grullon and Michae ly (2002) suggest that repurchases are an equivalent substitute for dividends. They show that the market reaction to dividend cuts is not significantly different from zero for firms that also repurchase their shares. Grullon and Michaely argue that repurchase program s should be superior to dividend payouts because repurchases disburse cash in a way that reduces shareholder tax liability. Repurchases also offer the firm flexibility in making payments and as such there is no long-term commitment associated with the methods of dis bursement. A dividend increase suggests a permanent plan for disbursement. Jagannathan, Stephens and Weisbach (2000) suggest that dividends ar e paid by firms with higher permanent
8 operating cash flows and repurchases are paid by firms with higher temporary nonoperating cash flows. Free Cash Flows or Undervaluation Many researchers have documented abnormal positive stock price returns to firms that repurchase their shares.8 These researchers have f ound that positive cumulative abnormal returns (CARs) occur in both shortterm and long-term studies. The short-term reaction could result from the firms conveyi ng revaluation information to the public. Chang (1993) suggests that repurchases are a cr edible informational signal if managers know more than investors and information is costly. The new information could indicate that the firm has free cash flows and purchas ing its own shares is a good investment, or that the firm believes that its shares are undervalued and purchasing shares is a rational investment in a positive ne t present value project. Researchers have put forth arguments s uggesting both that repurchases signal undervaluation and also that repurchases are an appealing use of free cash flows.9 For example, Ikenberry, Lakonishok and Verm aelen (1995) justified announcement date abnormal returns of 3.5 percent for open-mark et repurchase announcements to managers 8 Dann, 1981; Dann, Masulis, and Mayers, 1991; Ikenberry, Lakonishok, and Vermaelen, 1995; Lakonishok and Vermaelen, 1990, and Vermaelen 1981.Er win and Miller, 1998, show that in addition to positive stock price reactions for firms announcing re purchases, they also fi nd negative stock price reactions to rival firms. 9 Bagwell, 1991, presents another explanation by showing that an upward sloping demand curve exists. Thus, when a firm repurchases its shares, given heterogeneous valuations, shareholders with the lowest valuations will sell and the remaining shareholders w ill have higher valuations and the stock price must increase.
9 claims to repurchase their stock because prev ailing market prices are too low. Thus a repurchase is a good investment. The other hypothesis which explains pos itive stock price reactions is that a repurchase is an agency mitigating, and thus effective use of free cash flows. Jensen (1986) explains the problems of firms free cas h flow as follows: It would be optimal if managers owned 100 percent of the firm. Howe ver, due to the lega l structure in the United States we find that 50 percent of firm s have very broad ownership. Jensen puts forth the often-cited hypothesi s that given separation of ma nagement and ownership there will be agency costs. It is in the managers best interest, for example, to increase the value of his or her personal options by influencing the current market price of the stock. Obviously, this might also benefit the shareholders. In addition to agency mitigating, firms that choose to repurchase their own stock with free cash flows may be, in fact, choosi ng their best investment opportunity. The free cash flow hypothesis would support the pos itive abnormal stock price reactions empirically found at the announcement of a re purchase. Thus, repurchases are generally considered a good use of free cash flow because the repurchase reduces cash, which in turn reduces agency costs. On the other ha nd, Song (2002) argues that the open-market repurchase is not a credible commitment because repurchase distributions are not mandated by law and can occur at any time and by any method chosen by the manager.10 10Currently the Security Exchange Commission (SEC) has few regulations regarding disclosures of repurchase activities. In a nutshell, Securities and Exchange Commission release nos. 33-7881, 34-43154, IC-24599, file no. S7-31-99 states that an issuer conducting a repurchase program need not specify the amounts, prices, and dates that it will repurchase its se curities; rather, the issuer can adopt a written plan when it is not aware of material nonpublic information. On, December 10, 2002 th e SEC issued a proposed amendment to its 10b-18 rule regarding providing a safe harbor from liability for manipulation when a
10 Song (2002) 11 adds that United States corporate law does not mandate corporations to pay out a certain level of cash distributions, and in fact managers have strong incentives to avoid payouts. For example, managers are more risk-avers e than shareholders because it is difficult to diversify their own human capital. Thus, risk-averse managers have strong incentive to hold extra money for unexpected future hard times. Song concludes that managers have the opportunity to furt her their own interests, which, in turn, may result in significant social costs that ma y offset any benefits of the repurchase. Ikenberry and Vermaelen (1996) suggest that the market w ould only view openmarket repurchases as credible if they we re firm commitments. Companies choosing to falsely signal must be forced to repurchase shares at prices above their true long-run value in order for there to be a cost to false signaling. Ikenberry and Vermaelen question firm repurchases its common stock in the market in accordance with the rules manner, timing, price, and volume conditions. The SEC notes that an issuer may have the incentive to manipulate the price of its securities and one way to positively affect the price is to purchase securities in the open market. This amendment would require repurchase disclosures to Regulations S-K and S-B and Forms 10-Q, 10-QSB, 10-K, 10-ksB, and 20-F indicating the total number of shares purchased for the previous quarter, the average price paid per share, the identity of the broker, the number of shares purchases as part of a publicly announced plan and the maximum number of shares th at may yet be purchased. The SEC is also proposing footnote disclosure of the principal terms of publicly announced repurchase plans including 1) the date of the announcement, 2) the shar e or dollar amount approved, 3) the expi ration date (if any), 4) each plan that has expired during the period, 5) each plan the issuer ha s determined to terminate prior to expiration, and 6) each plan the issuer has not purchased under during th e period. In fact, as long as firms follow the SEC guidelines as set forth in SEC Rule 10b-18, firms w ill receive safe harbor protection from liability from purchasing their own stock. The 10b-18 Rule (SEC file no. S7-50-02) applies to bids for and purchases of an issuers common stock by or for an issuer. The safe harbor does not confer absolute protection from all liability for purchases (e.g. purchases that are part of a plan or scheme to evade fe deral security laws) even if made in technical compliance with the Rule. Rather the safe harbor provides only that certain, specific provisions of the securities laws w ill not be considered to have been violated solely by reason of the manner, timing, price, or volume of such repurchases, provided the repurchases are made within the limitations of the Rule. The four conditions of the rule are 1) issuer to use a single broker or dealer per day to bid or purchase its common stock; 2) issuer cannot bid for or purchase at the opening and during the last half hour of trading during the day; 3) highest price bid or pay for must be set by independent market forces; and 4) issuers may effect daily purchases up to 25 percent of the average daily volume. The Safe Harbor is a guideline for corporations to follow to avoid being accused of fraudulent trading practices. The guidelines are not mandatory. 11 Song, 2002, develops the theoretical rationale to argue for regulation in open-market repurchasing.
11 whether the news of an open-market repurchas e program should be viewed as a credible information signal and further suggest that the firms true intention is unknown. Nohel and Tarhan (1998) investigate the motive for repurchase and suggest that it is possible for different firms to repurchase for different reasons. For example, some firms may signal a bright future (undervalu ation), while others distribute excess cash. Nohel and Tarhan look at operating performan ce changes to see if firms are undervalued and are choosing to signal this information. If this were the case, Nohel and Tarhan argue, we should find operating performance impr ovement relative to what was expected and thus management would be signaling cr edible information. Specifically, Nohel and Tarhan hypothesize that high-gr owth firms may use the repurch ase to signal investment opportunity, while low-growth firms may dist ribute cash instead. Sorting by Tobins q, they unexpectedly find significant improvement s in operating performance following the repurchases from the low-growth firms. N ohel and Tarhan find repurchasers outperform their matched-control firms by 23.3 percent fo r low q-firms but are significantly negative for high-q firms. They also find that levera ge increases after th e repurchase, but the results are driven by high-q sample of firms. They find that market -to-book values of low q firms remain lower than their control gr oup, and post-repurchase performance at low q firms is correlated with assets sales, wh ich supports the free cash flow hypothesis. In another study, Evans and Gentry (1999) not only find little improvement but actually find underperformance by repurchasing firms. They find that firms that do not repurchase their shares create more long-run grow th in value than firms that incorporate a buyback strategy. Specifically, they find that small and mid-size non-repurchasing firms
12 outperform firms that repurchase their shares. Evans and Gentry attribute these results to the productive net working capital and cap ital projects investments made by the nonrepurchasing firms. Contrary to other researchers, their fi ndings do not support the theory that share-repurchase programs are related to management signaling an increase in a firm's long-run performance in the market, nor that a repurchase strategy signals that shares are undervalued. Type of Repurchase D'Mello and Shroff (2000) show with an earning-based valuation model that 74 percent of firms announcing fi xed-price tender offers we re undervalued. This would indicate that firms using a fi xed-price tender offer are sign aling a credible undervaluation message. Thus, it appears that managers can credibly convey their perceived dollar amount of stock undervaluation by the type of repurchase offer they choose. Traditional fixed-price tender offers specify a single pur chase price in advance, a number of shares sought, and an expiration date. Dutch-auction offers also specify a number of shares sought but at a range of prices within wh ich each tendering share holder chooses his or her minimum acceptable selling price. Genera lly, because the Dutch auction allows the managers to obtain a relatively low (minimum) o ffer price, it follows that Dutch auctions should provide a weaker signal of market price undervaluation than an otherwise equivalent fixed-price tender offer; that is, Dutch auction offers with low minimum offer prices should not be as convi ncing a signal of undervalua tion as a fixed-price tender offer. Comment and Jarrell (1991), in thei r study with 1984-1989 data, find average
13 excess returns of 11 percent at the announcemen t date with fixed-price tender offers and 8 percent for Dutch auctions. Furthermore, they find that open-market repurchase announcements are met with only two per cent average excess return. Comment and Jarrells results support the signaling hypothesis. Firms that send the strongest signal by attaching a clear premium are met with th e greatest event-day stock price reaction. Li and McNally (1999) employ a conditional study framework to determine the characteristics of firms choosing to make tender offers over those making open-market repurchase offers. They find that tender o ffering firms have higher cash flows, higher dividend yields, poorer investme nt opportunities, great er volatility of re turns and larger insider share holdings. Li and McNally note that this is consistent with large shareholders having the ability to mitigate agency problems of financia l slack. They conclude that firms exhibit comparative advantage in their choice of repurchase method. Gay, Kale and Noe (1996) present a theore tical argument of the advantages of Dutch-auction repurchases over that of fi xed-price tender offers. They confirm the intuitive argument that firms pay more than is needed to buy back the desired number of shares when conducting a fixed-price tender offer due to over-subscriptions. The authors suggest that if firms use a fi xed-price offer, then there wi ll be an excessive wealth transfer from remaining shareholders to exi ting shareholders. This raises the question as to whether there are repurchase situations where the firm pays back too much for its stock. In another study, Hodrick (1999) finds that the firms that will face greater stock price elasticity are more likely to choose a Dutch-auction repurchase offer over a fixed-
14 price tender offer. Other characteristics of the Dutch-auction c hoosing firms noted by Hodrick are that they have la rger market capitalizations, smaller insider holdings, larger institutional holdings, lower tr ading volume and tend to repur chase fewer shares during the repurchase program. The most common type of repurchase is accomplished through open-market repurchases where a firm announcing th e open-market repurchase provides no commitment to carry out its announced repurch ase intentions. The firm can buy back as many, or as few, shares as desired over a peri od of time at varying current market prices. Stephens and Weisbach (1998) find that between 46 percent and 75 percent of firms complete their repurchase plans within one year and 74 per cent and 82 percent of shares of all firms in the sample are actually acquired.12 Their study finds share repurchases are negatively related to prior stock-price performance and positively related to levels of cash flow. Stephens and Weisbach suggest that firms purchase the stock when it is undervalued and defer when it is overvalued and thus firms choose open-market repurchases over Dutch-auction or self-tende r because of the flexibility in magnitude repurchased and the timing of the repurchase.13 In an out-of-sample test, Ikenberry, Lakonishok and Vermaelen (2000) studied Canadian firms. Canadian firms are required to report each month the number of shares they actually repurchase. Surprisingly, 12 Altobelli and Wiggins, 1998, that on average firms repu rchase more shares than originally authorized and conclude that the open-market repurchase announcement is a credible signal. They also find that firms actively issue shares while repurchasi ng so that the average decrease in shares outstanding is only about 20% of the number repurchased. 13 Cook, Krigman, and Leach, 2001, find that while there is considerable variation across firms, NYSE firms under-going open-market repurchase activities on average beat their benchmark whereas Nasdaq firms do not. This suggests that NYSE firms are able to plan the timing of their open-market transactions to their advantage.
15 Ikenberry, Lakonishok and Vermaelen find that about 25 percent of firms never purchase any of their shares and less than 5 percen t of firms fully complete their repurchase programs (overall the mean completion rate is 28.6 percent). If firms use repurchases to signal undervaluation then why do we not see more firms using fixed-price and Dutch-auction a rrangements, since these two methods should provide a stronger signal of undervaluation? Furthermore, Song (2002) suggests that open-market repurchases are not credible commitments since the offers can be canceled anytime without legal or market punishment. Thus, there is no cost to the insiders and if there are no costs, such announ cements cannot be viewed as a signal. However, in spite of the weak and questionable signal, we fi nd the predominant structure of a repurchase accomplished through open-market repurchases (see Table 1). Firms choose to conduct an open-market repurchase six times as fre quently as fixed-price self-tender and Dutchauctions combined. Comment and Jarrell (1991) pr edict that fixed-price self-tender offers should be the strongest signal of firm undervaluation followed by Dutch-auction. The weakest signal should come from open-m arket repurchases because they have no attached premium. It is possible that each type of repur chase implies a different message. Persons (1994) presents a repurchase choice model. He finds that Dutch-auction repurchases are more effective takeover deterrents, while fi xed-price repurchases are more effective signals of undervaluation. Persons resear ch sheds light on why firms choose between Dutch-auction and fixed-price tender offers but does not help to explain why the open-
16 market procedure has become the pref erred method of cash disbursement to shareholders.14 Management Compensation and Options In addition to reducing free cash flow s through repurchases, increasing debt obligations, or increasing firms leverage or risk levels in order to mitigate agency problems, a firm can tie managers compensation to performance.15 Almost a decade ago, shareholders, led by institutional investors, pressured corporate boards and executives to tie managers compensation to performance. Under this pay for performance ideal, many companies began aggres sive stock option plans fo r managers and employees. Awarding stock options to employees and ex ecutives not only tied pay to performance but was also perceived to mitigate the principal-agent problem. Thus, while repurchases increased during the 1990s, there was an increasing reliance on stock options to compensate top managers.16 14 Although open-market transactions are conducted six times as frequently as fixed-price tender-offers and Dutch-auction combined, the market value of the repurcha se is considerably less. The total market value of open-market repurchases from 1984-2001 is one-third the market value of the other two methods combined. 15 Mehran, 1992, investigates the firms capital structure with managements incentive plans, managements equity ownership and several monitoring proxies, and finds results consistent with agency theory. 16 Top 200 US companies allocated a r ecord 15.2% of their shares to empl oyee stock options as a percent of shares outstanding in 2000, compared to 7.5% in 1990 (Yang and Carlton 2002). Also executive equity holdings account for nearly 70% of CEO compensation and most of the shares are the result from the exercise of their stock options (P earl Meyer & Partners, an executive co mpensation consulting firm). Also Bowen, 1996, reports in the Wall Street Journal that the 200 largest companies reserved more than ten percent of their common shares to be awarded to managers, which is an increase from six percent six years earlier.
17 Employees arguably have some informati onal advantage and th ere is, most likely, value in fortuitous timing of option grants and exercises.17 Yermack (1997) finds that managers receive stock-option grants shor tly before good news announcements regarding earnings, and delay such gran ts until after bad news a nnouncements. Yermack suggests that options are not so much an incentive de vice but rather a covert mechanism for selfdealing. On the other side, compensation-base d options do have a few value decreasing differences, including vesting restrictions and compensation based options that are nontransferable. Employees also tend to exerci se options early (American options) leaving some value unrealized. Stock options can encourage managers to choose repurchas es over conventional dividend payments because repurchases, unlik e dividends, do not reduce the per-share value of the stock. Jolls (1998) finds that fi rms that relied heavily on stock-option based compensation are significantly more likely to repurchase their stock than firms that do not. Furthermore, Jolls and othe rs attribute the growth in ope n-market repurchases to the increase in option grants.18 The granting of options to both managers of firms and employees has been lauded as a sound performance-based compensation plan for firms.19 Employees, managers, board members and sometimes consultants have been granted stock options, usually at 17 See Rothschild and Stiglitz, 1976, for discussion on imperfect information. 18 Arnold and Gillenkirch, 2002; Bens, Nagar, Skinner, and Wong, 2002; Fenn and Liang, 2001; Grullon and Michaely, 2002 and 2003; and Kahle, 2002. 19 Contrary to the argument that pay should be tied to performance, Elayan, Lau, and Meyer, 2001, find that companies which adopt incentive compensation schemes do not outperform companies without incentive compensation schemes.
18 the current market price.20 The general argument for pay-for-performance is that managers and employees will work harder if a portion of their pay is incentive-based. The goal for employees is to profit from in-the-money options, which is likely to occur when stock prices are high. Co mpanies presumably will pay smaller salaries and cash bonuses when options are substituted for direct compensation. Shareholders should benefit in this alignment of interest to increa se the value of the firm. In an ideal world, as the firms stock price increases, employees and managers receive higher compensation as their options move in-the-money. Shareholders also benefit as the value of their shares increases.21 Contrary evidence has been provided by Ye rmack (1995) in his investigation of why companies award stock opti ons to their top managers.22 He tests nine agency and financial contracting theories using Tobit and panel data analysis and finds little explanatory power with any of these prevai ling theories. Specifically, Yermack finds companies do not provide incentives from stock option awards in any significant association with the fraction of equity ow ned by the CEO. He also finds a negative association between incentives provided by stock options and the presence of growth opportunities which is counter to many other studies that suggest that firms with growth opportunities provide higher levels of CEO compensation. Furthermore, Yermack finds 20 According to generally accepted acc ounting standards firms only report the difference of option grant over the current market price as an expense during the year granted. As long as companies issue options at market price or out of the money, no expense needs to be recorded. Thus firms generally choose to issue at market. 21 David Aboody, 1996, analyzes the value of employee stock options (ESO) and finds ESOs to be negatively correlated with the firms stock prices. In early option vesting years there is a positive effect on firm value, but vested options are considered a net cost to the firms shareholders.
19 no significant association betw een financial leverage and in centives from stock option awards despite prediction from John and John (1993). Thus, although pay for performance has been argued to be the optim al compensation structure, it appears that there is a general absence of management incentives in CEO compensation packages. Interestingly, during most of the 1990s bull market, executiv es received huge compensation packages in the form of opti ons even when their companies lagged behind stock market averages, reports the Wall Street Journal.23 Executives and employees were happy and investors were satisfied as long as the stock prices c ontinued to increase. Unfortunately, there is no clear evidence that options generate be tter performance of a firms earnings, suggests the Wall Street Journa l. Also, it appears th at executive behavior changed in other less productive ways during that period, adds the Journal.24 Thus, although the argument for performance-base d compensation is sound, the performance measure may need adjustment (for example, tied to an index) in order that the compensation costs do not outweigh their benefits (Gillan (2001)) and to force the firm to conduct ongoing repurchases to m eet the option exercise demand. 22 Blasi, Kruse, and Berstein, 2003, argue that options can be effective only if they are granted to all employees and not just the top managers. 23 Lee, Susan, 2002, The Ugly Option, Wall Street Journal (New York). 24 One less productive management action has been the repricing of options. A consequence of later falling stock market prices was to make many employee and executive stock options essentially worthless, or to have been pulled underwater. This means that the exercise price of the options has fallen below the current market price of th e underlying security. Many employees ha d accepted compensation packages that included less salary, in hopes of large option payoffs. As the stock prices dived and the employees lost compensation, many may have chosen to relocate to another firm where they could receive an option package with a low exercise price at the new firm a nd ultimately a higher probability of realizing a gain. Managers of firms were thus under pressure to reprice the options in order to keep talented employees. Repricing options is a process of canceling existing outstanding options and reissuing at a lower strike price. Jin and Meulbroek, 2001, find that underwater options remain effective and conclude that restoring incentive alignment is not a sufficient reason for repricing options.
20 Kahle (2002) examines how stock options may have affected the firms decision to repurchase shares and finds that firms announce repurchases when executives have large numbers of options outstanding. So, during the 1990s as executive compensation increased in the form of options25 firms repurchased their stock at increasing prices. Bens, Nagar, Skinner and Wong (2002) substa ntiate this by finding that managers do repurchase shares to avoid the dilutive effects of employee stock option plans. Furthermore, the authors add that since repurchases involve paying out cash, thus reducing the future dollar return on that cash, the repurchase will ultimately result in reducing future earnings per share by reduci ng earnings. Thus, the original argument of initiating a repurchase to counter act dilution to presumably increase or at least maintain earnings per share may, in fact, in the long run decrease earnings per share. This is substantiated by Fenn and Liang (1998), who find negative relations hips between their proxies for investment opportunity and margin al financing costs a nd repurchases. Lie and McConnell (1998) state that when firms repurch ase shares to avoid dilution relating to option exercise there is a wealth transfer fr om shareholders to employees, since the cost of repurchase is much higher than the price at which empl oyees exercise their options. Klassen and Sivakumar (2001) argue that when repurchases are conducted because managers believe their stock to be underval ued, then positive information is conveyed to the market. A repurchase to reduce dilutive effects does not convey new information about future firm performance and on one hand should be irreleva nt. In this case we should not see any abnormal return reactions with a repurchase announcement to 25 Der Howanesian, Mara, 2002, The Buyback Boomerang, Businessweek.
21 counteract dilution. However, repurchasing sh ares does not effectively reduce economic dilution because the firm gives up a porti on of its assets to repurchase. Thus, repurchasing to avoid dilution may have the negative effect of passing up better investment opportunities. For these reasons, I suggest that repurchasing shares to counteract dilution should not result in positive stock price returns reactions and may, in fact, be negative, especially in the long run when the results of foregoing investment opportunities are realized. Klassen and Sivakumar (2001) note that the funding of stock option programs by repurchasing shares is an e xpensive strategy and suggest that option exercise could represent a real cost to the firm as a wealth transfer from shareholder to employee. The Economist (1/27/01) reports a study by Smithers and Co. that documents that if options had been accounted for at the time they were granted, the prof its of large-listed companies in 1998 would have been two-thirds lower. Klassen and Sivakumar note that repurchases increased from 19951999 when stock prices were soaring and then dropped during 2000. This implies firms were conducti ng buybacks as prices were increasing and stopped when their stocks became cheap. This is in contradiction to the often-cited undervaluation hypothesis. Yermack (1997) finds that managers rece ive stock option grants shortly before good news announcements and delay such gr ants until after bad news announcements. Yermack suggests that options are not so much an incentive device, but rather a covert
22 mechanism for self-dealing.26 Other researchers have suggested that managers behave opportunistically at the expens e of shareholders. For exampl e, Healy (1995) finds that firms are more likely to accrue discreti onary expenses during years in which their operating income exceeds the upper limits or falls below lower limits of managers' accounting-based bonus plans. In anothe r study, Jolls (1998) finds that option compensated managers substitute stock re purchases for dividend payments because managers normally do not share in dividends paid by the firm. Fenn and Liang (1998) study whether firms substitute repurchases for dividends when management options are at stake. They find that share repurchases are positively related to management stock options and divi dend increases are negatively related for the dividend-paying firms. Furthermore, Fenn and Liang found no statistical relationship between repurchases and management options for the non-dividend paying firms. In a similar vein, Lambert, Lanen and Larcher (1989) find that firms pay lower dividends after the adoption of stock option plans. Liang and Sharpe (1999) study S&P 500 firms to establish the effects of firms that repurchase and exercise stock options. Th ey find, in a 1994-1998 sample of S&P 500 firms, gross repurchases reduced shares outstanding two perc ent annually; however, owing to the exercise of employee stock opti ons, only about half of those shares were 26 Specifically, Yermack, 1997, finds that the average abnormal stock return to the CEO is $30,000 after 20 trading days and $48,900 after 50 days. As an asid e, options are awarded once a year by a compensation committee of the board acting under the authority of periodic shareholder votes. The options details are only disclosed in annual proxy statements, which could be as much as 15 months after the grant Shareholder votes usually occur once every five years and Yermack, 1997, reports that as of his paper a NYSE's proxy expert had no knowledge of one ever being rejected.
23 actually retired. According to Liang and Shar pe, when firms repurchase shares to avoid dilution relating to option exercise, there is a wealth transfer from shareholders to employees since the cost of repurchasing is hi gher than the employee ex ercise price. This would increase employee compensation and reduce the firms future net income, thus decreasing earnings per share available to th e stockholder. As an aside, a firm that chooses to issue new shares to facilitate opti on exercise would also experience a decrease in earnings per share due to the increase in num ber of shares rather than the decrease to earnings. In a September 2002 Business Week article, Der Howanesian (2002) suggests that stock repurchases can enrich executives comp ensation at investors expense. During the 1990s, cash rich firms purchased their stock at record high prices. Historically, buybacks were supposed to be a good use of free cas h flow and, as such, repurchasing activity made sense when a firms stock price was depressed, such as the period following the October 1987 crash. But, did it make sense dur ing the earlier 1990s when stock prices were booming? Thus, are buybacks simply a way for cor porate executives to maximize their own wealth, as Business Week suggests? Sin ce it is well-known that repurchase announcements are met with positive stock pr ice return reactions, can executives boost the price in the short-term and then sell thei r shares at a profit? This would have the effect of transferring wealth from the sharehol der or owners of the fi rm to the executives. If this is the case, shouldnt open-market repurchase announcements be met with possibly a negative, or at least non-positiv e, stock price return reaction?
24 Repurchases and Managerial Ownership Researchers have suggested that the agency problem between owners and managers can be mitigated if the managers have equity ownership. Morck, Shleifer and Vishny (1990) examine inside directors appointments and find that stock market reactions depend upon director ownership levels. Specifically, they find negative reactions when inside directors own less than five percent of the firms common stock, significantly positive when their ownership leve l is between five percent and 25 percent, and insignificantly different from zero wh en ownership exceeds 25 percent. Morck, Shleifer and Vishny attribute this to the a lignment of interests in the middle ownership range, but costs of entrenchment outweighing th e benefits of alignment of interests at high levels of ownership. Thus, it appears agency issues may be mitigated when managers own between five and 25 percent of the companys stock. At ownership levels of less than five percent, agency issues are a valid concern. McConnell and Servaes (1995) separate a la rge sample into high growth and low growth firms and investigate Tobins q, debt and equity ownership and find that firms with low (high) investment opportunity that q is positively (negatively) related to debt. McConnell and Servaes regress q with equ ity ownership and find some support that equity ownership is more important in low growth firms. This follows the Jensen (1986) argument that firms with poor investment oppor tunities are more likely to overinvest in negative net present value projects if free cas h flow is available. Concluding, many other researchers have investigated the relationship between corp orate value and the allocation
25 of shares among corporate insiders and while results differ there is a consensus that allocation of equity ownership matters.27 Hubbard and Palia (1996) propose the diversification-c ontrol hypothesis to address the impact of managers private bene fits of control as it relates to the bid premium paid in a merger. They develop a model and empirically show that 1) managers will indulge in non value-maximizing activit ies and will overpay the merger premium when the managers have a low ownership stak e; 2) managers and owners interest are aligned as the ownership stak e is increased and there is a negative relationship between the bid premium and managerial ownership; an d 3) at significantly high ownership levels managers are again willing to pay high premiums due to private benefits of control. This gives rise to a non-monotonic re lationship of first increasi ng (alignment of interest and value-maximizing behavior) and then decreasi ng as also shown by Morck, Shleifer and Vishny (1990). Ofek and Yermack (1997) argue that alt hough boards intend for stock options to boost the ownership positions of managers, there is no assurance that executive will behave accordingly. For example, modern por tfolio theory suggests that managers receiving additional stock should sell those shares or shares they already own in order to diversify away the unsystematic risk associat ed with concentrating wealth in a single asset. Furthermore, this risk is higher for managers than ordinary investors because of their human capital investment. Ofek and Ye rmack find that executives sell stock during 27 For example see Holderness and Sheehan, 1988, and Hermalin and We isbach, 1991, See also Lewellen, Loderer, and Rosenfeld, 1985, who find that return s to acquiring firms are positively correlated with the equity stake of the acquirer's top managers or low management ownership is associated with low returns.
26 years that they receive new options. Although consistent with modern portfolio theory, selling already owned shares counteracts the boards objec tives. Specifically they find that, on average, executives will sell 180 shares of stock for every 1000 new stock options awarded. Additionally, exec utives retain virtually none of the shares they acquire with the exercise of the options Thus, it appears that the gran ting of options may not have the intended purpose of aligning managers and stockholder objectiv es. Although option granting may not be the optimal method of turning ma nagers into owners, we do know that ownership matters..28 In the spirit of Morck, Shleif er and Vishny (1990), and of Raad and Wu (1994), I expect that the repurchase announcement should be pos itive for firms with high free cash flows, for it should align managerial and stockholders interests. However, if a firm has low managerial ownership and high free cash flows a repurchase coul d either help control the potential agency issues (good use of free cash fl ows) or it could be that the manager with significant options has the incentive to ma nipulate the firms stock price with a repurchase announcement. Another group which may have the incentiv e to manipulate and also the means to manipulate would be firms with high manage rial ownership (entrenched managers) and high free cash flow. This group may be able to manipulate the firms stock price in the short-run but may do poorly in the long-term. 28Morck, Shleifer and Vishny, 1990; McConnell and Servaes, 1995; and Rosenstein and Wyatt, 1997 Contrary to these and other researchers, Himmelberg, Hubbard, and Palia, 1999, use panel data to show that we cannot conclude that a firms performance is effected by managerial ownership.
27 At the other extreme are firms with high managerial ownership and low free cash flows. Firms with these characteristics shoul d not be susceptible to agency problems and thus should not require stock options to align shareholder an d manager incentives.29 In these firms the stock option grant may be an unnecessary and expensive compensation package. Furthermore, we may find the repurchase announcement to be non-positive because options may serve to exacerbate the entrenchment problem. I have made a theoretical argument that ow nership matters and different levels of managerial stock ownership and levels of op tion grants can provide different managerial incentives. However, I still may not be able to disentangle the opportu nistic behavior of managers. In other words, a self-serving mana ger who manipulates the price of the firms stock with an increase in market price is not necessarily acting against shareholders interests. In this case, it woul d be possible for everyone to gain. In reference to payouts a nd ownership stakes, Fenn and Liang (2001) show that the highest firm payouts (dividends and repurch ases) occur in firms that are more likely to have agency problems (for example, low executive ownership and high free cash flows). Jolls (1998) shows that the growth in re purchases is tied to the increase in options granted. It would appear that the firms that should derive the most benefit for granting options would be the firms with the greatest potential for agency problems. This would support significantly positive stock price reac tions of firms announcing repurchase offers to cover options if the firm also had low ma nagerial ownership and high free cash flows. 29 Although, these managers may not require options for agency mitigating problems, they may require options for what may be considered fair compensation in their industry.
28 Stock option grants may not depress current earnings directly; however, they will reduce future earnings. If a firm does not repurchase any shares, stock options will have a dilutive effect upon subsequent earnings per sh are. When a firm repurchases its stock, it reduces the number of shares over which earnings are divided. The cash used to finance the repurchase will reduce paid-in-capital, but will not directly affect earnings. Thus, some firms may have the incentive to comb ine on-going share repurchase programs with option programs to undo the erosion to earni ngs per share. Stock option grants should align shareholder objectives with firm managers objectives. However, there must be a trade-off between the benefit of aligning in centives and the cost of implementing or paying for the options as they become due. Fo r this trade-off reason, I expect to have a positive correlation between some lower level of option granting as incentives are aligned and compensation costs are not too high and a negative correl ation at some higher level of option granting as the costs of the gran ts outweigh the alignment benefit. As an example of a high level of option granting, Ba rrons Online reports th at Lehman Brothers issued 26 million options in 200230. In summary, there have been num erous studies surrounding repurchase announcements and the effects on firm perfor mance and stock price reactions. Although 30 Barrons Online reports that, The Street's most lavish options issuer continues to be Lehman Brothers. In 2002, it issued 26 million options, fully 10% of its shares outstanding, up from 21 million in the prior year. Lehman's reported compensatio n last year was 51% of revenues, but if the effective value of the options is factored in, that cost rises to almost 60% -a large wealth transfer from public shareholders to Lehman employees that isn't captured in the company' s income statement. Lehman spent heavily on share repurchases last year to offset its lavish option gran ts, acquiring $1.5 billion of stock, more than its $1 billion of net income.
29 some studies have suggested a downside to repurchasing activity,31 I am not aware of a study that attempts to disentangle the re purchase announcement to distinguish between the purposes of the repurchase announcement. My study will attempt to make this distinction by sub-sampling stock repurch ase announcements as specified in the following sections. Prediction, Data and Methodology Hypotheses Most research conducted on the repurchase announcement event finds that stock price returns are significantly positive on the event date and positive stock returns continue to persist for up to four years pos t the event date. These results are documented without questioning the motivation or the pur pose of the repurcha se. This research examines the purpose of the repurchase and I question whether a repurchase announcement is always viewed as good news and be met with significantly positive stock price reactions. Indeed, there are possibly opportunistic r easons for firms to repurchase their own stock. Firms may repurchase their own stock to fund an acquisition, to counteract dilution effects and to cover options. In very basic terms, when a firm grants options it will ultimately either have to increase its numb er of shares, which will reduce earnings per share, or it can avoid this potential earnings dilution by both granting options and repurchasing its shares. On the one hand, if option granting really does mitigate agency 31 Kahle, 2002; Yermack, 1997; and Evans and Gentry, 1999.
30 problems between the owners and managers of the firm and if repurchasing shares also has the desired effect to mitigate earnings dilution, then both activities could be positive news. On the other hand, it can be questione d whether repurchasing to counteract dilution is nothing more than an earning scheme by management. Thus, repurchasing to counteract dilution may maintain a constant number of shares outstanding (denominator in earning per share); however, earnings may be reduced (numerator) resulting in reduced earnings per share. Thus, when employee st ock options are granted the value of the existing owners stake in the firm is reduced. For these reasons, I do not think that a repurchase announcement should always be good news. Specifically, there are situations such as the funding of an acquisition, the covering of employee options, and the counter acting of dilution issu es, that should not necessarily be met with as positive stock pr ice reactions as those signaling undervaluation or enhancing stockholder value. Hypothesis 1 The abnormal return will be less-pos itive or non-positive for firms announcing repurchases for opportunistic reasons such as non-value-enhancing acquisitions, counteracting dilution effects, and co vering options as compared to the cumulative abnormal returns for all other repurchase motives. I have hypothesized that the return for some repurchasing activities will be less positive or perhaps non-positive. It is possible that this may not be immediately recognized or even distinguishable in the short run. Although the manager of a firm may be acting opportunistically, that does not neces sarily preclude that his or her actions are aligned with the interests of shareholders. In other words, if no difference in the stock
31 price reactions of different types of repurchase object ives are observed at the announcement date, it may be that these types of repurchases are st ill consistent with either the free cash flow hypothesis or th e undervaluation hypothesi s and distinguishing them may be an arduous task. Many researchers have shown the persistence in positive abnormal returns following stock repurchase announcements. This is a puzzle in that if a repurchase signals undervaluation or is an appropriate use of free cash flow, then positive announcement day reactions may be expected, but the reactions should not persis t threeto four-years post event.32 Furthermore, if the repurchases are conducted for opportunistic objectives, we should not see positive long-term abnormal returns. Hypothesis 2 The abnormal long-run returns will be less positive or non-positive for firms announcing repurchases for opportunistic reasons such as an acquisition, counteracting dilution effects, and the c overing of options as compared to the cumulative abnormal returns for all other repurchase motives. A lack of results or mixed results from hypotheses one and two could suggest that all repurchases on average are good news a nd positive stock price reactions should be anticipated. However, if that were the case firms should a nnounce repurchases even more frequently than they do, for there is no cost associated with a false signal.33 Another possibility is that opportunisti c repurchases do exist but thei r existence is dependent upon 32 Ikenberry, Lakonishok, and Vermaelen, 1995. 33 Jagannathan and Stephens, 2001, find that the market reactions to frequent vs. infrequent repurchase announcements are consistent with undervaluation. Infrequent repurchasers have lower market-to-book ratios suggesting that they are more likely to be undervalued, are preceded by negative abnormal returns, and are on average greeted with a stronger stock price return reaction than the frequent repurchasers. Thus,
32 managers ownership levels and whether or not managers and shareholders have similar goals. I suggest a positive relationship at low-op tion levels and a negative relationship at high-option levels. This is similar in spirit to the work by Rosens tein and Wyatt (1997) noted earlier. However, in their paper the co sts are due to entrenchment, whereas in this paper I am suggesting a wealth transfer. As a general clarification, compen sation-based options are different from exchange-traded options. One value-enhanci ng feature to the compensation-based option is its longer maturity date (five years ma y be typical). Thus, the value of options outstanding can be substantial. The number of options outstanding at the end of year is disclosed on firms proxy statements (s ince 1992); however, ba lance sheets do not include an allowance for this liability. In summary, I plan to test my hypothe sis using standard event study methodology using the constant mean retu rn and the market model to measure abnormal returns with the announcement of the repurchase. I plan to control the test with free cash flow variables, ownership variable s, executive option levels, grow th variables, firm size and the frequency of the repurchases by a firm. Sample The sample of repurchases is collected from Securities Data Corporations Mergers and Acquisitions database and Repur chases database. I begin by collecting all although the market generally views all repurchase s favorably frequent repurchasers are greeted less
33 open-market, Dutch-auction, and fixed-pric e self-tender offers with original announcement dates between January 1, 1988 and December 31, 2002. The Repurchases database began data accumulation with 1994 re purchases. My sample of repurchase firms dating from 1994 through 2002 was collected from this database. Prior to 1994, repurchases were obtained from the Merger s and Acquisition database. I found a firm choosing to repurchase its shares to be any fi rm which made an acquisition of its same cusip number. Thus, any firm with the same cusip number for the acquirer and the target firm is considered to have made a repurchas e. The Securities Data Corporation database includes all corporate transacti ons involving at least five pe rcent of the ownership of a company where the transaction was valued at $1 million or more (after 1992, deals of any value are covered) or where the value of th e transaction was undisclosed. Financial firms (SIC codes 6000-6999) and regulated utilit ies (SIC codes 4910-4949) were removed because they are believed to face a diffe rent incentive structure around repurchase activity. Imposing these restrict ions results in an initial sample of 3999 firm repurchase announcements including 177 Dutch-auction, 373 fixed-price tender offers and 3,449 open-market repurchases. Table 2-1 shows desc riptive statics for my sample repurchase firms. Along with repurchase announcements a nd the type of repurchase conducted, I have also obtained several more variable s from the Securities Data Corporations Repurchase database which ha ve been available since 1995.34 These variables include a enthusiastically. 34 I use the sample dating from 1988 to show the frequency of open-market repurchases, Dutch-auction and fixed-price tender offers. However, all empirical analysis is con ducted with repurchases initiated between
34 purchase code, SIC codes, and the total shares repurchased in number, dollar value or as a percent of the common shares outstanding. Th e purpose codes indicate a firm-reported description of the reason w hy the repurchase was conducted. The codes are to support an acquisition, to avoid dilution of earnings per share, to support an employee benefit plan, to enhance shareholder value, to support a stock option plan and to signal undervaluation. These codes will determine the motivation and the effect of the different repurchase announcements. I sub-divided my repurchase sample into four levels of executive stock option value (zero value of exercisable options, low value (option value is less than 20 percent of the executives salary a nd bonus) medium value (value of exercisable options is between 20 and 150 percent of the executives salary and bo nus); and a high option level (value of exercisable options is greater th an 150 percent of the executives salary and bonus)). Slightly over 60 percent of my sample of firms granted zero options and thus I placed those firms in one group. The low, medium, and high option levels were established by placing one-third of the remaining firms in each group. The stated percentile of options granted levels are simply the way the groups fell from this arrangement. The option data is obtained from the Standards and Poors ExecuComp database. This database contains inform ation on executive compensation and ownership for the S&P 1500 companies (S&P 500, S&P Midcap 400, and S&P Smallcap 600 indices), beginning in 1992. Using this databa se, I can calculate th e total number of 1995 and 2002. Thus, all statistical testing is conduc ted with the sample collected through Securities Data Corporations Repurchase database.
35 options held by top executives, the number of exercisable and unexercisable options held by top executives, and the shares owned by the same executives. In addition to my hypothesis variables, I include severa l control variables. Table 2-2 shows both control and hypothesis variables and the expected sign of their estimated coefficients. Table 2-3 explains the definitio ns of the variables used in table 2-2 along with theoretical justifications for the predicted signs. Methodology The research design to test my hypothes es uses a market model using both ordinary least squares and Scholes-Williams beta estimation and constant mean return model to calculate abnormal returns. I use dail y data and my periods of interest are the pre-event window (-30, +2), the announcement window (-1, +1) and post-event periods of (+2, +30), two-years and th ree-years after the announcement.35 Market model and constant mean return model I use event-study methodology using the c onstant mean return model and the market model estimated by both ordinary le ast square and the method of Scholes and Williams (1977) to measure abnormal returns (see Thompson (1995))36 The market model abnormal returns are defined as Ajt = Rjt ( j + jRmt) where Ajt is the abnormal return (or prediction error) fo r the common stock of the jth firm on day t, Rjt is the rate of 35I conduct both a t-test and a z-test statistic to make inference about the significance of my results. Both of these test statistics are calculate with Eventus software (see appendix A)
36 return of the common stock of the jth firm on day t, j and is the ordinary least squares estimate of j is a parameter that measures the sensitivity of Rjt to the market index Rmt. The least squares estimation period ends 46 days before the event date and is 255 days in length. The equally-weighted and valu e-weighted market indexes are both used as benchmarks in my study. The average abnormal return AARt is the sample mean. The cumulative abnormal returns are calculated as CART1T2 = (1/N) Ajt. In earlier drafts I also calculated buy-and-hold abnormal re turns by compounding successive daily returns.37 However, this produced problematic resu lts with questionable biases and thus those results have been omitted.38 I will show cumulative average abnormal returns for a pre-announcement period (-30, -2), event-day (-1, +1), and three post-event periods (+2, +30), (+31, two-years), (+31, three-years). I will also use the market model with the Scholes-Williams beta estimation. The beta estimator is j = ( j bar + j + j +) / (1 + 2 m) where j bar is the ordinary least squares (OLS) estimate from the regression of Rjt on Rmt-1, j is the OLS estimate form the regression of Rjt on Rmt+1, and m is the estimated first-order autocorrelation of Rm. The market adjusted returns are simply the difference between the actual return and the return of the market or Ajt = Rjt Rmt. The definitions of the average and 36 Thompson presents an excellent discussion of empirical methods of event studies in chapter 29 of Handbooks in Operations Research and Management Science, Volume 9. The chapter cites numerous studies and provides details of the empirical methodology that I will follow. 37 The buy-and-hold abnormal return is calculated with the BUYHOLD command in eventus. Eventus computes buy-and-hold abnormal returns by compounding successive daily raw returns and market index returns, then adjusts the raw returns according to the abnormal return method used. The calculation for the buy-and-hold abnormal return follows: 1 ) 1 ( ) 1 1 2 ( 1 1 ) 1 (2 1 ^ ^ 2 1 2 11T T t mt j T T t jt T jTR T T j R BHAR 38 See Canina, Michaely, Thaler, and Wom ack, 1998 for a complete discussion.
37 cumulative average abnormal returns follow those for the market model abnormal returns. The abnormal returns are calculated with Eventus software. Cross-sectional model Theoretical models often suggest that ther e should be an association between the magnitude of abnormal returns and charac teristics to the ev ent observation. To investigate this, I will use a cross se ctional regression of abnormal returns on characteristics, where Y, the dependent variable, is the observed CAR and X, the independent variables, are a matrix of charact eristics. My X variables are shown in table 2 with their predicted signs. The general re gression equation is that the announcement day excess return = + j (characteristics). The cross-sectional model will be carried out through SAS. This requires separate regres sions for each year. The general regression equation: CARit = 1 + KTKi + 2 D2i + 3 D3i + L PLi + 2 X2i + it Where TK = time series dummies from 1988 through 2002 D2i = 1 if Dutch-auction, 0 otherwise D3i = 1 if fixed-price, 0 otherwise PLi = purpose codes (ULV, DIL, ACQ, ESV, EBP, STP) X = all other firm characteristics (table 2) Market-to-book asset ratios also serv e as a proxy for investment opportunities (see e.g. Opler and Titman (1993)). All else be ing equal, higher market values suggest that the firm is not undervalued and, in fact, would make the repurchase of shares a more expensive undertaking. Thus, a high market-to-book (often referred to as a glamour stock) would suggest a negative correlation with repurchases. Alte rnatively, a low bookto-market would predict a positiv e correlation with repurchases.
38 Repurchases may also be the choice cash distribution for temporary cash flow availability and not a permanent cash flow in crease that would be more appropriately signaled by a dividend increase.39 Thus, if a firm chooses to repurchase rather than increase dividends, it may indicate only a te mporary cash flow increase and not the more desirable permanent cash flow increase. Previo us research has used debt/assets, book-tomarket, net operating cash flow/assets and payout ratios (common di vidends/net income) as controls. I will also include these controls. High debt levels may make managers more reluctant to distribute current cash flows. As leverage increases, the probabil ity of financial distress and hence external financing costs increases. Reducing debt is an alternative use of disgorging free cash flow;40 thus firms that rely more on debt will be less likely to repurchase. I will control for this with the long-term debt to assets variable. Firms with high levels of free cash flow ar e at a greater risk of overinvesting and hence derive greater benefits from repur chasing their own stock. Also, firms with relatively low marginal financing costs can di stribute more cash to shareholders knowing that if they must raise external funds in the future they will be ab le to do so relatively inexpensively. I include size as a proxy for financing costs and information asymmetry. I also include year indicator va riables to control for exogenous differences in repurchases that occur over the sample period. I also control for the frequency of the repurchase announcement.41 39 Jagannathan, Stephens, and Weisbach, 1998; Guay and Harford, 1999; and Gelb, 2000. 40 Jensen, 1986, and Berger, Ofek, and Yermack, 1997. 41 Jagannathan, Murali, and Clifford P. Stephens, 2003, Motives For Multiple Open-Market Repurchase Programs, Financial Management 32, 71-91. find frequent repurchasers to be much larger, have
39 Announcing a repurchase to signal firm undervaluation conveys good news to the market. If undervaluation is the message be ing conveyed, managers can clearly signal this with the premium offered. I have include d this distinction by th e type of repurchase offered.42 I expect all of the free cash flow contro l variables and the variables that suggest that the firm is undervalued to have positiv e estimated coefficients. If the underlying reason for the firm to make the repurchase de cision is value-enhanc ing, it should be met with positive stock price reacti ons. This would be consistent with past research studies.43 I include these variables to show both consis tency with other research and as control variables to help distingu ish between good news and no news or bad news repurchase announcements. Firms also self-report that repurcha ses are being carried out to enhance shareholder value. This term is ambiguous The firm could be either signaling undervaluation or conducting a repurchase as an effective use of free cash flows. If firms are credibly announcing a repurchase to enhan ce shareholder value, then the repurchase should be met with a positive stock price reaction. Results significantly less variation in operating income and have higher dividend payout ratios than infrequent repurchasers that are initiated by smaller firms with potentially higher degrees of asymmetric information. This suggests that smaller firms are more likely to repurchase for the positive motivation of signaling undervaluation. 42Comment and Jarrell, 1991 document that the stronge r the premium offered for shares, the stronger the signal of undervaluation and measure the undervaluation signal by the type of repurchase. 43 Dann, 1981; Vermaelen, 1981; Asquith and Mullin s, 1986; Ikenberry, Lakonishok, and Vermaelen, 1995; Stephens and Weisbach, 1998.
40 Table 2-4 presents the cumulative abnorma l returns (CARs) for the opportunistic purposes of funding a stock option plan, for an acquisition and to c ounteract dilution. The CARs for the Pre-event window, at the a nnouncement date and three post-announcement windows are shown using the ordinary leas t squares market model and the ScholesWilliams market model in comparison to the equal-weighted and value-weighted CRSP market indexes. Consistent with others research, the pre-event abnormal returns are consistently significantly nega tive for all models and repurchase purposes. Past research has shown significantly positive announcement day returns. My research also shows consistent positive returns. However, Hypothesis 1 is supported in that the abnormal returns are less positive for the opportunistic repurchase plans. For example, with the market model the (-1, +1) window for the nonopportunistic repurch ase plans has a CAR of 3.04 percent, whereas the stock-option repur chase plan, the plan to counteract dilution, and the plan to fund an acquisition have CARs of 2.22percent, 1.57 percent, and 1.82 percent, respectively. Consistent with others, I generally find persistence in positive abnormal returns. Kahle (2002) found that firms that repurchased shares to fund employee stock options were not as positive as firms that did not repurchase to cover employee options. Contrary to her work, I found that th e repurchases for funding an option plan performed well. Kahle compared firms that repurchased thei r shares with firms that increased their dividends during a four-year period. I choose to use a more robust sample of firms over an eight-year period that repur chased their shares for the purpose of funding an option plan, and compared my sample with both th e CRSP equal-weighted and value-weighted
41 benchmark portfolios. I found the market model equal-weighted CARs were 23.41 percent and 32.60 percent for the two-year a nd three-year post ev ent windows. These results were all significant and very similar to the non-oppor tunistic plans (21.00 percent and 31.77 percent for the two-year and threeyear post returns). My hypotheses proposed that opportunistic repurchase plans should not perform as we ll as undervaluation or free cash flow motivated plans. Although origina lly classified as an opportunistic plan, employee stock option plans are not serving on ly the firms executives interest, but may be serving all the firms employees. Later tables investigate the CEOs options and ownership separately in orde r to distinguish opportunistic repurchases in a clearer framework. Loughran and Vijh find that stock acqui rers earn 24.2 percen t less than their matched firms, on average, using buy-and-hol d returns over a five-year period; whereas cash acquirers earn 18.5 percent more than their matched firms. Rau and Vermaelen (1998) find that bidders in mergers underperform for up to three years after the merger is complete. Thus, I hypothesized that firms that used a repurchase plan to facilitate an acquisition should not perfor m as well as the non-opportunistic repurchase plans. Although the shorter postevent time windows were less pos itive (non-significant), it is interesting that the post two-year and th ree-year returns were significantly positive (CARs of 29.42 percent and 41.49 percent, respectively). Thus in the long-term, using a repurchase plan to finance an acquisition can be beneficial to the bidder and appears to act more like a cash-financed acquisition. This will be explored further in the next chapter.
42 Finally, the repurchase plan to counteract dilution reacted as predicted. All postevent day CARs are small and not significantly different from zero. Furthermore, the two-year (-2.51 percent) and th ree-year (-0.62 percent) are ne gative. Thus, to counteract dilution does not carry with it the positive abnormal returns that other repurchase purposes carry. In order to further differentiate the purpos e of repurchase, I i nvestigate the levels of free cash flows in my sample of firms and also the CEO ownership levels, and present the results in table 2-5A and 2-5B. These tabl es look at the difference in mean CARs for the pre-event window, at the announcement da te and a short window of 30 days post announcement, and finally two-years and th ree-years post announc ement. Table 2-5A presents a 3X5 matrix of free cash flows (hi gh, medium and low) and ownership levels (0 percent, less than 1 percent, 1 percent to 5 percent, 5 percent to 25 percent, and greater than 25 percent). The event-day CARs are in terestingly negative fo r the high free cash flow and CEO stock ownership of greater than 25 percent (-4.82 percent), the medium level of free cash flow and greater than 25 percent ownership (-0.48 percent), and high free cash flows and ownership level between fi ve percent and 25 percent (-0.45 percent). The p-values of .006 and .000 for the two-ye ar and three-year indicate that these groups are significantly different from each other. At two-year post announcement the high free cash flow and five percent to 25 percent ownership group has a CAR of -47.5 percent while there are strong positive CARs for low free cash flows and five percent to 25 percent ownership (+86.6 percent) and low cash flows and zero ownership (+68.8 percent). At three-year post event the high free cash flow a nd five percent to 25 percent
43 ownership group has a CAR of -44.2 percen t and the high free cash flows and one percent to five percent group has a CAR of -5.23 percent, while there are strong positive CARs for low free cash flows and five per cent to 25 ownership (+157.8 percent) and low free cash flows and zero ow nership (81.5 percent). Table 2-5B presents a 3X3 matrix of fr ee cash flows (high, medium and low) and ownership levels (high, medium and low). This table shows significant CARs at two-year and three-year post announcement. The only negative CARs shown are for the high free cash flow and low ownership group (-4.1 percen t and -2.2 percent for post two-years and three-years respectively). Arguably this group has the most likely agency problems. The highest positive or best performing group is the low free cash flow and high ownership group (45.3 percent and 66.3 percent). This hi gh ownership group presumably has little agency problems and thus is conducting the re purchase with both owners and managers objectives. Since cash flows are low, this gr oup is likely to have a strong motivation to make the cash draining decision to repurchase it s shares. It is likel y that this group is making a very credible announcement that its current market price is low. I noted earlier the somewhat surprising result where repurchasing to fund an option plan has very similar re sults to all the non-opportunistic repurchases plans. Tables 2-6 and 2-7 explore CEO options further to distinguish between funding for an employee option plan, which may not be opportunistic and the level of CEO options, which may lend itself to opportunistic motivations. In table 2-6 the repurchase sample is gr ouped by the value of the exercisable stock options held by the CEO divided by the CEO s salary and bonus. Four groups were
44 determined: no options held, low option value (v alue of exercisable options is less than 20 percent of the CEOs salary and bonus), medium option value (value is between 20 percent and 150 percent) and hi gh option value (great er than 150 percent). The results in table 2-6 are not significant; however, it should be noted that the highest CARs in the long term are shown for the medium option leve ls. It may be that th ere exist an optimum level of option grants; high enough to motivate th e manager to act in the owners interest and low enough to keep the cost of the op tion less than the mo tivational benefit. Table 2-7 presents a 5X4 matrix of CE O ownership levels (zero, less than one percent, one to five percent, five to 25 percent and greater than 25 percent) and option levels (no, low, medium and high). Panel A of this table shows significant CARs mean differences at two-year a nd three-year post announcemen t. The groups with negative CARs at two-year post repurchase announcemen t are: five-to 25 pe rcent ownership and medium options (-54.8 percent); five-to-25 percent ownership and low options (-51.5 percent); one-to-five percen t ownership and zero options ; (-9.9 percent), and zero ownership and no options (-0.1 percent). Due to some small sample size groups, the one to five percent ownership and zero options (n=122) is the onl y group with negative abnormal returns from which we can state that its negative returns are significant. The one-to-five percent ownership and zero options group also shows the same lackluster returns for the three-year post period of 10.8 percent. It is interesting to point out that the motivation to repurchase stock for any group s receiving no options cannot be to cover options. Thus the poor performance of the one -to-five percent group with no options is not a reflection of repurchasing due to the necessity of covering option commitments.
45 Focusing on just the two-y ear and three-year signifi cant periods and only groups with greater than 20 observations, it is intere sting to note that the best two-year return performance is the zero ownership and no option group. This group also has the distinction of being the second best performers at the three-year post announcement. This is consistent with option granting and CEO ownership not influencing performance or possibly the cost of the options does not outweigh the benefit of improved CEO performance. In other words it may be that op tion granting may increase the effort of the CEO, but not enough to affect the bottom line. Th is is an area that requires future study. The worst return performance groups CEOs ow n one to five percent of the corporations stock and receive no options. Consistent with agency theory, the be st group performance three-years post announcement is the one-to-five ownership group receiving a medium level of options. I expect this groups interests w ould be well aligned with those of the shareholders because the managers have a moderate level of ow nership, but not too much to be guilty of entrenchment and also receive a moderate level of options that should motivate the managers without the options costing so much too outweigh the benefit. Strong positive abnormal returns (two-y ear 52.4 percent and three-year 65.1 percent) are found for the 1 percent to 5 percent CEO ownership and middle option group (n=34) and (two-year 39.3 percent and threeyear 56.4 percent) for the zero ownership and no option (n=81). In order to address the probl em of some of the small sample-sized groups, panel B of table 2-7 shows just seven groups form ed by combining the medium and low option
46 groups, including the zero ownership groups wi th the less than one percent ownership groups, and blending the one to five percent ownership and five-to-25 percent ownership into a one-to-25 percent ownership group. The be st performances are seen by the greater than 25 percent ownership group with two-year and three-year abnormal returns of 47.3 percent and 41.2 percent respectively and the one-to-25 percent ownership and the medium and low option group with two-year and three-year abnormal returns of 44.4 percent and 28.7 percent respectively. This group would be the optimal alignment of managers and stockholders interests group. Un expected performance is observed by the one to 25 percent ownership with no option gr oup with two-year a nd three-year abnormal returns of 0.4 percent and 9.7 percent, respec tively. Thus, it app ears that not only does the existing ownership level matter, but it is also necessary to provide future incentives in the form of modera te option grants. Much of the recent literature has focu sed attention on open-market conducted repurchases due to the growth in numbers and the increase in the value of the repurchases. Researchers have suggested th at the repurchases are substitutes for dividends or are the direct result of compensation-based option grants and the needs of the firms to cover their options and avoid d ilution effects. There is a general consensus that the increase in open-market repurchases is due at least in part to the growth in option grants.44 Thus, it is likely that the incentives fo r open-market repurchases have changed over time. However, what has happened to Du tch-auction and fixed-price tender offers over time? During the 1990s the frequency of initiation has not increased as open-market
47 repurchases have proliferated. Do firms still elect to use these methods to signal undervaluation (see Comment and Jarrell (1991) ) or as a takeover deterrent (see Persons (1994)? Table 2-8 displays the cumulative abnor mal returns for the market model and Scholes-Williams model for the different types of repurchases, Dutch-auction, fixed-price tender offers and open-market repurchases Data is shown yearly from 1995-2002. Using the results from the value-weighted market model returns, open-market repurchasers perform the worst prior to the announ cement with a CAR mean over time of 7.17 percent and fixed-price offers perf ormed well with a mean over time of 4.32 percent. The at-announcement date abnormal returns for open-market repurchases range from a mean of 2.24 percent (1997) to 3.91 pe rcent (2000), the smallest sample year. Both fixed price and Dutch-auction carry the hi gher event-day returns (fixed price ranges from 2.93 percent (1996) to 19.67 percent (1995) and Dutch auction from 4.55 percent (1995) to 13.23 percent (2002)) and at firs t glance there is no trend over time. A review of the long-term abnormal retu rns for Dutch auction repurchases shows erratic returns ranging from 25.52 percent ( 2001) to 86.9 percent (2000) for the twoyear post returns and ranging from-21.98 pe rcent (1996) to 118.89 percent (2000) for three-years. The fixed price tender offers al so show a wide range of returns over the years. The erratic returns over years for the Du tch-auction and fixed price tender offers is most likely influenced by small sample sizes ra nging from as low as 6 events in a year to only a high of 24 repurchase announcements. 44 Jolls, 1998; Dittmar, 2000; Liang and Sharpe, 2000; Jagannathan and Stephens, 2001; Klassen and
48 Table 2-9 is a multi-model table addressing several hypotheses. The five panels A-E display the coefficients for pre-a nnouncement, at the announcement and postannouncement time periods (30 day, two-year an d three-year). Six fixed-effects models are displayed in panel A. Although my hypotheses did not address the cumulative abnormal returns prior to the repurchase announcement, this panel shows some interesting results. The share percent variable (aggregate shares held by the CEO divided by the number of common shares outstanding) is significantly negative. This implies that the greater the percent of shares owned by th e CEO the more negative the returns prior to making a repurchase announcement. The three-year least squares annual growth rate of net income variable is also significantly negative implying the higher the net income growth the more negative the cumulative abnormal returns prio r to the repurchase initiation. Not surprisingly, firms that a nnounce a repurchase due to undervaluation also have significantly negative CARs before th e announcement, whereas firms that plan to fund a stock option plan perform well. My hypotheses questions opportunistic re purchases at the announcement date. Panel B displays 8 models addressing the event date announcement CARs. A look at the significant estimates reveals the stock value (a ggregate value of stock options held by the CEO as a percentage of salary and bonus) is pos itive. This implies a direct relationship to the CEOs option value and the stocks retu rn reaction at the repurchase event. As predicted, firms that announce a repur chase to counteract dilution effect do poorly (significantly negative in models six and eight). Also, predicted the nonSivakumar, 2001; Bens, Nagar, Skinner and Wong, 2002; and Kahle, 2002.
49 opportunistic purpose of enhanci ng shareholder value is positive (significant in models one and two). I had also predicted, non-pos itive returns for both repurchases for acquisitions and for stock option plans. Although negative estimators appear in several models, none are significant. The strongest estimator is with firms announcing a Dutch-auction repurchase. Although the positive returns with a Dutch-auction is consiste nt with others research, it is not consistent to find the returns stronger than the fixedprice tender offe rs (negative, not significant) which carries a higher premium by virtue of its makeup. Panel C displays the short-term stock pr ice reactions from two days to 30 days post announcement. For this period both c ontrol measures of price to book were significantly negative. As with the event time period, stock value is slightly positive. In the short-term none of the repurchase purpos es show any significance. The frequency control is negative in one model and positive (not significant) in another. Dutch-auction and fixed-price repurchases show expected strong positive estimators; however, similar to the event day results the Dutch-auction fi rms have stronger positive returns than the fixed-price offers. Panels D and E show two-years and th ree-years post announcement, respectively. During these time periods, Dutch-auction and fixed-price repurchases no longer show any superiority in returns over open-market repu rchase plans. Generally for both long-term periods net income to assets is slightly ne gative, share percent is slightly positive and both income growth and earnings per share growth negative.
50 Overwhelming the purpose to counteract di lution is significantly negative in all models as predicted by the hypotheses. Howeve r, the hypothesis also predicted negative return reactions for firms funding stock option pl ans. This did not bear out and in fact by three-years post announcement funding the stock option plan was significantly positive. I had also predicted that the purpose of f unding an acquisition would have the negative results found with most acquisition research. This estimator although generally negative was only significant in one model. Two surprisingly results of slightly negativ e returns at three-years post repurchase announcement for the purpose of enhancing shareholder value and for the purpose of undervaluation. It appears that the managers purpose of repurch ase did not occur. In summary, the purpose to counteract dilu tion is strongly nega tive in all models. This supports the argument that opportunistic repurchases do not perform as well as other repurchases. Contrary to the hypothesis, I again find that stock option plans have significantly positive long-term cumulative abnormal returns. The acquisition motivated repurchase is negative, as expected, but not significant. In the long-term, whether the repurchase was completed with a Dutch auc tion or fixed price tender offer was not significant. Conclusion The purpose of the repurchase announcem ent matters. At the announcement date managers intention of avoiding dilution is significantly negative and enhancing shareholder value is significantly positive. However, more interesting results are observed at two-years and three-years post a nnouncement. Counteracting dilution is not a
51 good reason to conduct a repurchase and althoug h not as strongly negative, enhancing shareholder value does not bear out its announc ement promise. Consistent with a rich history of acquisition work, conducting an ac quisition in conjunction with a repurchase seems to carry the negative attributes of the acquisition-driven mo tivation. This will be explored further in the next chapter. The strongest positive reason to conduct a repurchase is to initiate or fund an employee stock option plan. Both employee be nefit plans and stoc k option plans carry positive abnormal returns well into future time periods. This is contrary to the results found by Kahle (2002). Although it was expected that stock options would fall under opportunistic CEO behavior, very different results were found. A probable explanation is that stock options not only benefit the CEO but also employees and outside shareholders. Furthermore, when the ownership structure of the sample of repurchase firms was investigated, a few unexpected results surfaced. For example, the largest positive abnormal returns were found for a subsample of firms in which the CEOs had no stock ownership and no option grants. It is not known whether these CEOs have an alternative form of incentive based pay or if the cost of the options si mply outweighs the benefit of improved CEO performance.
52 Table 21 Types of Repurchase Announcements Number of Repurchase Announcements Dutch-auction Fixed-price Open Market 1988 17 36 136 1989 16 56 279 1990 6 51 396 1991 3 52 115 1992 8 36 114 1993 5 37 260 1994 7 38 408 1995 9 8 493 1996 17 11 691 1997 21 14 664 1998 24 6 647 1999 18 14 415 2000 17 13 320 2001 9 9 177 2002 11 16 399 Total 188 397 5515
53 Table 22 H ypothesized Relationships Hypothesized relationships between repurchas e event return and firm characteristics Hypothesis Variable Predicted Sign Hypothesis variables Offset dilution Fully diluted eps/basis eps Dummy variable for firms reporting as purpose Acquisition Dummy variable fo r firms reporting as purpose Managerial CEO Equity Ownership Ownership Share Ownership Value Share Ownership Value + Stock Option Value Control variables Free cash Operating income before depreciation/assets + Net income before taxes and minority interest/assets + Net income before extraordinary items and discontinued operations/assets + Net income before extraordinary items and discontinued operations less preferred dividends/assets + Under-valuation Book-to-market + Dummy variable for frequent repurchasers Dummy variable for firms reporting as purpose + Signaling Type of Repurchase Fixed +++ Dutch ++ Open + Enhance Sh. Value Dummy variable for firms reporting as purpose + Risk Beta + Leverage: Long-term debt/assets + Long-term debt/market value + Debt/Assets + Other Controls Size Year ? Dummy variable for frequent repurchases
54 Table 23 Variable Definitions Acquisition dummy variable = dumACQ = SDC repurchase purchase code ACQ. Beta = CRSP beta Black Scholes option value = BlkVal = The a ggregate value of stock options granted to the executive during the year as valued using S&P's Black Scholes methodology (ExecuComp BLK_VALU). DA = Repurchase announcement was a Dutch-auction offer. Debt/Assets = Total debt (Compustat DLTT + LCT) / asse ts (Compustat TA) Dilution dummy variable = dumDIL = SD C repurchase purchase code DIL. Employee Benefit Plan dummy variable = dumEBP = SDC repurchase purchase code EBP. Enhance shareholder value dummy variab le = dumESV = SDC repurchase purchase code ESV. EpsDil-1 = the diluted eps/basi s eps for the year prior to the repurchase announcement. EpsDil_0 = = the diluted eps/basis eps for th e year of the repurchase announcement= Fully diluted eps/basis eps =Fully dilu ted reported since December 15, 1997 (APB opinion No. 15 and Financial Accounting St andards No. 128) Basic EPS is earning available to common shareholders divide d by the weighted average of shares outstanding. Diluted EPS increase the number of shares in the denominator to reflect the dilutive effects of convertible securities and stock op tions and adds back to earnings the interest payments that would not have to be made by the firm upon the conversion of bonds/preferred stock to common stock. EPS growth = EPSgrow = the three-year l east squares annual growth rate of Net Income before extraordinary items and disc ontinued operations less preferred dividend requirements (ExecuComp EPSEX3LS). Exercisable unexercised options = The num ber of unexercised options that the executive held at the end of the year that were vested options (ExecuComp UEXNUMEX).
55 Table 23 Variable Definitions (Continued) Exercised options = Soptexsh = the numb er of options exercised by the executive during the year (Exe cuComp SOPTEXSH). Exercised options value = Soptexer = the net value realized from exercising options. It is the difference in value between the exer cise price of the options and the market price of the company's stock on the date of exercise (ExecuCo mp SOPTEXER/1000). Exercised to total available options = Exercised options (SOPTEXSH) / Exercised options + exercisable unexercised (S OPTEXSH + UEXNUMEX) This gives the percentage of options exercised that were exercisable (vested). FP = Repurchase announcement was a Fixed-price tender offer. Frequent Repurchaser = Freq = Firm made more than one repurchase announcement during the year. Less Frequent Repurchases = Somefreq = Firm made more than one repurchase during the three-year pe riod, but not more than one during the year. Long-term debt/assets = long-term debt (C ompustat DLT) / assets (Compustat TA) Long-term debt/market value = long-term debt (Compustat DLT) / market value (Compustat MKVALF) Market value = MktVal = market value (Compustat MKVALF/1000). Net income before extraordinary item s and discontinued operations/assets = ExecuComp NIBEX/Assets Net income before extraordinary items a nd discontinued operations less preferred dividends/assets = NI/assets = ExecuComp NIAC/Assets. Net income before taxes and minority in terest/assets = ExecuComp PRETAX/Assets NI growth = net income growth = The 3-year least squares annual growth rate of Net Income (ExecuComp NI3LS) Operating income before depreciation/assets = ExecuComp OIBD/Assets Operating Income growth = In cGrow = The three-year least squares annual growth of Operating Income before deprec iation (ExecuComp OIBD3LS).
56 Table 23 Variable Definitions (Continued) Option grant value = Soptval = the aggregate value of all options granted to the CEO during the year as valued by the company (ExecuComp SOPTVAL). Payout ratio = Total common dividends (item 21) / Net income Percent Sought = Total share repurchased (SDC total shares repurchased) / Total shares outstanding (SDC Number of securities outstanding) Purpose Code: The code describing the pur pose of the repurchase program. Examples are: to enhance shareholder value (ESV) to offset diluti on (DIL) to support a stock option program (STP), to indicate unde rvaluation (ULV), to fund and acquisition (ACQ), for an employee benefit plan (E BP), and for general business (GEN). Price to book = The market value of assets divided by the book value of assets, where the market value of assets is the book value of assets plus the market value of equity (Compustat item #24 times Compustat item #25) minus the book value of equity. PrcBk-1 = Price to book for the year endi ng prior to the repurchase announcement. PrcBk_0 = Price to book for the year of the repurchase announcement. Share % = CEO equity = The aggregate shares held by the CEO divided by the number of common shares outsta nding (ExecuComp SHROWN/SHRSOUT) Share Ownership Value = Market value of the common shares held by executive divided by the executive's salary an d bonus (ExecuComp PRCC*SHROWN/TCC) Share Ownership Value + Stock Option Valu e = Market value of the common shares held by executive plus the aggregate value of stock options granted to the executive during the year as valued using S&P' s Black Scholes methodology divided by the executive's salary and bonus (Execu Comp (PRCC*SHROWN + BLK_VALU)/TCC) Stock Option Plan dummy variable = dumST P = SDC repurchase purchase code STP. Stock Option Value = StkVal = The aggregat e value of stock opti ons granted to the executive during the year as valued us ing S&P's Black Scholes methodology divided by the executive's salary and bonus (ExecuComp (BLK_VALU)/TCC). Type of Repurchase = SDC Technique code. OP = open market, DA = Dutch=auction, and FPOL = fixed-price tender-offer Undervaluation dummy variable = dumUVL = SDC repurchase purchase code UVL.
57 Table 24 Retu rns to Repurchase Purpose Cumulative Abnormal Market model re turns (CARs) for both equal and value weighted portfolios of all firms Announcing Repurchase Plans to Fund a Stock Option Plan (n=482, beta = 1.06) Market Model Scholes-Williams Market Model Window Equal Wt Value Wt Equal Wt Value Wt -30, -2 -3.82% *** -4.87% *** -4.14% *** -5.04% *** -1,+1 2.22% *** 2.02% *** 2.15% *** 1.97% *** +2,+30 1.98% ** 1.67% 1.93% ** 1.62% +31,+504 23.41% *** 24.20% *** 21.46% *** 23.14% *** +31,+756 32.60% *** 33.88% *** 29.96% *** 33.03% *** significant at .10, ** significant at .05, *** significant at .01 Cumulative Abnormal Market model re turns (CARs) for both equal and value weighted portfolios of all firms Announcing Repurchase Plans to Fund an Acquisition (n=96, beta = 0.75) Market Model Scholes-Williams Market Model Window Equal Wt Value Wt Equal Wt Value Wt -30, -2 -2.84% -2.67% -1.14% -3.64% -1,+1 1.57% ** 1.68% *** 1.67% ** 1.56% ** +2,+30 0.91% 1.97% 1.05% 1.27% +31,+504 29.42% *** 30.47% *** 49.41% *** 17.94% *** +31,+756 41.49% *** 45.92% *** 72.15% *** 26.75% *** significant at .10, ** significant at .05, *** significant at .01
58 Table 24 Returns to Repurchase Purpose (continued) Cumulative Abnormal Market model re turns (CARs) for both equal and value weighted portfolios of all firms Announcing Repurchase Plans to Counteract Dilution (n=126, beta = 1.27) Market Model Scholes-Williams Market Model Window Equal Wt Value Wt Equal Wt Value Wt -30, -2 -3.72% ** -4.33% ** -3.74% ** -4.27% ** -1,+1 1.82% ** 1.75% ** 1.75% ** 1.67% ** +2,+30 2.03% 1.68% 1.61% 1.20% +31,+504 -2.51% 3.07% -2.28% 2.54% +31,+756 -0.62% 6.16% -0.60% 6.63% significant at .10, ** significant at .05, *** significant at .01 Cumulative Abnormal Market model re turns (CARs) for both equal and value weighted portfolios of all firms Announcing Repurchase Plans for purposes other than to fund a stock opti on plan, to fund an acquisition or to counteract dilution (n=3814, beta = 0.94) Market Model Scholes-Williams Market Model Window Equal Wt Value Wt Equal Wt Value Wt -30, -2 -5.36% *** -6.44% *** -5.42% *** -6.53% *** -1,+1 3.04% *** 2.94% *** 3.05% *** 2.92% *** +2,+30 1.75% *** 1.83% *** 1.86% *** 1.57% *** +31,+504 21.00% *** 22.40% *** 21.53% *** 19.63% *** +31,+756 31.77% *** 34.84% *** 32.56% *** 31.23% *** significant at .10, ** significant at .05, *** significant at .01
59 Table 2-5 Free Cash Flow and Executive Ownership A. The share ownership percentage is th e number of shares held by the top executive (excluding stock options) divided by th e number of common shares outstanding (ExecuComp variables SHROWN/SRSOUT). I have determined five levels of executive ownership; ZeroOwn (executive owns zero sh ares), Less1Own (less than one percent), 1_5Own (between one and five percent), 5_25Own (between five and 25 percent), Great25Own (greater than 25 percent). Free cash flows are proxied with three calculations; operating income before depreciation divided by assets, net income before tax di vided by assets, and net income available divided by assets. I placed the highest 30 percen t of the free cash flow calculations in the high cash flow group, HCF; the next 40 percent of the free cash flow calculations in the medium group, MCF; and the lowest 30 percen t of the free cash flows in the low free cash flow group, LCF. In some cases the thr ee free cash flow proxies yielded a different classification. In such cases, the average classification was used. For example if one proxy calculation classified as high, another as medium and the final as low, the firm would be classified as MCF. Finally, 15 groups were form by taking members of each ownership group (5) and combining it with members of each free ca sh flow group (3). ANOVA follows for the group differences of the abnormal returns for fi ve event time periods. Groups are sorted from lowest to highest average abnormal return. Group n (-30, -2) (-1, +1) (+2, +30) (+31, +504) (+31, +756) HCF 1_5Own 63 -0.084 0.021 0.009 0.003 -0.052 HCF 5_25OWN 13 -0.087 -0.005 0.100 -0.475 -0.442 HCF Great25Own 6 -0.063 -0.048 0.008 0.106 0.336 HCF Less1Own 207 -0.085 0.008 0.018 0.015 0.069 HCF ZeroOwn 22 -0.068 0.023 0.021 0.568 0.645 LCF 1_5Own 32 -0.132 0.000 0.012 0.271 0.357 LCF 5_25OWN 14 -0.031 0.041 0.074 0.864 1.578 LCF Great25Own 5 -0.116 0.106 -0.003 0.477 0.356 LCF Less1Own 150 -0.048 0.017 0.035 0.206 0.453 LCF ZeroOwn 13 -0.087 0.038 0.031 0.688 0.815 MCF 1_5Own 70 -0.087 0.021 -0.001 0.082 0.112 MCF 5_25OWN 37 -0.049 -0.005 0.007 0.055 0.174 MCF Great25Own 12 -0.094 0.031 0.015 0.550 0.613 MCF Less1Own 306 -0.044 0.016 0.013 0.016 0.031 MCF ZeroOwn 31 -0.032 0.003 -0.020 0.046 0.138 P-value 0.453 0.086 0.783 0.006 0.000 Table 2-5 Free Cash Flow and Executive Ownership (Continued) B Ownershi p is p roxied with three calculations ; executive shares owned divided b y
60 common shares outstanding, the market value of the common shares held by executive divided by the executives salary and bonus, and the share ownership value plus the Black Scholes option value divided by the executives salary and bonus. The highest 30 percent of the executive ownership level were placed in the high ownership, HOwn; the next 40 percent in the medium group, MOwn ; and the lowest 30 percent in the low ownership group, LOwn. In cases where the th ree ownership proxies yielded a different classification, the average classification was used. Free cash flows are proxied by three calculati ons; operating income before depreciation divided by assets, net income before tax di vided by assets, and net income available divided by assets. The highest 30 percent of the free cash flow calculations were placed in the high cash flow group, HCF; the next 40 percent in the medium group, MCF; and the lowest 30 percent in the low free cash fl ow group, LCF. In some cases the three free cash flow proxies yielded a different classificat ion. In such cases, the average was used. Finally, nine groups were form by taking members of each ownership group (3) and combining it with members of each free ca sh flow group (3). ANOVA follows for the group differences of the abnormal return s for five event time periods noted. Panel A Groups n (-30, -2)(-1, +1)(+2, +30)(+31, +504) (+31, +756) HCF Hown 88 -0.077 0.006 0.020 0.108 0.218 HCF Lown 89 -0.081 0.013 -0.001 -0.041 -0.022 HCF Mown 134 -0.089 0.011 0.033 0.038 0.035 LCF Hown 53 -0.084 0.010 0.029 0.453 0.663 LCF Lown 80 -0.054 0.027 0.040 0.183 0.477 LCF Mown 81 -0.059 0.018 0.028 0.292 0.482 MCF Hown 138 -0.066 0.014 0.021 0.080 0.138 MCF Lown 112 -0.053 0.002 0.014 0.018 0.043 MCF Mown 206 -0.041 0.022 -0.005 0.038 0.056 P-value 980 0.539 0.385 0.417 0.087 0.037 Panel B Groups (+31, +504) (+31, +756) HCF Lown 88 -0.041 -0.022 LCF Mown 81 0.292 0.482 P-value 0.037 0.016 Panel C Groups (+31, +504) (+31, +756) HCF Lown 88 -0.041 -0.022 LCF Hown 53 0.453 0.663 P-value 0.013 0.007
61 Table 2-6 Executive Options Executive option value is determined by the va lue of the exercisable stock option divided by the total salary and b onus (ExecuComp variables SOPTEXER/TCC). I have determined four option value levels They ar e no or zero value of exercisable options, NoOpt; low value of exercisable options, LOpt (value of exercisable options is less than 20 percent of the executives salary and bonus ); medium value of exercisable options, MOpt (value of exercisable options is be tween 20 and 150 percent of the executives salary and bonus); and a high op tion level, HOpt (value of ex ercisable options is greater than 150 percent of the execu tives salary and bonus). ANOVA follows for the group differences of th e abnormal returns for five event time periods: 30 days to two days prior to the repurchase a nnouncement (-30, -2), the event period (-1, +1), 30 days post-announcemen t (+2, +30), two-years post-announcement (+31, +504), and three-years post-announcement. Panel A Groups n (-30, -2) (-1, +1) (+2, +30) (+31, +504) (+31, +756) Hopt 159 -0.070 0.024 0.008 0.132 0.215 Lopt 79 -0.070 0.018 0.026 0.164 0.249 Mopt 156 -0.044 0.006 0.035 0.183 0.308 NOopt 583 -0.068 0.014 0.013 0.054 0.116 P-value 976 0.579 0.191 0.409 0.436 0.358 Panel B Groups n (-30, -2) (-1, +1) (+2, +30) (+31, +504) (+31, +756) Hopt 159 -0.070 0.024 0.008 0.132 0.215 Lopt 79 -0.070 0.018 0.026 0.164 0.249 Mopt 156 -0.044 0.006 0.035 0.183 0.308 P-value 393 0.491 0.063 0.388 0.896 0.804 Panel C Groups n (-30, -2) (-1, +1) (+2, +30) (+31, +504) (+31, +756) Hopt 159 -0.070 0.024 0.008 0.132 0.215 Lopt&Mopt 235 -0.053 0.010 0.032 0.177 0.288 P-value 393 0.424 0.046 0.187 0.656 0.571
62 Table 27 Executive Ownership and Options The share ownership percentage is the num ber of shares held by the top executive (excluding stock options) divided by the num ber of common shares outstanding. Five levels of executive ownership are ZeroOwn (executive owns zero shares), Less1Own (less than one percent) 1_5Own (between one and five percent), 5_25Own (between five and 25 percent), Great25Own (greater than 25 percent). Executive option value is determined by the va lue of the exercisable stock option divided by the total salary and bonus. F our option value levels are no or zero value of exercisable options, NoOpt; low value of exercisable options LOpt (value of exercisable options is less than 20 percent); medium value of exercisable options, MOpt (value of exercisable options is between 20 and 150 percent); a nd a high option level, HOpt (value of exercisable options is greater than 150 pe rcent of the executives salary and bonus). Finally, 18 groups were form by taking members of each ownership group (5) and combining it with members of each option level group (4) (Two groups have no members.). ANOVA follows for the group differe nces of the abnormal returns for five event time periods: 30 days to two days pr ior to the repurchase announcement (-30, -2), the event period (-1, +1), 30 days post-announcement (+2, +30), two-years postannouncement (+31, +504), and three-years post-announcement. Panel A Groups n (-30, -2) (-1, +1) (+2, +30)(+31, +504) (+31, +756) 1_5Own HOpt 52 -0.084 0.013 0.046 0.167 0.294 1_5Own LOpt 15 -0.124 0.052 -0.044 0.462 0.694 1_5Own MOpt 34 -0.056 0.024 0.041 0.524 0.651 1_5Own NOpt 122 -0.081 0.013 0.002 -0.099 -0.108 5_25Own HOpt 10 0.008 0.010 0.012 0.363 0.653 5_25Own LOpt 2 -0.065 -0.171 -0.083 -0.515 -0.583 5_25Own MOpt 11 -0.061 -0.014 0.014 -0.548 -0.450 5_25Own NOpt 60 -0.061 0.015 0.037 0.192 0.521 Great25Own HOpt 2 -0.886 -0.004 0.179 1.509 1.586 Great25Own LOpt 1 -0.112 0.066 0.059 1.148 0.863 Great25Own NOpt 21 -0.074 0.029 -0.027 0.332 0.420 Less 1Own HOpt 145 -0.062 0.021 0.006 0.101 0.218 Less 1Own LOpt 93 -0.051 0.021 0.026 0.244 0.325 Less 1Own MOpt 166 -0.034 0.008 0.016 0.139 0.299 Less 1Own NOpt 513 -0.063 0.009 0.016 0.065 0.138 ZeroOwn HOpt 4 -0.059 0.048 0.020 -0.001 0.138 ZeroOwn MOpt 4 -0.133 -0.008 0.132 0.867 0.938 ZeroOwn NOpt 81 -0.053 0.013 -0.013 0.393 0.564 P-value 1335 0.000 0.156 0.559 0.003 0.006
63 Table 2-7 Executive Ownership and Option (Continued) Panel B Groups n (-30, -2) (-1, +1) (+2, +30)(+31, +504) (+31, +756) 1_25Own/Hopt 62 -0.069 0.012 0.040 0.198 0.352 1_25Own/MLopt 62 -0.074 0.018 0.012 0.287 0.441 1_25Own/Nopt 182 -0.074 0.013 0.013 -0.004 0.097 Great25 28 -0.115 0.023 0.005 0.412 0.473 Less1/Hopt 150 -0.062 0.022 0.006 0.098 0.216 Less1/MLopt 263 -0.042 0.012 0.021 0.187 0.318 Less1/Nopt 594 -0.061 0.009 0.012 0.110 0.197 P-value 1340 0.444 0.717 0.857 0.197 0.339
64 Table 28 Type of Repur chase Abnormal Returns Pre-Event CAR Dutch-Auction Market Model Scholes-Williams Year n B Equal Wt Value Wt Equal Wt Value Wt 1995 9 0.96 -0.45% 1.24% -1.14% 0.47% 1996 17 0.61 1.15% -1.10% 0.77% -0.94% 1997 21 1.01 -4.38% -1.69% -5.81% -2.73% 1998 24 0.85 -5.07% -5.89% -4.67% -5.76% 1999 18 0.78 0.43% 1.16% 0.68% 1.15% 2000 17 0.58 -2.12% -0.40% -0.98% 0.00% 2001 9 1.05 -7.67% 1.35% -9.39% 0.38% 2002 11 0.56 -1.66% -2.87% -1.35% -2.43% 1995-2002 126 0.84 -2.49% -1.51% -2.66% -1.67% Fixed Price Market Model Scholes-Williams Year n B Equal Wt Value Wt Equal Wt Value Wt 1995 8 0.96 4.39% 6.00% 6.10% 5.83% 1996 11 0.61 0.42% 1.67% 0.80% 1.69% 1997 14 1.01 -2.86% -2.23% -2.74% -2.11% 1998 6 0.85 6.57% -0.57% 4.34% -1.21% 1999 14 0.78 2.02% 3.77% 1.33% 3.18% 2000 13 0.58 11.82% 9.36% 12.88% 10.47% 2001 9 1.05 20.43% ** 21.19% ** 19.83% ** 20.96% ** 2002 16 0.56 0.91% -0.26% 0.70% 0.16% 1995-2002 91 0.84 4.61% 4.32% 4.63% 4.40% Open Market Market Model Scholes-Williams Year n B Equal Wt Value Wt Equal Wt Value Wt 1995 493 0.96 -2.52% *** -2.75% *** -2.60% *** -6.06% *** 1996 691 0.61 -3.27% ** -4.30% *** -3.31% *** -4.33% *** 1997 664 1.01 -4.08% *** -5.24% *** -3.80% *** -5.85% *** 1998 647 0.85 -7.90% *** -12.44% *** -8.07% *** -12.32% *** 1999 415 0.78 -8.76% *** -7.54% *** -9.15% *** -7.92% *** 2000 320 0.58 -11.75% *** -13.88% *** -11.83% *** -13.52% *** 2001 177 1.05 -5.57% *** -6.17% *** -6.62% *** -6.08% *** 2002 399 0.56 -4.48% *** -6.99% *** -4.83% *** -6.76% *** 1995-2002 3806 0.84 -5.65% *** -7.17% *** -5.78% *** -7.29% ***
65 Table 28 Type of Repurchase (continued) At Announcement CARs (-1,+1) Dutch-Auction Market Model Scholes-Williams Year n B Equal Wt Value Wt Equal Wt Value Wt 1995 9 0.96 4.41% *** 4.55% *** 4.37% 4.38% 1996 17 0.61 9.32% *** 9.12% *** 9.32% *** 9.14% *** 1997 21 1.01 9.66% ** 10.06% *** 9.63% *** 10.14% *** 1998 24 0.85 8.24% *** 8.17% *** 8.36% *** 8.25% *** 1999 18 0.78 7.93% *** 7.91% *** 7.91% *** 7.78% ** 2000 17 0.58 12.96% *** 12.79% *** 12.91% *** 12.81% *** 2001 9 1.05 9.50% ** 9.80% ** 9.53% *** 9.85% *** 2002 11 0.56 13.28% *** 13.23% *** 13.12% *** 13.00% *** 1995-2002 126 0.84 9.47% *** 9.50% *** 9.46% *** 9.49% *** Fixed Price Market Model Scholes-Williams Year n B Equal Wt Value Wt Equal Wt Value Wt 1995 8 0.96 19.54% *** 19.67% *** 19.52% *** 19.55% *** 1996 11 0.61 2.91% 2.93% 2.86% 2.85% 1997 14 1.01 4.16% ** 4.13% ** 4.11% ** 4.09% ** 1998 6 0.85 5.18% *** 4.66% ** 5.22% *** 4.75% ** 1999 14 0.78 12.70% *** 13.08% *** 12.65% *** 13.13% *** 2000 13 0.58 11.08% *** 10.67% *** 11.21% *** 11.18% *** 2001 9 1.05 10.97% *** 10.35% *** 11.39% *** .1079* *** 2002 16 0.56 7.93% *** 7.59% *** 7.77% *** 7.53% *** 1995-2002 91 0.84 9.07% *** 8.92% *** 9.08% *** 9.01% *** Open Market Market Model Scholes-Williams Year n B Equal Wt Value Wt Equal Wt Value Wt 1995 493 0.96 2.48% *** 2.47% *** 2.48% *** 2.45% *** 1996 691 0.61 2.41% *** 2.35% *** 2.44% *** 2.33% *** 1997 664 1.01 2.34% *** 2.24% *** 2.36% *** 2.03% *** 1998 647 0.85 3.45% *** 3.23% *** 3.54% *** 3.24% *** 1999 415 0.78 3.57% *** 3.72% *** 3.59% *** 3.75% *** 2000 320 0.58 4.20% *** 3.91% *** 4.24% *** 4.01% *** 2001 177 1.05 4.86% *** 3.74% *** 4.54% *** 3.95% *** 2002 399 0.56 3.29% *** 3.05% *** 3.31% *** 3.07% *** 1995-2002 3806 0.84 3.07% *** 2.91% *** 3.08% *** 2.90% ***
66 Table 28 Type of Repurchas e (continued) Short-term (+2,+30) CARs Dutch-Auction Market Model Scholes-Williams Year n B Equal Wt Value Wt Equal Wt Value Wt 1995 9 0.96 0.00% 0.94% -0.17% 0.27% 1996 17 0.61 1.07% 0.13% 1.14% -0.11% 1997 21 1.01 -5.83% -3.10% -7.42% -4.71% 1998 24 0.85 3.01% 2.37% 3.95% 2.75% 1999 18 0.78 -4.71% -0.45% -3.96% -0.46% 2000 17 0.58 13.47% 10.55% 12.36% 10.04% 2001 9 1.05 4.36% 4.73% 4.44% 5.05% 2002 11 0.56 2.91% 2.47% 2.91% 2.82% 1995-2002 126 0.84 1.53% 1.93% 1.33% 1.64% Fixed Price Market Model Scholes-Williams Year n B Equal Wt Value Wt Equal Wt Value Wt 1995 8 0.96 -7.04% -5.82% -6.64% -6.40% 1996 11 0.61 2.35% 0.90% 0.56% -0.55% 1997 14 1.01 4.55% 6.58% 4.58% 6.44% 1998 6 0.85 -18.28% -21.51% ** -20.68% ** -23.41% *** 1999 14 0.78 -3.74% -1.98% -5.04% -2.18% 2000 13 0.58 -4.26% -1.67% -3.66% -2.24% 2001 9 1.05 -0.15% 0.34% -1.25% -0.58% 2002 16 0.56 -0.39% -0.87% -0.14% -0.68% 1995-2002 91 0.84 -2.24% -1.47% -2.62% -2.01% Open Market Market Model Scholes-Williams Year n B Equal Wt Value Wt Equal Wt Value Wt 1995 493 0.96 0.17% 0.49% 0.14% 0.11% 1996 691 0.61 0.89% *** 2.35% *** 0.85% *** 0.55% 1997 664 1.01 0.68% 0.61% 0.83% -0.53% 1998 645 0.85 3.59% *** 1.73% 3.78% *** 1.73% 1999 415 0.78 2.18% 4.89% *** 2.10% 4.34% *** 2000 320 0.58 2.93% *** 1.73% ** 3.05% *** 2.11% ** 2001 177 1.05 5.24% 6.47% ** 5.81% 6.40% ** 2002 399 0.56 1.93% 1.69% 1.96% 1.62% 1995-2002 3804 0.84 1.84% *** 1.73% *** 1.92% *** 1.43% ***
67 Table 28 Type of Repur chase (continued) Two-year post CARs Dutch-Auction Market Model Scholes-Williams Year n B Equal Wt Value Wt Equal Wt Value Wt 1995 9 0.96 15.16% 9.84% 10.57% 2.50% 1996 17 0.61 29.08% 2.99% 24.87% 1.33% 1997 21 1.01 -25.93% ** -17.44% ** -44.54% ** -33.91% *** 1998 24 0.85 -3.08% 11.60% -2.47% 10.84% 1999 18 0.78 21.38% 52.79% *** 22.70% 56.29% *** 2000 17 0.58 98.74% *** 86.99% *** 94.99% *** 85.17% *** 2001 9 1.05 -60.34% ** -25.52% -66.22% ** -32.68% ** 2002 11 0.56 -15.43% -7.76% -9.11% -5.06% 1995-2002 126 0.84 10.73% 16.90% 6.62% 13.19% Fixed Price Market Model Scholes-Williams Year n B Equal Wt Value Wt Equal Wt Value Wt 1995 8 0.96 -0.27% -3.44% -4.96% -10.29% 1996 11 0.61 47.80% *** 38.55% *** 45.09% *** 38.73% *** 1997 14 1.01 30.24% 24.42% 28.38% 24.11% 1998 6 0.85 22.99% 16.63% 25.85% 18.58% 1999 14 0.78 9.60% 28.14% 11.82% 33.77% ** 2000 13 0.58 61.40% *** 52.45% *** 62.76% *** 59.80% *** 2001 9 1.05 16.55% 19.33% 19.69% 21.34% 2002 16 0.56 -3.15% 8.45% -7.07% 2.39% 1995-2002 91 0.84 23.33% *** 24.39% *** 22.63% *** 24.89% *** Open Market Market Model Scholes-Williams Year n B Equal Wt Value Wt Equal Wt Value Wt 1995 493 0.96 20.32% *** 16.15% *** 19.59% *** 12.45% *** 1996 691 0.61 25.58% *** 1.99% *** 24.87% *** 1.52% *** 1997 664 1.01 -5.58% *** -14.52% *** -2.21% *** -30.87% *** 1998 643 0.85 0.04% *** 24.41% *** 1.01% *** 22.15% *** 1999 415 0.78 41.18% *** 66.57% *** 37.95% *** 64.43% *** 2000 320 0.58 44.47% *** 26.75% *** 43.60% *** 33.09% *** 2001 177 1.05 -22.96% *** -8.44% *** -19.77% *** -9.82% *** 2002 399 0.56 -14.28% *** 0.56% *** -12.45% *** -2.07% *** 1995-2002 3802 0.84 11.99% *** 13.23% *** 12.43% *** 9.37% ***
68 Table 28 Type of Repur chase (continued) Three-year CARs Dutch-Auction Market Model Scholes-Williams Year n B Equal Wt Value Wt Equal Wt Value Wt 1995 9 0.96 17.32% 6.43% 11.92% -4.68% 1996 17 0.61 12.60% -21.98% 6.91% -24.31% 1997 21 1.01 -37.61% *** -16.66% *** -62.20% *** -38.12% *** 1998 24 0.85 26.53% 49.05% *** 28.88% ** 50.48% *** 1999 18 0.78 44.61% ** 93.86% *** 45.23% ** 98.89% *** 2000 17 0.58 138.77% *** 118.89% *** 132.27% *** 116.48% ** 2001 9 1.05 -79.17% *** -0.30% -86.24% *** -40.78% ** 2002 11 0.56 -15.43% -7.76% -9.11% -5.06% 1995-2002 126 0.84 19.62% ** 30.21% *** 14.02% ** 25.55% *** Fixed Price Market Model Scholes-Williams Year n B Equal Wt Value Wt Equal Wt Value Wt 1995 8 0.96 17.51% 6.95% 4.94% -7.22% 1996 11 0.61 45.69% *** 32.70% ** 38.89% *** 30.09% ** 1997 14 1.01 17.03% 12.01% 13.35% 12.12% 1998 6 0.85 22.90% 12.67% 23.69% 11.26% 1999 14 0.78 24.95% ** 46.25% *** 29.64% ** 57.40% *** 2000 13 0.58 69.43% *** 61.28% *** 68.63% *** 67.59% *** 2001 9 1.05 28.34% 34.83% 34.57% 39.13% 2002 16 0.56 -3.15% 8.45% -7.07% 2.39% 1995-2002 91 0.84 27.34% *** 28.02% *** 25.50% *** 28.34% *** Open Market Market Model Scholes-Williams Year n B Equal Wt Value Wt Equal Wt Value Wt 1995 493 0.96 38.74% *** 21.92% *** 38.21% *** 16.59% *** 1996 691 0.61 25.11% *** -9.15% *** 23.92% *** -9.76% *** 1997 664 1.01 -13.98% *** 14.58% *** -8.48% *** -38.89% *** 1998 643 0.85 19.00% *** 52.11% *** 19.47% *** 53.07% *** 1999 415 0.78 65.06% *** 103.45% *** 58.69% *** 99.39% *** 2000 320 0.58 46.14% *** 26.00% *** 45.69% *** 33.63% *** 2001 177 1.05 -36.28% *** -12.74% *** -30.28% *** -0.15% *** 2002 399 0.56 -14.28% *** 0.56% *** -12.45% *** -2.07% *** 1995-2002 3802 0.84 18.16% *** 20.39% *** 18.64% *** 15.28% ***
69 Table 2-9 Market Reaction to Repurchase Announcements Models 1 through 6 include 913 firms that announce open-market repurchases from 1995 to 2002. Return it = b 0 + b 1 X it + b 2 Year Indicators + e it (t-statistics in parentheses; ***, **, and denote significant levels at 1%, 5% and 10% levels). A. Return for (-30, -2) Panel A is 1 of 2 Variable (1) (3) (5) (6) Intercept -0.07482 -0.04384 -0.05735 *** -0.04381 (-.81) (-1.75) (6.67) (-1.69) PrcBk-1 -0.00007 0.00014 0.00014 (-0.71) (0.42) (0.43) EpsDil-1 0.03759 (0.41) NI/assets -0.04718 (-0.63) Share% -0.07526 ** (-2.14) StkVal -0.00128 (-0.08) IncGrow -0.00032 -0.00057 ** -0.00057 ** (-1.66) (-2.25) (-2.24) EpsGrow 0.00000 0.00000 (0.25) (0.32) MktVal 0.00034 (0.97) BlkVal 0.00039 (0.66) Soptexer 0.00051 0.00050 (0.68) (0.66) DumACQ 0.00598 0.00667 0.02134 0.01139 (0.09) (0.11) (0.30) (0.18) DumDIL 0.01781 0.01478 0.01877 0.01712 (0.63) (0.48) (0.67) (0.56) DumEBP 0.00539 -0.00762 -0.00630 -0.00664 (0.21) (-0.26) (-0.22) (-0.22) DumESV -0.02367 -0.00148 -0.02299 -0.00185 (-1.22) (-0.07) (-1.26) (-0.09) DumSTP 0.04380 ** 0.04167 ** 0.03036 0.04277 ** (2.13) (2.04) (1.53) (2.07) DumUVL -0.08009 *** -0.11400 *** -0.08829 *** -0.11392 *** (-3.11) (-4.21) (-3.43) (-4.19) Freq -0.01310 (-0.69) Somefreq 0.00623 (0.37) Year Yes Yes No Yes Adj. R2 0.0365 0.0517 0.0121 0.0494
70 Table 2-9 (Continued) Models 7 and 8 include 963 firms that announce open-market, Dutch-auction or fixed-price repurchases from 1995 to 2002. A. Return for (-30, -2) Panel A is 2 of 2. Variable (2) (4) (7) (8) Intercept -0.00872 -0.06282 *** (-0.07) (-2.76) PrcBk_0 -0.00059 -0.00017 (-0.71) (-0.52) EpsDil_0 0.04454 (0.36) Share% -0.07525 ** (-2.04) IncGrow -0.00049 ** (-2.27) EpsGrow 0.00000 (0.38) MktVal 0.00028 (0.81) Soptexer -0.00004 0.00052 (-0.03) (0.67) DumACQ 0.00855 (0.12) DumDIL 0.01317 (0.43) DumEBP -0.00437 (-0.15) DumESV -0.03493 (-1.60) DumSTP 0.04245 ** (1.97) DumUVL -0.08121 *** (-2.87) Year Yes Yes Adj. R2 0.0422 0.0144
71 Table 2-9 (continued) B. Return for (-1 +1) Panel B is 1 of 2 Variable (1) (3) (5) (6) Intercept -0.03232 0.00069 0.01572 *** -0.00184 (-0.83) (0.06) (4.50) (-0.16) PrcBk-1 -0.00001 -0.00019 -0.00019 (-0.18) (-1.33) (-1.34) EpsDil-1 0.03675 (0.95) NI/assets -0.00686 (-0.22) Share% -0.01687 (-1.14) StkVal 0.00135 (1.90) IncGrow -0.00009 -0.00007 -0.00006 (-1.13) (-0.61) (-0.52) EpsGrow 0.00001 ** 0.00001 ** MktVal -0.00008 (-0.56) BlkVal 0.00004 (0.15) Soptexer 0.00048 0.00049 (1.45) (1.07) DumACQ -0.01342 0.01246 0.00938 0.00908 (0.47) (0.45) (0.32) (0.33) DumDIL -0.01188 -0.02039 -0.00829 -0.02241 (-0.99) (-1.52) (-0.73) (-1.65) DumEBP -0.00812 -0.01216 -0.00449 -0.01263 (-0.66) (-0.93) (-0.39) (-0.96) DumESV 0.01024 0.00831 0.00718 0.00772 (1.25) (0.89) (0.97) (0.82) DumSTP -0.00019 -0.01008 -0.00389 -0.01136 (-0.02) (-1.12) (-0.48) (-1.25) DumUVL 0.00239 -0.00985 0.00015 -0.01064 (0.22) (-0.82) (0.01) (-0.89) Freq 0.00641 (0.76) Somefreq 0.00244 (0.33) Year Yes Yes No Yes Adj. R2 -0.0062 0.0062 -0.0041 0.0052
72 Table 2-9 (continued) B. Return for (-1 +1) Panel B is 2 of 2. Variable (2) (4) (7) (8) Intercept -0.00352 0.00384 0.00596 0.01180 (-.007) (0.39) (0.53) (1.29) PrcBk_0 0.00004 -0.00019 (0.13) (-1.34) EpsDil_0 0.00529 (0.11) Share% -0.01509 (-1.01) IncGrow -0.00012 -0.00012 (-1.33) (-1.08) EpsGrow 0.00001 ** 0.00000 MktVal -0.00004 (-0.30) Soptexer 0.00001 0.00049 0.00039 (1.69) (1.46) (1.16) DumACQ 0.01274 0.00983 0.01381 (0.44) (0.35) (0.47) DumDIL -0.01956 -0.01793 -0.00754 (-1.60) (-1.30) (-0.65) DumEBP -0.00823 -0.01211 -0.00395 (-0.64) (-0.91) (-0.34) DumESV 0.01963 ** 0.00512 0.00745 (2.22) (0.56) (0.97) DumSTP -0.00182 -0.00960 -0.00324 (-0.21) (-1.05) (-0.39) DumUVL 0.00026 -0.00405 0.00470 (0.02) (-0.34) (0.44) Freq 0.01 (0.77) Somefreq 0.00 (0.06) DA 0.07 *** 0.06122 *** (5.11) (4.69) FP -0.02 -0.01425 (-0.68) (-0.48) Year Yes Yes Yes Yes Adj. R2 0.0002 0.0052 0.0390 0.0139
73 Table 2-9 (continued) C. Return for (+2 +30) Panel C is 1 of 2 Variable (1) (3) (5) (6) Intercept -0.05920 0.00274 0.01746 ** 0.01165 (-0.77) (0.14) (2.44) (0.56) PrcBk-1 -0.00004 -0.00089 *** -0.00088 (-0.49) (-3.45) (-3.44) *** EpsDil-1 0.05954 (0.78) NI/assets 0.06832 (1.11) Share% 0.00731 (0.25) StkVal 0.00247 (1.77) IncGrow -0.00019 0.00041 ** 0.00037 (-1.20) (2.01) (1.82) EpsGrow 0.00000 0.00001 (0.57) (0.97) MktVal 0.00058 (0.20) BlkVal -0.00007 (-0.15) Soptexer -0.00053 -0.00063 (-0.96) (-1.04) DumACQ 0.00179 0.01022 0.00542 0.02843 (0.03) (0.20) (0.09) (0.57) DumDIL -0.01649 -0.00534 -0.02721 0.00491 (-0.70) (-0.22) (-1.16) (0.20) DumEBP -0.02076 -0.01848 -0.01060 -0.01550 (-0.87) (-0.77) (-0.44) (-0.65) DumESV -0.00223 0.00829 0.00542 0.00986 (-0.14) (0.49) (0.36) (0.58) DumSTP 0.01879 0.00397 0.02073 0.00996 (1.11) (0.24) (1.26) (0.61) DumUVL 0.01361 0.02833 0.01873 0.03119 (0.64) (1.30) (0.87) (1.44) Freq -0.04007 *** (-2.64) Somefreq -0.00017 (-0.01) Year Yes Yes No Yes Adj. R2 -0.0009 0.0452 -0.0021 0.0572
74 Table 2-9 (continued) C. Return for (+2 +30) Panel C is 2 of 2. Variable (2) (4) (7) (8) Intercept -0.01616 0.00549 -0.00635 0.00638 (-0.16) (0.31) (-0.16) (0.23) PrcBk_0 0.00058 -0.00078 *** (0.85) (-3.07) EpsDil_0 -0.01730 (-0.17) Share% 0.00372 (0.12) IncGrow -0.00007 0.00026 (-0.39) (0.68) EpsGrow 0.00000 0.00000 (0.54) (0.28) MktVal 0.00018 (0.67) Soptexer 0.00103 *** -0.00055 0.00039 (0.91) (-0.90) (-0.33) DumACQ 0.00401 -0.00125 -0.00343 (0.07) (-0.01) (-0.04) DumDIL -0.01632 -0.00955 -0.02943 (-0.65) (-0.20) (-0.82) DumEBP -0.01894 -0.04501 -0.02623 (-0.73) (-0.98) (-0.72) DumESV 0.00220 -0.03871 -0.01688 (0.12) (-1.22) (-0.72) DumSTP 0.00817 -0.01210 0.01125 (0.46) (-0.38) (0.45) DumUVL 0.00529 0.00080 0.00735 (0.23) (0.02) (0.23) Freq 0.00173 (0.06) Somefreq 0.02599 (1.03) DA 0.20396 *** 0.18427 *** (4.53) (4.59) FP 0.17624 0.17596 (1.79) (1.94) Year Yes Yes Yes Yes Adj. R2 -0.0016 0.0419 0.0376 0.0220
75 Table 2-9 (continued) D. Return for (+30 Two years) Panel D is 1 of 2 Variable (1) (3) (5) (6) Intercept -0.13381 -0.01625 0.12715 *** -0.01137 (-0.74) (-0.12) (2.97) (-0.08) PrcBk-1 0.00051 0.00190 0.00189 (1.10) (1.15) (1.15) EpsDil-1 0.41952 (0.93) NI/assets -0.62612 (-1.71) Share% 0.35803 ** (2.08) StkVal 0.01030 (1.25) IncGrow -0.00409 *** -0.00668 *** -0.00669 *** (-4.34) (-5.25) (-5.23) EpsGrow 0.00002 0.00002 (0.42) (0.42) MktVal -0.00123 (-0.72) BlkVal -0.00115 (-0.40) Soptexer -0.00353 -0.00353 (-0.93) (-0.93) DumACQ -0.20677 -0.26211 -0.24351 -0.26028 (-0.62) (-0.84) (-0.69) (-0.82) DumDIL -0.14856 -0.25748 -0.28589 ** -0.25591 (-1.07) (-1.68) (-2.05) (-1.66) DumEBP -0.04357 0.04580 0.09425 0.04574 (-0.31) (0.31) (0.66) (0.31) DumESV -0.09432 -0.01325 -0.14440 -0.01168 (-0.99) (-0.12) (-1.58) (-0.11) DumSTP 0.24495 ** 0.25132 *** 0.22910 ** 0.25270 ** (2.57) (2.45) (2.33) (2.44) DumUVL -0.08316 -0.06036 -0.13960 -0.05890 (-0.66) (-0.44) (-1.09) (-0.43) Freq 0.00067 (0.01) Somefreq -0.01090 (-0.13) Year Yes Yes No Yes Adj. R2 0.0816 0.1253 0.0097 0.1224
76 Table 2-9 (continued) D. Return for (+30 Two years) Panel D is 2 of 2. Variable (2) (4) (7) (8) Intercept -0.50772 -0.09659 -0.02546 -0.01469 (-0.82) (-0.81) (-0.19) (-0.13) PrcBk_0 0.00202 0.00088 (0.50) (0.53) EpsDil_0 0.60061 (1.01) Share% 0.34833 (1.95) IncGrow -0.00456 *** -0.00613 *** (-4.33) (-4.82) EpsGrow 0.00002 0.00002 (0.46) (0.63) MktVal -0.00057 (-0.33) Soptexer 0.00232 -0.00316 -0.00390 (0.35) (-0.81) (-1.01) DumACQ -0.23201 -0.26787 -0.22517 (-0.67) (-0.83) (-0.65) DumDIL -0.29160 ** -0.28226 -0.26150 (-1.99) (-1.81) (-1.91) DumEBP -0.04190 0.02893 0.04388 (-0.27) (0.19) (0.32) DumESV 0.11796 -0.04740 -0.09898 (-1.11) (-0.46) (-1.09) DumSTP 0.20114 0.22129 ** 0.16910 (1.93) (2.13) (1.75) DumUVL -0.06279 -0.06986 -0.19176 (-0.46) (-0.51) (-1.53) Freq 0.00320 (0.03) Somefreq 0.03316 (0.40) DA 0.12428 0.16774 (0.84) (1.08) FP -0.04013 0.10648 (-0.12) (0.31) Year Yes Yes Yes Yes Adj. R2 0.0754 0.0842 0.1212 0.0534
77 Table 2-9 (continued) E. Return for (+30 Three years) Panel E is 1 of 2 Variable (1) (3) (5) (6) Intercept -0.84043 0.05133 0.20817 *** 0.08689 (-1.47) (0.30) (3.78) (0.50) PrcBk-1 0.00004 0.00455 ** 0.00452 ** (0.06) (2.17) (2.15) EpsDil-1 0.99994 (1.76) NI/assets -0.81467 (-1.77) Share% 0.39822 (1.84) StkVal 0.00361 (0.35) IncGrow -0.00490 *** -0.00455 ** -0.00460 ** (-4.13) (-2.39) (-2.42) EpsGrow -0.00539 *** -0.00550 *** (-3.63) (-3.69) MktVal -0.00244 (-1.13) BlkVal 0.00059 (0.16) Soptexer -0.00632 -0.00638 (-1.30) (-1.32) DumACQ -0.44883 *** -0.54254 -0.54550 -0.53149 (-1.07) (-1.36) (-1.20) (-1.32) DumDIL -0.24654 ** -0.38236 -0.38534 ** -0.32085 (-1.40) (-1.71) (-2.16) (-1.63) DumEBP 0.07503 0.10720 0.27925 0.10691 (0.42) (0.56) (1.53) (0.56) DumESV -0.12850 -0.07849 -0.19422 -0.06694 (-1.07) (-0.58) (-1.66) (-0.49) DumSTP 0.34207 *** 0.27753 ** 0.33184 *** 0.28678 ** (2.70) (2.13) (2.64) (2.18) DumUVL -0.06817 -0.06211 -0.11837 -0.05082 (-0.43) (-0.36) (-0.72) (-0.29) Freq 0.00949 (0.08) Somefreq -0.08198 (-0.75) Year Yes Yes No Yes Adj. R2 0.07560 0.12420 0.01560 0.12240
78 Table 2-9 (continued) E. Return for (+30 Three years) Panel E is 2 of 2. Variable (2) (4) (7) (8) Intercept -0.89011 0.00417 0.09190 0.00902 (-1.13) (0.03) (0.53) (0.06) PrcBk_0 -0.00027 *** 0.00430 ** (-0.05) (2.06) EpsDil_0 1.03764 (1.36) Share% 0.41224 (1.80) IncGrow -0.00568 *** -0.00432 ** (-4.24) (-2.31) EpsGrow -0.00720 *** -0.00438 *** (-5.69) (-3.21) MktVal -0.00152 (-0.71) Soptexer -0.00558 -0.00636 -0.07150 (-0.66) (-1.31) (-1.47) DumACQ -0.49438 -0.50009 -0.48365 (-1.12) (-1.23) (-1.10) DumDIL -0.40336 ** -0.37268 -0.37950 ** (-2.16) (-1.88) (-2.183) DumEBP 0.08885 0.09630 0.19319 (0.46) (0.51) (1.09) DumESV -0.15604 -0.08623 -0.13244 (-1.15) (-0.64) (-1.14) DumSTP 0.28975 ** 0.27379 ** 0.25864 ** (2.18) (2.09) (2.11) DumUVL -0.04656 -0.05351 -0.20479 (-0.27) (-0.31) (-1.29) Freq -0.04723 (-0.40) Somefreq -0.02409 (-0.23) DA -0.03018 -0.00015 (-0.14) (0.00) FP -0.14752 0.01255 (-0.29) (0.02) Year Yes Yes Yes Yes Adj. R2 0.06820 0.11180 0.10570 0.04820
79 Chapter 3 Why do firms repurchase stock to acquire another firm? In chapter 2, I suggest that using a re purchase to fund an acquisition may be an opportunistic motivation. This categorizati on was based on the hubris hypothesis of why we continue to see poor stock re turns associated with mergers.45 This chapter investigates the enigmatic decision by a firm to take on the extra transactional step to repurchase its shares with cash and then use those shares to finance an acquisition, rather than use the cash to directly finance the acquisition. It woul d seem to be far easier, if a firm has the cash available, to acquire the target firm with the cash. This is even more of an enigma when it is well known that cash offerings perform better than stock offerings.46 I find that firms that repurchase shar es to finance an acquisition are well compensated for their efforts, The most compelling argument as to why firms would take on the extra financing step is to achieve the best of both the stockfinancing acquisitions and cash-financing acquisitions. These firms experience risk sharing with the target firms, counteract the negative effects of diluti on by repurchasing shares first, and enjoy a tax advantage for their efforts. 45 Roll, 1986, The Hubris Hypothesis of Corporate Takeovers 46 Martin, 1986, The Method of Payment in Corporate Acquisitions, Investment Opportunities, and Management Ownership.
80 The organization of this chapter proceed s as follows. The first part discusses merger and acquisition lite rature. The second section develops the hypotheses and methodology. The third section reports the em pirical findings a nd the last section summarizes and concludes the chapter. Literature Takeovers can occur through mergers, te nder offers, or proxy contests. This research focuses on mergers, which are ge nerally stock-financed, and tender offers, which are generally cash-financed. Mergers are negotiated directly with the target managers and require approval from the target firms board of directors. Tender offers are offers to buy shares made directly to target shareholders, bypassing the target managers. This research will also investigate the me thod of payment choice. Specifically, the research asks why firms would choose to repurc hase their shares with cash and then use those shares to finance an acquisition, rather than use cash to finance the acquisition directly. Several hypotheses have been put forth to help explain the markets behavior toward mergers and acquisitions. Myers and Majluf (1984), in thei r seminal paper that develops pecking order, address the benef its of financial slack. Convention assumes managers should accept all positiv e net present value (NPV) proj ects. However, if a firm can only issue risky debt, the firm may rationally pass up on positive NPV projects. Having financial slack enables the firm to take all positive NPV projects. Thus, firms with financial slack will issue stock only when their stock is overvalued. As a result, the
81 equity offering may send a signal that manage rs believe that thei r stock is overvalued. Thus, consistent with the Cash Availability Hypothesis, firms with financial slack will prefer to finance an acquisition with cash if their stock is undervalued. In a similar vein to the Cash Ava ilability Hypothesis, the Investment Opportunities Hypothesis argues that managers with growth opport unities will prefer to raise capital with equity because it allows more flexibility in the use of funds than debt financing.47 Martin (1996) tests this hypothesis with three proxies for investment opportunity (Tobins q, five-years sales growth and the recent run-up of the firms stock price) and finds that both acquiring firms with high growth opportunity and the firms with recent stock price run-ups are more likely to use stock. The Control Hypothesis argues that firm s with large managerial ownership positions should prefer to use cash to finance an acquisition because the alternative would dilute the managers control position.48 Martin (1996) suggests that managerial ownership may be nonlinear on the choice of stock financing. At very low and high ownership levels managers may not be very concerned about the im pact of dilution of control. However, the middle range may be very concerned. Martin uses a spline variable approach to define the ownership levels. Low ownership groups have less than five percent manager and director combined owners hip; the middle range is defined as greater than five percent and less than 25 percent a nd the high owner firms have greater than 25 percent manager and director control. Mart ins sample of 721 fi rms found 425 in the low 47 Myers (1977) ties existence of growth opportunities to debt. Myers and Majluf (1984) and Jung, Kim, and Stulz (1995) put forth this hypothesis.
82 group, 299 in the middle range and 125 in the high ownership group.49 Martin found that the low ownership group was not significant and he suggests that managers with low ownership were not concerned with dilution effects. Martin found that the middle range considered ownership important (significantly negative relationships in three of his four logistic regressions which sugge st that the firm is not lik ely to use stock financing). Finally, Martin found the high ownership group to be significantly negative in one of four regressions, substantia ting his nonlinear choice of stock financing Martin (1996) uses the Risk Sharing Hypothesis to suggest that as target firms size increase (as measured by market value) and the addition of target firms to bidder firms becomes more significant, the bidder will prefer to use stock in order to share the risk with the target. On the other hand, if th e acquiring firms size is significantly larger than the target firm, the acquirer will be less lik ely to feel the need to share risk since the target will not have as much an impact to the combined firm. In this case, the bidder will be less likely to use stock. Hansen (1987) m odels the payment choice under the condition of asymmetric information. If the target knows its value better than th e bidder, the bidder will prefer stock in order to force the target to share in the post-acquisition reevaluation effects. Martins initial investigation fi nds no support. Martin refines his test by establishing four distant groups of firms; both the bidder and target firms have high Tobins q ratios (q>1), both firms have low q ratios (q<1), and groups of one firm high q and one firm low q. Martin finds that 68 percen t of firms choose stock financing and only 48 Stulz (1988) and Jung, Kim, and Stulz (1995) suggest that managers will not want issue stock if in doing so will dilute their control position. Amihud, Lev, and Travlos (1990) find evidence to support the control hypothesis.
83 16 percent choose cash if both the bidder and ta rget have high investment opportunities. Furthermore, if both the parties have low i nvestment opportunities, then only 26 percent use stock financing and 42 percent use cash financing50. The Outside Monitoring Hypothesis is suggested by Jensen (1991) and Black (1992). They argue that active investors and in stitutional shareholders undertake costly monitoring and thus can take actions to align managers in terests with those of the shareholders. Martin (1996) finds support in that the presence of blockholders and institutional holders results in a hi gher likelihood of stock financing. Another argument supporti ng stock-financed acquisi tions was put forth by Aboody, Kasznik and Williams (2000). They su ggest that managers of acquiring firms should prefer pooling versus purchasi ng accounting because purchase accounting requires the firm to book as an asset the difference between the purchase price of the acquisition and the book value of th e target firm. This asset is expensed into the future. This additional expense dampens net income and earning per share for years to come. As an aside, the authors did not address the positive tax consequences of these future expenses, but instead focused attention on th e negative impacts of the publicly reported earnings.51 Aboody, Kasznik, and Williams suggest th at because managerial incentive contracts are often tied to net income and earnings per share, managers should prefer pooling accounting. However, pooling accounting cannot be accomplished in conjunction 49 This cutoff choice is consistent with Morck, Shleifer and Vishny (1988). Furthermore Martin (1995) performed robustness checks on percentage chan ges with the groups and found little difference. 50 Martins findings are in a total sample in which 40 percent of firms use stock financing and 35 percent cash.
84 with an acquiring firm repurchasing its shares APBO No. 16 prohibits changes in equity interests of voting stoc k with pooling accounting.52 Therefore, firms cannot repurchase shares and then use those shares to pool asse ts with a target firm. The authors conclude that although managers should prefer pooling assets in stoc k for stock acquisitions, the transaction cannot be supported with the acqui ring firms repurchase of its own shares. Aboody, Kasznik and Williams (2000) also argue that managers should prefer pooling. However, this might be shortsighted by the managers. Goodwill is created when a firm purchases another firm (or assets of a nother firm) for greater than its market value. The difference between the fair market value and the amount paid is considered goodwill. Historically, financial statement reporting for a purchase of another firm books goodwill as an asset and the goodwill is amortized fo r up to 40 years. On the other hand, when firms merge through pooling accounting, no goodwill is created. The combined firms simply add previously recorded book values together. Thus, for financial statement reporting the purchase accoun ting results in higher booke d assets and thus higher amortization expenses. This decreases the combined firms reported earnings and lowers earning per share; thus most mangers woul d seek to meet the criteria of pooling accounting. 51 Title 26 of the Internal Revenue Code; subtitle A; Ch apter 1; Subchapter B; Part VI; section 197 effective August 1993 states that a taxpayer is entitled to an amortization deduction for goodwill over a period of up to 15 years. 52 Accounting Principles Board Opinion No. 16 (August 1070) establishes 12 specific criteria that most be met in order to qualify for pooling accounting. These 12 criteria include the use of exchange of common stock (90% substantially all rule); no equity change s in contemplation of combination (two-year rule) and shares can be reacquired only for purposes other than business combination. If an y criteria are not met, the purchase method must be used.
85 Financial statement reporting and tax code reporting can be different. For most of the US tax code history, amortization of goodw ill was not a tax-deductible expense. Thus mangers were justifiably reluctant to inher it the appearance of lower earnings per share that resulted with purchas e accounting. However, effec tive August 1993, section 127 of the US tax code regarding amortization of goodwill and other intangibles changed to allow taxpayers the amortization deduction. Although there have been recent changes regarding purchase versus pooling accounting, these changes do not affect the section 127 tax code53. Since 1993 acquiring firms mana gers who chose purchase accounting received a tax benefit (reduc tion in taxes payable due to increase in amortizable deductions). Although the purchas e method does have the appear ance of lower earnings per share; in reality the purc hase method through its real ta x benefit affords higher cash flows to the purchasing firm. My sample of firms falls under this tax code. As an example, lets assume that a firm used purchase accounting with the acquisition of another firm. A very simp le income statement might look like the following: 53 SFAS No. 141, Business Combinations, issued on 7-20-01 requires that the purchase method of accounting be used for all business combinations in itiated after 6-30-01. This new purchase accounting method does not allow for goodwill amortization for financial repor ting. Instead companies will recognize goodwill as an asset on financial statements and present it as a separate item on the balance sheet. Companies will then conduct an annual impairment test and goodwill will remain on the balance sheet as an asset subject to impairment. Some effects of this change may be that companies will no longer worry about structuring a deal in order to comply with po oling, goodwill will be more reflective of value and not a system of arbitrary amortization, and impairment charges could be costly and bumpy The new standards for goodwill accounting should contribute to more mean ingful financial statements, improved transparency and greater consistency among compan ies. Three drawbacks are the lack of an international convergence, inconsistency with the existing tax code where g oodwill remains amortizable, and the removal of the discipline of writing off goodwill which allows for more management discretion.
86 Pooling Purchase Operating Income 50 50 Additional depreciation Due to purchase 0 10 Earnings before taxes 50 40 Taxes at 40 percent 20 16 Net Income 30 24 Thus, net income and earnings per share are greater under pooling accounting. However, depreciation is a non-cash expense and as a ma tter of fact the sources of cash are greater with purchasing. Therefore, a more insightfu l manager might realize the tax savings of the purchase accounting will add value to the co mpanys cash flows. Additionally, this is what should matter to the investors. Thus, purchase accounting can be beneficial due to the tax advantage. Normally, purchase account ing is conducted with cash. However, if a firm believes its stock is underv alued and if the firm also has ample cash flow, managers may be able to avail themselves of the best of both worlds. They can first take advantage of their undervalued stock by announcing a repurch ase and then use that stock to support an acquisition and by APBO No. 16 be forced to use purchase accounting which will save the firm future taxes. This scenario give s a strong argument for w hy firms would take on extra transactions in order to acquire another firm. They proc eed with this method to take advantage of undervalued stock and to re duce the firms future tax burden. Most research on mergers and acquisitions finds that, on average, target firms gain value and the bidder firms lose value.54 For example, Travlos (1987) explores the 54 Bradley, Desai, and Kim, 1988, find average re turns to bidders are non-positive. Lang, Stulz, and Walkling, 1989, find that having a low Tobins Q (proxy for poor quality of bidding firm management) reduces the bidders return (see al so Servaes, 1991). Dennis and McConnell, 1986, investigate firms senior securities and find that the bidder firms convertib le preferred stockholders have significantly positive returns while all other security holders are not significantly different from zero. Jensen and Ruback, 1983, provide summary of the literature.
87 Method of Payment Hypothesis and finds th at bidders who use stock have significant negative abnormal returns while bidders who make cash offers experience normal returns. Specifically, Travlos finds that bi dders using stock have significant negative stock price return reaction of -0.78 percent on the day pr ior to the announcement and 0.67 percent on the day of the announcement. Travlos concludes that, on average, stockholders of acquiring firms experience si gnificant losses when their firm acquires another firm through the exchange of comm on stock. In contrast, bidders using cash offers have a positive two-day (-1,0) signifi cant cumulative abnormal return reaction of 0.24 percent. Travlos suggests that his findi ngs are consistent with the Signaling Hypothesis. He explains that firms signal overvaluation if they finance a takeover with stock. Thus, managers will prefer cash if they believe their stock is undervalued, while a stock-financed offer will be preferred in the opposite case. Accordingly, the market participants respond favorably to cash offers and negatively to stock offers. If a firm announces that it will repurchas e its own shares prior to or in conjunction with an acquisition announcement, then the market participants are not likely to perceive that the firm is overvalued. In fact the market should conclude quite the opposite. Taxes may have an important impact on st ock price reactions of the bidding and target firms. Cash offers generate tax ob ligations for the target firms stockholders. Wansley, Lane and Yang (1987) suggest that bi dding firms will pay a higher premium in a cash offer to compensate a target firm's shareholders for their tax burden of tendering shares. Thus, the target firms gain may simply be compensation for its upcoming tax burden. Again it should be noted that the firm using repurchased shares as the financing
88 vehicle would not have to offer the bidder fi rm additional compensation for the targets tax consequences. However, it should be noted that the bidders shareholders may suffer a capital gains tax loss from tendering shares. A cash offer affects the taxes of the ac quiring firms by raising the depreciation basis of acquired assets to th e assets market values. This is advantageous to increasing the firm's expenses and decr easing the firm's tax liability, but also has the effect of reducing booked net income, which may have perceived adverse effects to the market participants. Wansley, Lane, and Yang (1987) in vestigate the gains to bidder firms and find significant positive stoc k price returns of 6.17 pe rcent for an 81-day period surrounding a cash acquisition announcement and a zero gain for stock acquisitions. The authors suggest that the offer to the target re flects the bidders expecta tions of the targets value. Alternatively, if the bidder believes that the target is overva lued the bidder will choose to finance with stock and if the bi dders believes the target is undervalued the bidder will prefer cash55. Servaes (1991) and Lang, Stulz and Wa lkling (1991) also explore the takeover gains. Lang, Stulz, and Walkling document th at abnormal returns in tender offers are related to Tobin's q (high q is defined as q> 1 and low as q<1) and find that bidders with high q ratios have significant positive a bnormal returns and low q bidders have significant negative returns. Lang, Stulz, and Walkling find the highest value is created when a high q firm takes over a low q firm a nd the value is destroyed when a low q firm takes over a high q firm. Servaes adds to the research of Lang, Stulz, and Walkling with 55 Hansen (1987)
89 the addition of merger offers in his extende d sample of 704 successful takeovers over the period 1972-1987. Servaes finds for the total sa mple that target returns are positive (23.64 percent) and significant, bidder negative (-1.07 percent) and to tal return s positive (3.66 percent). Servaes splits his sample based on the method of payment and finds that a cash offer results in a 26.67 per cent return to the target firm 3.44 percent return to the bidder and an 8.41 percent to the combined firm (calculated as a weighted average). He finds that a stock-financed acquisition, on aver age, results in a 20.47 pe rcent return to the target firm, a -5.86 percent to the bidder, and a -3.03 percent to the combined firm. Finally, a combined stock and cas h offer results in a 21.05 percen t return to the target, a 3.74 percent return to the bidder, and 5.64 to the combined firm. Morck, Shleifer and Vishny (1990) inves tigate why returns to bidding firms are negative and find that returns are most negative when the firm acquires another firm as a diversification, when it acquire s a growing firm, or if it has poor management. Morck, Shleifer, and Vishny suggest that managers make acquisitions to pursue personal objectives other than maximization of sharehol der value. Thus, managers are willing to pay more for targets than they are worth if the acquisition will serve their personal benefits of improving their j ob security or diversifying th eir human capital. Loughran and Vijh (1997) calculate the total wealth gains for mergers and acquisitions and find that all returns to the bidder in a stock merger are ne gative (and cash is positive). Specifically, Loughran and Vijh find that stock acquirers earn 24.2 percent less than their matched firms, on average, using buy and hold return s over a five-year period; whereas cash acquirers earn 18.5 percent more than their matched firms. Loughran and Vijh suggest
90 that cash tender offers obtain considerable gains because of the associated disciplinary actions afforded to the cash acquisition, such as the ability to appoint new managers. Rau and Vermaelen (1998) investigate long-term bi dder performance and find that bidders in mergers underperform for up to three-y ears after the merger is complete. Rather than explaining the persistent negative and positive stock price reactions to acquisitions as a method of payment issu e, Megginson, Morgan and Nail (2004) show that the primary determinant of long-term perf ormance in strategic mergers is related to changes in corporate focus. The authors find that the change in focus is significantly related to performance after cont rolling for the form of payment56, and firm value ratios such as book-to-market ratio.57 The authors find that focus-decreasing mergers as defined by the Herfindahl index which quantifies th e revenue changes of each line of business (SIC codes)58, result in significantly negative long-term performance.59 Furthermore Megginson, Morgan and Nail find that increasing or at least preservi ng the focus of the firms lines of business result in margin al long-term performance improvements. Megginson, Morgan and Nails re gression results reveal that their measure of focus change is the only variable with signifi cant relationships to long-term buy-and-hold abnormal returns. 56 Loughran and Vijh (1997), Ghosh (2001), Linn and Switzer (2001). and Martin (1996) 57 Rau and Vermaelen (1998) and Martin (1996) 58 The degree of focus is found using the Herfindahl index for both revenue and assets. This looks at square of revenue (assets) of each division divided by the total revenue. Thus, the revenue of division A is divided by the revenue of the entire firm and then squared. 59 Megginson etal find an average loss in stockholder wealth, firm value and cash flows of 18 percent, 9 percent and 2 percent respectively for up to three years post-merger.
91 Sicherman and Pettway (1987) find that buyers of divested assets gain wealth if the firms have related assets as defined by th eir two digit SIC codes. The authors studied 127 firms that acquired divested assets from 1893-1985 and found that related asset acquisitions resulted in an average 3.975 per cent gain over firms that acquired unrelated divested assets. Sicherman and Pettway add th at shareholders obtain higher returns when the acquirer purchases related lines of busin ess. They further investigate how insider ownership affects the choice of relatedness. Copeland and Weston (1988) discuss that ma nagers may be motivated to acquire unrelated assets in order to reduce personal risk. A substantial portion of the managers wealth is invested as human capital from employment. Thus, purchasing a divested unrelated line or simply diversifying the firms assets helps reduce the managers employment assets, which reduces personal risk. Sicherman and Pettway use insider ownership as a percentage of total ownership and find that firms acquiring related assets have a greater equity ownership than firms acquiring unrelated assets. Thus, firms that have high manager ownership are more likely to act in the shareholders best interests, whereas low ownership managers may be motivat ed to be more self-serving and prefer to reduce their own risk. Many studies note this effect. Kaplan and Weisbach (1992) find that bidder results are slightly negative and the combined firms returns are pos itive. Thus, targets appear to be worth less than bidders pay, but are worth more than the targets market value prior to the takeover suggesting that the acquisi tion increases the combined shareholder wealth. Kaplan and Weisbach focus on 1971-1982 acquisitions that later
92 divested and classify some as successful. Alth ough the emphasis of th eir paper is to show that not all firms that acquire another firm and then later divest are unsuccessful, they discovered an interesting outcome, showing th at the markets initial reaction to the acquisition did a good job predicting whether th e later divestiture was a success or not. They define an unsuccessful acquisition as one that reports an accounting loss at the later dates divestiture. Kaplan and Weisbach find th at the combined returns at the acquisition announcement are significantly lower for the acq uisitions that will, in hindsight, be classified as unsuccessful as compared to the corresponding returns for the successful divestitures and for the acquisi tions that do not divest. Although it has been substantiated that bidders often lose in an acquisition, merger activity continues. Ro ll (1986) explains this acquisition fever in his hubris hypothesis. In summary, the hubris hypothesis suggests that managers are infected by hubris and so overpay for targets because they overestimate their own ability to run the merged firm.60 Furthermore, managers believe that they are better at estimating valuation than the merger evidence would indicate. A dditionally, Jensen ( 1986) and others have noted that managers may want to increase th e firms size because they value the status associated with a larger firm and furthermor e their compensation may be tied to the size of the firm. Thus, managers do not want to reduce firm size by distributing assets to shareholders (see Roll (1986)). Managers ar e more concerned with growth in market share, in labor employment and in new lines of business than in maximizing shareholder wealth (see Jensen and Meckling (1976)). Thus, managers are more concerned with size
93 and things money cannot buy, such as perks, prestige and future employment, than they are with maximizing the value of the firm.61 This desire to empire-build (see Roll (1986)) may lead to the acceptance of negative net present value projects; that is, managers may overinvest. Since managers are risk averse, they will not choose to increase th eir debt to the point of risking bankruptcy. Thus, rational self-serving managers will not want to increase interest payments and commitment levels in order to empire build ; however, they will be more likely to make poor investment choices when free cash flow is available and bankruptcy is less likely. Poor investments could be carried out with cash or stock financing. If the firm chooses to use stock financing, managers may also c onduct repurchasing activit ies to finance the stock-based acquisition. Thus, a repurchase in order to condu ct an acquisition may signal an empire-building strategy and would not necessarily be good news to the market. Furthermore, stock-financed acquisitions have been shown to decrease the value of the acquiring firm.62 Thus, firms that repurchase to facili tate an acquisition may find that the market reactions to the repurchase announcem ents are similar to non-positive market reactions to acquisition announcements. Repurchasing stock in order to finance an acquisition creates one more step in the empire-building firm strategy. There are costs as sociated with this extra step (time lost and transaction fees) and therefor e it would seem that there must be a benefit. It is likely 60 Shleifer and Vishny, 1988, suggest that managers willingly overpay for an acquisition to improve their job security. 61 Graham and Harvey, 1999, survey 392 chief financial officers and find that executives are not concerned about many financial theories such as asset substitutio n, free cash flows or asymmetric information, but rather are concerned with earning per share dilution and recent stock price appreciation. 62 Loughran and Vijh, 1997, and Travlos, 1987.
94 that both the signaling hypothesis and tax hypothesis play an important role in the financing plan. Firms that repurchase shares may signal undervaluation and firms that use stock to finance the acquisition in this situation must use purchase accounting. And although purchase accounting will reduce the boo k value of earnings per share, it also will reduce the firms taxes and thus increase its cash flows. Prediction, Data and Methodology Hypotheses In my previous chapter I suggest that an opportunistic reason for a repurchase is to fund an acquisition, and thus positive stoc k price reactions may not be anticipated. Based on previous research,63 if a repurchase is conducted in order to finance an acquisition it may also carry with it the poor stock return reacti ons that have been associated with bidder firms conducting acqui sitions. However, researchers have made a clear distinction betwee n cash-financed acquisitions and st ock-financed acquisitions. If a firm uses cash to repurchase shares which are then used to acquire a target firm, this is not straight cash or straight stock-financed. Many research ers have documented losses to bidding firms that use stock. The use of repur chased shares to conduct an acquisition is stock-financed and may result in the negative abnormal re turns associated with stockfinanced acquisitions. On the other hand, using repurchased stock to finance an acquisition is just adding a step to a cash-f inanced acquisition and t hus may act according 63 Travlos, 1987; Morck, Shleifer and Vishny, 1990; Loughran and Vijh, 1987; and Rau and Vermaelen, 1982.
95 to previous research and have no negative a bnormal returns or possibly slightly positive returns. Additionally, using a repurchase to f acilitate an acquisition begs further investigation. Why would a firm go through such transactional gymnastics? It would be simpler and less costly in time and dollars to just conduct an acquisition with cash.64 Therefore, there must be some benefit to taki ng on this additional cost. It may be that the premium to acquire is less with a stock-fina nced acquisition than with a cash-financed acquisition for the bidding firm will not need to compensate the target firm for its immediate tax consequences. 65 It is possible that the repurchase ann ouncement gives managers the anticipated positive stock price return reaction which more than offsets the anticipated decrease in stock price with an acquisition announcemen t. In a sense, this may extinguish the negative return reactions associated with a straight stock offering and allow bidder managers to pay a smaller premium at the acqui sition. If this is the case, I expect that these firms may have better long-term performa nce than firms that do not take the extra transactional step since they would be less likely to overpay for the acquisition. 66 Finally, purchasing accounti ng does carry a long term tax advantage. Normally stock offered acquisitions do not use purchas e accounting. However, if the firm uses 64 Loughran, Tim, and Anand M Vijh, 1997, Do Long-Term Shareholders Benefit From Corporate Acquisitions?, Journal of Finance 52, 1765-1790. During a five-year period following the acquisition, on average, firms that complete stock mergers earn significantly negative excess returns of -25.0 percent whereas firms that complete cash tende r offers earn signific antly positive excess returns of 61.7 percent. 65 Martin, 1996, finds that the higher the bidding firms investment opportunity set the more likely the firm will choose to use stock financing. 66 Roll, Richard, 1986, The Hubris Hypothesis of Corporate Takeovers, The Journal of Business Vol. 59, pp. 197-216. Roll argues that firms on average pay too much for an acquisition and thus the poor post
96 repurchased shares it can only proceed with purchase accounting. This is an advantage to the long-term cash flows of the combined firms. In order to test, I will conduct a diffe rence in means between firms announcing both a repurchase jointly with an acquisi tion and firms that announce an acquisition without a repurchase. Hypothesis 1 Abnormal return (at the announcement date and long-term post announcement) will be less negative for firms that a nnounce repurchase intentions with an acquisition announcement than for firms that only announce the acquisition. This test will be performed at the announcement date for announcement date effects and also three-year and four-year post announcement. Table 3-1 summarizes the hypotheses put fo rth in the literature. Most of the hypotheses make predictions on the method of payment choice. I question why firms would use cash to repurchase shares in order to conduct a stock-financed acquisition. Since the bidder firm wealth is not hurt by cash acquisitions and the combined firm wealth is, on average, better with cash, it is perplexing as to w hy a firm would incur additional transactions fees a nd most likely incur labor costs to take this extra financing step that at first glance does not appear to carry benefits. I review the hypothesis with this question in mind. My sample is of firms which either have the cash available at the repurch ase announcement or did not make a credible repurchase announcement. If they have the ca sh available, then according to the cash availability hypothesis they will prefer to us e it if they are underv alued. Since the firm announcement returns s hould be expected. So, if firms are able to decrease the premium paid, their post
97 has chosen not to use the cash for the acquis ition, but rather for the repurchase, the cashavailability hypothesis suggests that the firm is overvalued. However, if the firm is overvalued it is not likely it would choose to repurchase its own stock (see Chapter 2). Thus, it is feasible that using cash directly to purchase another firm or using cash indirectly with repurchased sh are financing is inconsequential to the cash availability hypothesis in that both announcements are i ndicative of undervaluated bidder shares. The investment opportunity hypothesis predicts that a high-growth bidder will prefer stock because it will afford the high-grow th firm with future financial flexibility. This hypothesis is not applicable to cash flus h firms with moderate growth. The signaling hypothesis is a little problematic in that the repurchase signals undervaluation and the subsequent stock financing signals overvaluatio n. Although it is unlikely that a firm sets out to send mixed signals, it is possible that a fi rm prefers to use stock (ie. for risk sharing and future tax benefits) and plans to mitigate the bad news of overvaluation indicated with a stock financing by offsetting with the undervaluation signal of the repurchase announcement. The risk-sharing hypothesis is consistent w ith the extra financing. If a firm is concerned about the post-merger performance of the target firm then stock financing will mitigate this concern. Thus, if the target firm will represent a significant portion of the combined firm, it may be the preference of the bidder firms managers to share the risks, even if evidence of poor stock fina nced acquisitions is predominant. returns may not be as poor.
98 The target firm managers may have a pr eference for stock financing in order to maintain some control in the merged firm.67 Thus, if the target is large enough in comparison to the bidder and the target firms managers have some control, they may be in the position to influence the financing decisi on. In the extreme, the target may be able to influence the bidder to first repurchase its shares and then to pass the shares on to the target firms shareholders. This argument ma y hold for the target manager shareholders; however, the argument fails for all the other target sharehol ders who should prefer cash due to the higher premium. It has also been suggested; however, that the higher premium is nothing more than compensation for the fo rced tax consequences and thus the high premium quickly disappears net of taxes. The control hypothesis states that if a ma nager desires to maintain his ownership position in the firm, he or she will prefer stock to finance an ac quisition in order to maintain control. A repurchase decreases the total outstanding shares and thus serves to increase the ownership positi on of the non-tendering sharehol ders. Thus, managers with a high concern for their ownership position wo uld favor repurchase of shares first to mitigate the loss in ownership position if a stock-financed acquisition was pursued over the preferable cash acquisition. Pooling accounting (stock financing) and repurchasing ac tivities are both consistent with manager objectives of increas ing earnings per share. Thus, if a managers compensation were tied to earnings per sh are, both repurchasing shares and stock financed acquisitions would supplement the managers compensation. Thus, the pooling 67 Ghosh and Ruland, 1998.
99 versus purchasing hypothesis would be cons istent with the do ubled transactions. Furthermore, the doubled transactions may create favorable tax results. Purchase accounting creates a tax burden on the target firm. Thus, stock financing is beneficial for both risk-sharing and tax consequences. Cash financing has historic ally better returns. Thus, it is possible that by taking on the extra transactions the firm is taking advantage of both types of financing and ente ring into a win-win situation. Finally, if it is not the method of paym ent that matters but only whether the acquisition is a good fit and increases the focus of the firm, then the transactions that preceded the acquisition may not be the impor tant issue. This argument suggests that although it appears inefficient to use cash to repurchase shares to be used for the acquisition of another firm, th is method of payment may not be predictive of poor postmerger stock price returns that have been documented by numerous researchers. If the bidder acquires a firm that increases its focu sed line of business then value should be enhanced and the method of payment is immate rial. Similarly, if the bidder attempts a diversifying acquisition, the market would be expected to respond negatively. These studies suggest that a viable control for a valueenhancing merger versus a value-decreasing merger could be determin ed by whether the merger increases or maintains its focus or decreases its focus in diversification attempts. Flanagan and O'Shaughnessy (2003) use primary SIC codes to cl assify transactions core-related in their paper that explores which firm charact eristics influence the size of acquisition
100 premiums.68 Flanagan and OShaughnessy classify an acquisition as core-related if both the acquiring and target firms share the same three or four digit SIC code. I will separate my firms that announce repurchase intenti ons to conduct an acquisition as value enhancing if the firms have the same three or four digit SIC code and are thus corerelated focus increasing or preserving firms. Firms will be considered focus decreasing if the acquiring and target firms do not share th ree or four digit SIC codes and appear unrelated. Sample The sample of firms announcing a repurchas e in order to facilitate an acquisition are collected from Securities Data Corporations Mergers and Acquisitions database and Repurchases database. I begin by collecting all repurchase offers with an acquisition (ACQ) purpose. Financial firms (SIC codes 6000-6999) and regulated utilities (SIC codes 4910-4949) were removed because they are believed to face a different incentive structure around repurchase activ ity. Imposing these restrictio ns results in an initial sample of 103 firms with a repurchase announcement between 1995 and 2002. The sample is reduced to 96 firms with usable return information available from CRSP. Using the same database, I searched fo r acquisition announcement dates one year before and one year after the sample firms repurchase announcement and found that two-thirds (66) of the sample firms made both the repurchase and acquisition announcement on the same date. Of those 66 fi rms, nearly one-half (32) had announced 68 Flanagan and O Shaughnessy, 2001 explore the relationship between relatedness and takeover premiums and find that acquires that are not core-related to pay very high premiums when multiple bidders are
101 the acquisition at a previous date in a ddition to the second announcement made in conjunction with the repurchase announcemen t. Of the firms that did not make the acquisition announcement of the same date as the repurchase announcement, ten of them made the acquisition announcement prior to the repurchase announcement and eleven made the acquisition announcement after the re purchase announcement. Subsequent to the acquisition announcement, twelve firms withdrew their announcement. Methodology ( See chapter 2, page 35. ) Results Table 3-3 presents abnormal return data for my 96 firms that announce repurchase-financed acquisitions. The returns are relative to the repurchase announcement date with the exception of Panel C, which is at an acquisition announcement date and Panel G which is at the repurchase withdr awal date. Panel A shows the full sample of 96 firms. The CA Rs show the genera lly positive abnormal returns consistent with other researchers results. Two-thirds of the sample firms anno unce the repurchase and the acquisition on the same date. The abnormal returns to this group shown in Panel B are similar, if not a little more significant then the entire sample of firms. Panel C is very interesting in that onethird of the repurchas ed-financed firms had acquisition announcements prior to making the repurchasing announcements. Thus, at the present. Although, the emphasis of the paper is not on the independent core-related variables, I plan to follow their procedure for defining core-relatedness by SIC codes.
102 time of the first acquisition announcement there would be little indication that the firms planned repurchase as part of the financing. Thus, ex ante the abnormal re turn reaction should be similar to all other acquisitions. Panel C, although only slightly significant, shows generally positive results, which is contrary to the prevailing documentation on acquisition returns. Panel D is a very small group of only ten firms that made an acquisition announcement and later announced a repurchase. Panel D shows very little significance due to the small sample size. Panel E is also a very small group. However, these eleven firms show some very significantly positive results. These firms made their repurchase announcement in advance of their acquisition announcement and have had exceptional market model value weighted CARs of 96.94 percent two-year s post the repurchase announcement and 159.64 percent three-years post. Th e beta of this group of firms is only .68 and thus the argument of being compensated for risk does not seem viable. Twelve of my 96 firms later withdrew th eir repurchase intent ions. The returns of these firms are displayed in Panel F. It appears that at the initial repurchase announcement these firms enjoy similar posi tive reactions accorded to repurchase announcing firms. Thus, there is no indicat ion that the market expects the later withdrawal. However, in the long run these fi rms do not do as well as firms that carry out the repurchase plan. Panel G presents the sa me sub-sample of firms at the withdrawal date.
103 Table 3-4 directly tests h ypothesis 1 and finds that prior to the acquisition announcement both the stock-financed and cashfinanced firms show the characteristic negative abnormal returns. The repurchas e group shows no abnormal returns and the groups are not different from each other. At the acquisition announcement event date all groups show moderate positive abnormal returns. The most interesting results begin to appear within 90 days of the acquisition announcement where the groups become very different from each other. The cash-financed (-1.8 percent) and stock-financed (-6.3 percent) acquisitions show negative abnormal returns, whereas the repurchase-financed acquisition is slightly positive (1.3 percent). Th is distinction continues into the long-term with significantly negative abnormal return s for both the cash-financed (-33.4percent for two-year post) and stock-financed (-99.7 pe rcent for two-year post) acquisitions and significantly positive for the repurchase-financed (11.8 percent for twoyear post) returns. Thus, firms that take on the extra transacti ons seem to be well-compensated for their efforts. This table strong ly supports my hypothesis. No t only do these repurchasedfinanced acquisition firms not exhibit the char acteristic negative abnormal returns of both cash-financed and stock-financed, these firms show positive CARs two-years, three-years and four-years post announcement. I attribute this to the firms reducing their tax burden by completing a purchase accounting acquisition. Straight cash-financed acquisitions also have this advantage; however, a cash-financed acquisition is not able to share the risk with the target shareholders in the merged firm. Furthermore, a straight cash-financed acquisition may have to pay a premium to targ et shareholders to compensate the target
104 shareholders with an increase in tax burden due to their most likely gain on the stock sale.69 Table 3-6 is a multi-model table. The five panels A-E display the coefficients for pre-announcement, at the announcement and post-announcement time periods (30 day, two-year and three-year). Four multi-variate models are displayed in each panel. Each model uses a combination of control variab les (market value three-year growth, threeyear sales growth, beta, earnings per share th ree-year growth, price to book ratio, and free cash flow) as well as the firms method of acquisition payment. These models show a dummy variable if the acquisition was finan ced with a repurchase or another dummy variable if the acquisition was financed w ith 100 percent cash. The base case is if the acquisition is 100-percent stock financed, and thus the para meter is zero and is not shown. Panel A shows the cumulative abnorma l returns prior to the acquisition announcement. There appears to be no differen ce in the abnormal return performance of the sample of firms based on their method of financing their acquisitions prior to the announcement. Panel B displays the event-date anno uncement CARs. At the announcement three of the models show significant positive coefficients for both cash-financed acquisitions in comparison to the base case of the stock-financed acquisition, and one model shows significant positive coefficients for the repurchased-financed acquisitions. Panel C 69 As an aside it may be that some firms with available cash do not take advantage of this double transactional step due to the advantage of cash-finan ced acquisitions being quick, allowing firms to avoid undue competition.
105 displays the short-term stock price re actions from two-days to 30-days post announcement. This time periods results ar e very similar to the announcement-event period return. However, the size-control variab le; market value growth for the last three years, and the risk control variable; beta, both show negative coefficients. As shown in the first two panels, panel C also displays a very small R-square. Panels D and E show very interesting re sults and also much higher R squares (ranging between 26 and 41 percent) for the two-years and three-years post announcement, respectively. During these time pe riods, we find that the coefficients for the cash-financed acquisition are mixed and not significant in comparison to the stockfinanced acquisition. However, in the longterm the repurchase acquisition groups coefficients become strongly significantly posit ive. Consistent with table 34, it appears that firms that finance acquisitions with repur chased shares do very well in the long term. Similarly to chapter 2, ExecuComp data was obtained for the three types of distinct acquisition financ ing groups. Data on 175 firms us ing only cash financing, 100 firms using only stock financi ng and only three firms using repurchases financing were found. Due to the extremely small sample si ze of the repurchasing acquisition firms, no further testing was attempted to differentiate the officers stock ownership or options. There was, however, information that c ould be obtained through the compustat database to differentiate the firm choice of acquisition financing. The mean market value of the firms, the return on assets (ROA), re turn on equity (ROE), net income, and free cash flows for all three groups of firms is shown in table 3-5. Firms choosing to repurchase shares to finance an acquisition ar e larger, have higher returns on assets and
106 returns on equity and have significantly higher free cash flows. Furthermore, all firms that conduct repurchases financing have pos itive net incomes duri ng the year of the acquisition whereas; only 80 pe rcent of the cash-financed firms and 73 percent of the stock-financed firms can make the same claim. Conclusion Firms that take on the extra transactional st ep of repurchasing shares to finance an acquisition are well compensated for their effort s, especially in the long run. These firms have cash available and positive earnings but on average have negative abnormal returns prior to their repurchase announcements. Thus these firms are likely to be undervalued and therefore choose this method of financi ng to signal undervaluation in the market place. These firms experience risk sharing with the target firms, counteract the negative effects of dilution by repurchasi ng shares first, and enjoy a ta x advantage for their efforts. My results raise the question as to why more firms do not take advantage of this win-win situation. Aboddy, Kasnik and Willia ms (2000) argued that managers should prefer pooling accounting because all else e qual, purchase accounting hurts net income and earning per share for years to come. I s uggest that since management compensation is likely tied to these performance meas ures, most managers do prefer pooling accounting. However, this is shortsighted.
107 Table 3-2 Variable Definitions Beta = CRSP beta. Cash 100% = Bidding firm acquired target firm with 100 percent cash financing. EPS growth = EPSgrow = the three-year least squares annual growth rate of Net Income before extraordinary items and disconti nued operations less preferred dividend requirements (ExecuComp EPSEX3LS). Free CF = The free cash flow concept is Op erating Activities Net Cash Flow minus Cash Dividends minus Capital Expenditures (Compustat OANCF-DV-CAPX). Market Value = This data item provides a pr e-calculated company-level market value based upon the sum of all the companys trad ing issues multiplied by their respective closing price (Compustat PRCC CSHO) a nd is reported in millions of dollars. Market Value3 = The 3-year least square s annual growth rate in market value (ExecuComp MKTVAL3LS). NI = The income or loss reported in millions of dollars by a company after expenses and losses have been subtracted from all revenue s and gains for the fiscal period including extraordinary items and discontinued opera tions (Compustat annual data item A172). Price to book = The market value of assets divided by the book value of assets, where the market value of assets is the book value of assets plus the market value of equity (Compustat item #24 times Compustat item #25) minus the book value of equity. ROA = Return on Assets is Income Before Extraordinary Items Available for Common, divided by Total Assets, which is defined as th e sum of current assets, net property, plant, and equipment, and other non-current assets This is then multiplied by 100 (Compustat IBCOM/AT)*100. Repurchase = Bidding firm acquired targ et firm with repurchased shares. ROE = Return on Equity is Income Before Extraordinary Items Available for Common, defined as income before extraordinary it ems and discontinued operations less preferred dividend requirements, but before adding sa vings due to common stock equivalents, divided by Common Equity To tal, which is defined as the common shareholders' interest in the company. The result is multiplied by 100 (Compustat IBCOM/CEQ)*100). Sales3 = The 3-year least squares annual gr owth rate in sales (ExecuComp SALES3LS). Stock 100% = Bidding firm acquired target firm with 100 percent stock financing.
108 Table 3-3 Repurchase to Fund an Acquisition Cumulative Abnormal Market Model Returns (CARs) for both equal and value weighted portfolios of firms announcing a repurchase from 1995-2002 for the purpose of c onducting an acquisition. Panel A: All Firms Announcing Repurchase Plans to Fund an Acquisition. (n = 96, mean beta = 0.75). Abnormal returns are relative to the repurchase announcement date. Window Market Model Scholes-Williams Equal Wt Value Wt Equal Wt Value Wt -30-2 -2.84% -2.67% -1.14% -3.64% -1,+1 1.57% ** 1.68% *** 1.67% ** 1.56% ** +2,+30 0.91% 1.97% 1.05% 1.27% +31,+504 29.42% *** 30.47% *** 49.41% *** 17.94% *** +31,+756 41.49% *** 45.92% *** 72.15% *** 26.75% *** significant at .10, ** significant at .05, *** significant at .01 Panel B: This panel is a sub-samp le of panel A. This sub-sample includes the firms making both announcements on the same date. (n = 66, mean beta = 0.92) Window Market Model Scholes-Williams Equal Wt Value Wt Equal Wt Value Wt -30-2 -1.05% -0.39% -1.42% -0.80% -1,+1 2.31% *** 2.36% *** 2.29% *** 2.25% *** +2,+30 2.40% 3.22% 2.02% 2.73% +31,+504 34.20% ** 34.78% ** 32.22% *** 30.82% ** +31,+756 46.11% ** 50.36% *** 42.37% *** 44.05% *** significant at .10, ** significant at .05, *** significant at .01 Panel C: This panel is a sub-sample of panel B. The repurchase and the acquisition were announced on the sa me date, however the acquisition also had an earlier announcement. (n = 32, mean beta = 0.90) Window Market Model Scholes-Williams Equal Wt Value Wt Equal Wt Value Wt -30-2 3.98% 4.65% 3.64% 3.82% -1,+1 1.27% 1.28% 1.18% 1.43% +2,+30 1.40% 0.16% 0.64% 0.27% +31,+504 15.11% 16.68% 13.18% 13.93% +31,+756 30.72% ** 32.77% ** 28.93% ** 29.81% significant at .10, ** significant at .05, *** significant at .01
109 Table 3-3 Repurchase to Fund an Acquisition (continued) Panel D: This is a sub-sample of firms that only announced their acquisition intentions in advance of their repurchase announcement. The returns are shown relative to the repurchase announcement. (n=10, mean beta = 0.76) Window Market Model Scholes-Williams Equal Wt Value Wt Equal Wt Value Wt -30-2 -6.16% -3.01% -4.52% -2.82% -1,+1 0.83% 1.42% 0.89% 1.35% +2,+30 -4.59% -1.12% -4.47% -1.46% +31,+504 -1.39% 15.07% 1.95% 16.96% +31,+756 1.17% 24.83% 6.42% 29.59% significant at .10, ** significant at .05, *** significant at .01 Panel E: This sub-sample of firm s made acquisition announcements and a subsequent repurchase announcement. The returns are shown relative to the repurchase announcement. (n=11, mean beta 0.68) Window Market Model Scholes-Williams Equal Wt Value Wt Equal Wt Value Wt -30-2 -1.50% -1.86% -2.05% -2.66% -1,+1 2.88% 3.08% 2.89% 2.97% +2,+30 11.86% 10.17% 12.29% 11.01% +31,+504 93.94% *** 96.94% *** 91.29% *** 98.81% *** +31,+756 158.45% *** 159.64% *** 151.81% *** 163.23% *** significant at .10, ** significant at .05, *** significant at .01
110 Table 3-3 Repurchase to Fund an Acquisition (continued) Panel F: A small sub-sample of firms announced repurchases and later withdrew. This is at the repurch ase announcement date. (n=12, beta = 1.18) Window Market Model Scholes-Williams Equal Wt Value Wt Equal Wt Value Wt -30-2 2.18% 1.69% 2.93% 1.42% -1,+1 2.00% ** 1.62% 2.04% ** 1.64% +2,+30 1.39% 0.24% 1.00% 0.11% +31,+504 -18.90% -19.49% -17.79% -21.54% +31,+756 -22.62% -32.44% -20.08% -34.58% significant at .10, ** significant at .05, *** significant at .01 Panel G: A small sub-sample of firms announced repurchases and later withdrew. This is at the withdr awal date. (n=12, beta = 0.92) Window Market Model Scholes-Williams Equal Wt Value Wt Equal Wt Value Wt -30-2 0.46% -0.88% 0.92% -0.44% -1,+1 -1.37% -1.30% -1.34% -1.22% +2,+30 0.69% 1.95% 0.70% 1.66% +31,+504 6.75% 8.86% 3.07% 8.95% +31,+756 5.32% 5.10% 3.26% 5.71% significant at .10, ** significant at .05, *** significant at .01
111 3-4 Comparison of Acquisiti on with and without a Repurchase CARs using the market model are shown fo r firms announcing an acquisition from 1995 2002. The firms are separated into firms that also announced a plan to repurchase stock in conjunction with the acqui sition announcement (Re purchase) and firms that financed the acquisition with 100 percent cash (Cash) a nd firms that financed the acquisition with 100 percent stock (Stock). ANOV A differences follow for the group differences of the abnormal returns for the seven event time pe riods: 30 days to two days prior to the acquisition announcement (-30, -2), the event period (-1, +1), 30 days postannouncement (+2, +30), 90 days post-announcement (+31, +90), one-year, two-years, and three-years post-announcement (+31, + 252), (+31, +504), (+31, +756), respectively. Repurchase Cash 100% Stock 100% P-value Beta 0.98 1.12 1.24 n 436 9205 4497 (-30,-2) 0.000 -0.005 ***-0.006 0.960 (-1,+1) 0.005 0.011 ***0.009 0.449 (+2,+30) 0.013 -0.018 ***-0.063 *** 0.000 (+31,+90) -0.018 ***-0.041 ***-0.102 *** 0.000 (+31,+252) 0.009 -0.158 ***-0.441 *** 0.000 (+31,+504) 0.118 ***-0.334 ***-0.997 *** 0.001 (+31,+756) 0.199 ***-0.417 ***-1.355 *** 0.199 Table 3-5 Acquiring Firm Characteristics Data is obtained from Compustat for market value, return on assets return on equity, net income and free cash flows for firms choos ing to acquire firms by financing with repurchases, cash, and stock. Mean values are shown. Market ValueROA ROE NI Free Cash Flow Repurchase 14,327 3.6 15.2 1,122 1,962 Cash 100% 9,425 1.3 1.2 261 164 Stock 100% 6,955 -175.2 -5.3 145 48 p-value 0.004 0.277 0.353 0.000 0.000
112 Table 3 6 Market Reaction to Type of Financed Acquisition Announcement The acquisitions are financed by a re purchase, 100 percent stock or 100 percent cash. Shown are the dummy variables if the acquisition was financed with a repurchase and finan ced with cash, otherwise stock. (tstatistics in parentheses; ***, **, a nd denote significant levels at 1%, 5% and 10% levels). Models include firms announcing acquisitions from 1995 to 2002. Returnit = b0 + b1Xit + b2year Indicators + eit A. Return for (-30, -2) Variable (1) (2) (3) (4) Intercept 0.01705 0.02048 -0.02080 0.00551 (0.84) (1.03) (-3.52) (0.76) Dummy if cash -0.00939 -0.01214 0.02319 ***-0.01197 (-.0.60) (-0.78) (4.52) (-1.58) Dummy if rep. -0.06016 -0.06254 0.01239 -0.00790 (-1.00) (-1.05) (1.14) (-0.42) Mkt Val 3 year 0.00015 0.00014 0.00004 -0.00008 (0.97) (0.91) (1.37) (-1.50) Sales 3 year -0.00019 -0.00017 -0.00007 0.00006 (-0.66) (-0.61) (-0.97) (0.94) Beta -0.02375 **-0.02517**0.00081 (-2.01) (-2.17) (0.30) EPS 3 year -0.00020 -0.00020* (-1.63) (-1.65) Price/Book 0.00023 0.00012 (0.37) (0.19) Free cash flow 0.00000 (-0.25) R-square 0.0121 0.0126 0.0178 0.0017 n 782 793 1235 2664
117 References Aboody, David, 1996, Market valuation of employee stock options, Journal of Accounting and Economics 22, 357-391. Aboody, David, Ron Kasznik, and Michael W illiams, 2000, Purchase versus pooling in stock-for-stock acquisiti ons: Why do firms care?, Journal of Accounting and Economics 29, 261-286. Amihud, Yakov, Baruch Lev, and Nicholas G. Travlos, 1990, Corporat e control and the choice of investment financing: The case of corporate acquisitions, Journal of Finance 45, 603-616. Andrade, Gregor, 1999, Do Appearances Matt er? The Impact of EPS Accretion and Dilution on Stock Prices, Working Paper, Harvard Business School 1-60. Arnold, Marcus C., and Robert M. Gill enkirch, 2002, Stock Options as Incentive Contracts and Dividend Policy, Working Paper 1-26. Asquith, Paul, and David Mullins, 1986, Equity Issues and Offering Dilution, Journal of Financial Economics 15, 61-89. Bagwell, Laurie Simon, 1991, Share Repurchas e and Takeover Deterrence" Rand Journal of Economics, The Rand Journal of Economics 22, 72-88. Barth, M., H. Elliot, and M. Finn, 1995, Market Rewards Associated with Increasing Earnings Patterns, Working paper, Cornell University Bens, Daniel A., Venky Nagar, Douglas J. Skinner, and M.H. Franco Wong, 2002, Employee Stock Options, EPS Dilution, and Stock Repurchases, Working Paper 1-50. Berger, Philip G., Eli Ofek, and David L. Yermack, 1997, Managerial Entrenchment and Capital Structure Decisions, The Journal of Finance 52, 1411-1438. Black, Bernard, 1992, Agents watching agents: The promise of institutional investor voice, UCLA Law Review 39, 811-893. Blasi, Joseph, Douglas Krus e, and Aaron Bernstein, 2003. In The Company of Owners: The Truth About Stock Options and Why Every Employee Should Have Them (Basic Books, New York). Bowen, Sarah, 1996, Enough! Shareholders lik e the idea of using stock as a pay incentive, but now, some way, it has gone to far, Wall Street Journal (New York). Bradley, Michael, A. Desai, and E. Han Ki m, 1988, Synergistic gains form corporate acquisitions and the division between the stockholders of target and acquiring firms, Journal of Financial Economics 21, 3-40. Bradley, Michael, Gregg Jarrell, and E. Han Kim, 1984, On the existence of an optimal capital structure: Theory and evidence, Journal of Finance 39, 857-878.
118 Canina, Linda, Roni Michaely, Richard Thaler, and Kent Womack, 1998, Caveat Compounder: A Warning about Using the Daily CRSP Equal-Weighted Index to Compute Long-Run Excess Returns, The Journal of Finance 53, 403-416. Chang, Chun, 1993, Payout Policy, Capital Struct ure, and Compensation Contracts when Managers Value Control, Review of Financial Studies 6, 911-933. Comment, Robert, and Gregg Jarrell, 1991, The relative signaling power of DutchAuction and fixed-price se lf-tender offers and open-market share repurchases, Journal of Finance 46, 1243-1271. Conrad, Jennifer, and Gautam Kaul, 1993, Longterm market overreaction or biases in computed returns?, Journal of Finance 48, 39-63. Cook, Douglas O., Laurie Krigman, and J. Chris Leach, 2001, On the Timing and Execution of Open Market Repurchases, Working Paper Copeland, Thomas E., and J. Fred Weston, 1988. Financial Theory and Corporate Policy (Addison-Wesley Publishing Company, Reading). Dann, Larry Y, 1981, Common Stock repurchas es: An Analysis of returns to Bondholders and Stockholders, Journal of Financial Economics 9, 113-138. Dann, Larry Y., Ronald W. Masulis, and Da vid Mayers, 1991, Repurchase tender offers and earnings information, Journal of Accounting and Economics 14, 217-251. Dennis, Debra K., and John J. McConnell, 1986, Corporate Mergers and Security Returns, Journal of Financial Economics 16, 143-186. Der Howanesian, Mara, 2002, The Buyback Boomerang, Businessweek. Dittmar, Amy K., 2000, Why Do Firms Repurchase Stock?, The Journal of Business 73, 331-355. D'Mello, Ranjan, and Pervin K. Shroff, 2000, Equity Undervaluation and Decisions Related to Repurchase Offers: An Empirical Investigation, Journal of Finance 55, 2399 2424. Erwin, Gayle R., and James M. Miller, 1998, The Intra-Industry Effect s of Open Market Share Repurchases: Contagion or Competitive?, The Journal of Financial Research 21, 389-406. Evans, John P., and James A. Gentry, 1999, Do Strategic Share Repurchase Programs Create Long-Run Firm Value, Universi ty of Illinois Working Paper #99-0119. Fenn, George W., and J. Nellie Liang, 2001, Corporate Payout Policy and Managerial Stock Incentives, Journal of Financial Economics 60, 45-72. Fenn, George W., and Nellie Liang, 1998, Good News and Bad News about Share Repurchases, FEDS Paper No. 98-4. Flanagan, David J., and K. C. O'Shaughne ssy, 2003, Core-related acquisitions, multiple bidders and tender offer premiums, Journal of Business Research 56, 573-585. Gay, Gerald D., Jayant R. Kale, and Th omas H. Noe, 1996, (Dutch) Auction Share Repurchases, Economica 63, 57-80. Gelb, David, 2000, Payout Composition and Inve stors' Reaction to Dividend and Stock Repurchase Announcements, Working Paper 1-26. Gillan, Stuart L., 2001, Option-based Comp ensation: Panacea or Pandora's Box?, Journal of Applied Corporate Finance 14.
119 Graham, John R., and Campbell R. Harvey, 2001, The Theory and Practice of Corporate Finance: Evidence from the Field, Journal of Financial Economics 60, 187-243. Grullon, Gustavo, and Roni Michaely, 2002, Dividends, Share Repurchases, and the Substitution Hypothesis, Journal of Finance 57, 1649 1684. Grullon, Gustavo, and Roni Michaely, 2003, The Information Content of Share Repurchase Programs, Journal of Finance forthcoming. Guay, Wayne, and Jarrad Harford, 2003, Th e Cash-Flow Permanence and Information Content of Dividend Incr eases vs. Repurchases, forthcoming, Journal of Financial Economics Guay, Wayne R., 2002, Discussion of: Real Inve stment Implications of Employee Stock Option Exercises, Journal of Accounting Research 40, 395-406. Hansen, Robert G., 1987, A theory for the ch oice of exchange medium in mergers and acquisitions, Journal of Business 60, 76-95. Harford, Jarrad, 1997, Manage rial Discretion in Repur chase Tender Offers: Do Shareholders Benefit?, Working Paper 1-36. Healy, Paul M., 1995, The effect of bonus schemes on accounting decisions, Journal of Accounting and Economics 7, 85-107. Hermalin, Benjamin E., and Michael S. Weisbach, 1991, The effects of board composition and direct incentives on firm performance, Financial Management 20, 101-112. Himmelberg, Charles P., R. Glenn Hubbard, and Darius Pa lia, 1999, Understanding the Determinants of Managerial Ownership and the Link Between Ownership and Performance, Journal of Financial Economics 53, 353-384. Hodrick, Laurie Simon, 1999, Does stock price elasticity affect corporate financial decisions?, Journal of Financial Economics 52, 225-256. Holderness, Clifford G., and Dennis P. Sheeh an, 1988, The role of majority shareholders in publicly held corporations, Journal of Financial Economics 20, 317-346. Hubbard, R. Glenn, and Darius Palia, 1996, Bene fits of Control, Managerial Ownership, and the Stock Returns of Acquiring Firms, NBER Working Paper.W5079 1-24. Ikenberry, David L., Josef Lakonishok, and Th eo Vermaelen, 2000, Stock Repurchases in Canada: Performance and Strategic Trading, Journal of Finance 55, 2373-2397. Ikenberry, David L., and Theo Vermae len, 1996, The Option to Repurchase Stock, Financial Management 25, 9-24. Ikenberry, David, Josef Lakonishok, and Th eo Vermaelen, 1995, Market underreaction to open market share repurchases, Journal of Financial Economics 39, 181-208. Jagannathan, Murali, and Clifford P. Stephe ns, 2003, Motives For Multiple Open-Market Repurchase Programs, Financial Management 32, 71-91. Jagannathan, Murali, Clifford P. Stephens and Michael S. Weisbach, 2000, Financial Flexibility and The Choice Between Dividends and Stock Repurchases, Journal of Financial Economics 57, 355-384. Jagannathan, Murali, Clifford P. Stephens and Michael S. Weisbach, 1998, Financial Flexibility and The Choice Between Dividends and Stock Repurchases, Working Paper 1-35.
120 Jensen, 1986, The Agency Costs of Free Cash Flow, Corporate Finance, and Takeovers, American Economic Review 76, 323-329. Jensen, Michael C., 1991, Corporate c ontrol and the politi cs of finance, Journal of Applied Corporate Finance 4, 13-33. Jensen, Michael C., and Richard S. Rubac k, 1983, The Market for Corporate Control, Journal of Financial Economics 11, 5-50. Jensen, Michael, and W Meckling, 1976, Theory of the Firm: Managerial Behavior, Agency Costs, and Capital Structure, Journal of Financial Economics 3, 305-360. Jin, Li, and Lisa K. Meulbroek, 2001, Do Underwater Executive Stock Options Still Align Incentives? The Effect of Stock Price Movements on Managerial IncentiveAlignment, Harvard Business School Working Paper No. 02-002; EFA 2002 Berlin Meetings Discussion Paper 1-71. John, Teresa A., and Kose John, 1993, TopManagement Compensation and Capital Structure, The Journal of Finance 48, 949-974. Jolls, Christine, 1998, Stock Repurchases and Incentive Compensation, National Bureau of Economic Research (NBER) 1-36. Jung, Kooyul, Yong-Cheol Kim, and Rene M. Stulz, 1996, Timing, investment opportunities, managerial discretion, and the security issue decision, Journal of Financial Economics 42, 159-186. Kahle, Kathy, 2002, When a Buyback Isn't a Buyback: Open Market Repurchases and Employee Options, Journal of Financial Economics 63, 235-261. Kaplan, Steven N., and Michael S. We isbach, 1992, The Success of Acquisitions: Evidence from Divestitures, The Journal of Finance 47, 107-138. Klassen, Kenneth J., and Ranjini Sivakumar, 2001, Stock Repurchases Associated with Stock Options Do Represent Dollars Out of Shareholders' Wallets, EFMA 2001 Lugano Meetings 1-31. Lakonishok, Josef, and Theo Vermaelen, 1990, Anomalous Price Behavior Around Repurchase Tender Offers, Journal of Finance 45, 455-477. Lambert, Richard A., William N. Lanen, and David F. Larcher, 1989, Executive stock option plans and corporate dividend policy, Journal of Financial and Qualitative Analysis 24, 409-425. Lang, Larry H. P., Rene M. Stulz, and Ral ph A. Walkling, 1991, A test of the free cash flow hypothesis: The case of bidder returns, Journal of Financial Economics 29, 315-336. Lang, Larry, Rene M. Stulz, and Ralph A. Walkling, 1989, Tobin's q and the gains form successful tender offers, Journal of Financial Economics 24, 137-154. Lee, Susan, 2002, The Ugly Option, Wa ll Street Journal (New York). Lewellen, Wilbur, Claudio Loderer, and Ah ron Rosenfeld, 1985, Merger decisions and executive stock ownership in acquiring firms, Journal of Accounting & Economics 7, 209-231. Li, Kai, and William J. McNally, 1999, Op en Market versus Tender Offer Share Repurchases: A Conditional Event Study, University of British Columbia, Finance Working Paper No. 98-2
121 Liang, J. Nellie, and Steven A. Sharpe 1999, Share Repurchases and Employee Stock Options and their Implications for S& P 500 Share Retirements and Expected Returns. Lie, Erik, 2000, Excess Funds and Agency Pr oblems: An Empirical Study of Incremental Cash Disbursements, Review of Financial Studies 13, 219-248. Lie, Erik, and John McConnell, 1998, Earnings signals in fixed-price and Dutch-auction self-tender offers, Journal of Financial Economics 49, 161-186. Linn, Scott C., and Jeannette A. Switzer, 2001, Are cash acquisitions associated with better postcombination operating perf ormance than stock acquisitions?, Journal of Banking & Finance 25, 1113-1138. Loughran, Tim, and Anand M Vijh, 1997, Do Long-Term Shareholders Benefit From Corporate Acquisitions?, Journal of Finance 52, 1765-1790. Martin, Kenneth J., 1996, The Method of Paym ent in Corporate Acquisitions, Investment Opportunities, and Management Ownership, Journal of Finance 51, 1227-1246. McConnell, John J., and Henri Servaes, 1995, Equity ownership a nd the two faces of debt, Journal of Financial Economics 39, 131-157. Megginson, William L., Angela Morgan, and Lance Nail, 2004, The determinants of positive long-term performance in strategic mergers: Corporate focus and cash, Journal of Banking & Finance 28, 523-552. Mehran, Hamid, 1992, Executive Incentive Plans, Corporat e Control, and Capital Structure, The Journal of Financial and Quantitative Analysis 27, 539-560. Morck, Randall, Andrei Shleifer, and Robe rt W Vishny, 1990, Do Managerial Objectives Drive Bad Acquisitions?, Journal of Finance 45, 31-48. Morck, Randall, Andrei Shleifer, and R obert W. Vishny, 1988, Management Ownership and market valuation: An empirical analysis, Journal of Financial Economics 20, 293-316. Myers, and Majluf, 1984, Corporate Financi ng and Investment Decisions when Firms Have Information that Investors Do Not Have, Journal of Financial Economics 13, 187-221. Myers, Stewart C., 1977, Determinants of Corporate Borrowing, Journal of Financial Economics 5, 147-175. Narayanan, M. P., 1985, Managerial In centives for Short-Term Results, Journal of Finance 40, 1469-1484. Nohel, Tom, and Vefa Tarhan, 1998, Share repurchases and firm performance: new evidence on the agency costs of free cash flow, Journal of Financial Economics 49, 187-222. Ofek, Eli, and David Yermack, 1997, Taking stock: Does equity-based compensation increase managers' ownership?, Working Paper 1-26. Opler, Tim, and Sheridan Titman, 1993, The Determinants of Leveraged Buyout Activity: Free Cash Flow vs Financial Distress Costs, The Journal of Finance 48, 1985-1999. Persons, John C., 1994, Signaling and Takeove r Deterrence with Stock Repurchases: Dutch Auctions Versus Fi xed Price Tender Offers, The Journal of Finance 49, 1373-1402.
122 Rau, P. Raghavendra, and Theo Vermaelen, 1998, Glamour, value and the postacquisition performance of acquiring firms, Journal of Financial Economics 49, 223-253. Rau, R. P., and Theo Vermaelen, 1998, Glamour, value and the post-acquisition performance of acquiring firms, Journal of Financial Economics 49, 223-253. Roll, Richard, 1986, The Hubris Hypothesis of Corporate Takeovers, The Journal of Business Vol. 59, pp. 197-216. Rosenstein, Stuart, and Wyatt, 1997, Inside Directors, Board Effectiveness and Shareholder Wealth. Rothschild, Michael, and Joseph Stiglitz, 1976, Equilibrium in Competitive Insurance Markets: An Essay on the Econom ics of Imperfect Information, Quarterly Journal of Economics 90, 629-649. Scholes, Myron, and Joseph T. Williams, 1977, Estimating Betas form Nonsynchronous Data, Journal of Financial Economics 5, 309-327. Servaes, Henri, 1991, Tobin's Q and the Gains from Takeovers, The Journal of Finance 46, 409-419. Servaes, Henri, 1991, Tobin's Q and the Gains from Takeovers, Journal of Finance 66, 409-419. Shleifer, Andrei, and Robert W. Vishny, 1988, Value maximization and the acquisition process, Journal of Economic Perspectives 2, 7-20. Sicherman, Neil W., and Richard H. Pettway, 1987, Acquisition of Di vested Assets and Shareholders' Wealth, The Journal of Finance 42, 1261-1273. Song, Ok-Rial, 2002, Hidden social costs of open market share repurchases, Journal of Corporation Law 27, 425-480. Stephens, Clifford P., and Michael S. We isbach, 1998, Actual Share Reacquisitions in Open-Market Repurchase Programs, The Journal of Finance 53, 313-334. Stulz, Rene M., 1988, Managerial control of voting rights: Financial policies and the market for corporate control, Journal of Financial Economics 20, 25-54. Thompson, Rex, 1995. Handbooks in Operations Research and Management Science: Finance (Elsevier, Amsterdam). Travlos, Nicholas G., 1987, Corporate Takeove r Bids, Method of Payment, and Bidding Firms' Stock Returns, Journal of Finance 42, 943-963. Vermaelen, Theo, 1981, Common stock re purchases and market signaling, Journal of Financial Economics 9, 139-183. Wansley, James W, William R. Lane, and Ho C. Yang, 1987, Gains to Bidder Firms in Cash and Securities Transactions, The Financial Review 22, 403-414. Weston, J. Fred, and Juan A. Siu, 2002, Changing Motives for Share Repurchases, Working Paper 1-57. Yermack, David, 1995, Do corporations aw ard CEO stock options effectively?, Journal of Financial Economics 39, 237-269. Yermack, David, 1997, Good Timing: CEO St ock Option Awards and Company News Announcements, The Journal of Finance 52, 449-476.
123 Bibliography Blasi, Joseph, Douglas Kruse, and Aaron Bernstein, 2003. In The Company of Owners: The Truth About Stock Options and W hy Every Employee Should Have Them (Basic Books, New York). Thompson, Rex, 1995, Empirical Methods of Ev ent Studies in Corporate Finance, Handbooks in Operations Research and Ma nagement Science 9, Finance, Chapter 29, 963-992.
125 Appendix A: T-Test The time series standard deviation met hod (t-test) calculate s a single variance estimate for the entire portfolio. Its drawback is that it does not take into consideration unequal return variance across securities. However, on the other hand, it does avoid potential problems of crosssectional correlation of security returns. The estimated variance of AARt is (AA R t AAR mean )2 = D-2 where D = estimation period and AA R t AARmean = D and the portfolio test statistic for day t is AA R t t = AAR and the test statistic for CAART1T2, assuming time-series independence is CAA R t ______ t = (T2-T1+1)1/2 AAR In order to calculate the z test statistic, it is a ssumed under the null hypothesis that Ajt has mean zero and variance 2Ajt. The maximum likelihood estimate of the variance is 1 (R mt -R m mean)2 S2 Ajt = S2 Aj 1 + Dj + (Rmt-Rmmean)2 where S2 Ajt = A2 jk / (Dj 2) Rmt is the observed return on the market index on day t, Rmmean is the mean market return over the estimation period and Dj is the number of non-missing trading day returns in the D-day interval. The standardiz ed abnormal return is defined as SARjt = Ajt/SAjt Under the null, each SARjt follows a students t distribution with Dj-2 degrees of freedom. Summing SARjt across the sample TSARt = SARjt and the expected variance of TSARt is
126 Qt = (Dj 2) / (Dj 4) and the test statistic for the null hypothesis that CARRT1T2 = 0 is ZT1,T2 = 1 / N1/2 Z T1T2, where ZT1,T2 = 1/ Q T1,T2 1/2 SARjt and QT1T2 = (T2 T1 + 1) (Dj 2) / (Dj 4)
About the Author Robin Wilber received a Bachelors Degree in Civil Engineering from the University of Florida in 1980 and a M.B.A. fr om the University of South Florida in 1986. She worked as an engineer and a corporate controller until she entered the Ph.D. program in Finance at the University of South Florida. Ms. Wilber began teaching many finance courses as a Ph.D. student and also served one year as an instructor at the Univ ersity of South Florida at St. Petersburg. Ms. Wilber is currently an Assistant Profe ssor of Finance at Niagara University.
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001680953
007 cr mnu|||uuuuu
008 051202s2005 flu sbm s000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001168
Wilber, Robin S.
Disentangling the repurchase announcement
h [electronic resource] :
b an event study analysis to the purpose of repurchase /
by Robin S. Wilber.
[Tampa, Fla.] :
University of South Florida,
Thesis (Ph.D.)--University of South Florida, 2005.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains 132 pages.
ABSTRACT: Researchers have consistently shown that a firms repurchase announcement is met with positive abnormal stock price return reactions. Open-market repurchases are extremely flexible, non-committal and non-punitive; thus, it is puzzling that the mere announcement of an open-market repurchase will likely increase a firms stock price. I propose to disentangle a firms choice to repurchase its stock to determine when a repurchase announcement is good news for shareholders and when the announcement is not. I find that the purpose of the repurchase announcement matters. At the announcement date, managers intention of avoiding dilution is significantly negative and enhancing shareholder value is significantly positive, as expected.
Adviser: Jianping Qi.
Method of payment.
x Business Administration
t USF Electronic Theses and Dissertations.