Stock Picking in Disguise?

New Evidence that Hedge Fund Activism Adds Value

Benjamin S. Solarz Economics

April 6, 2009 Advisor: Dean Takahashi

Abstract Using a large hand-collected data set spanning from 2000 through 2008, I find substantial evidence that activist hedge funds do more than pick stocks; they improve them. Controlling for differences in target firms, activist investments outperform the same hedge funds’ passive investments. Over two months, they earn 3.8% greater returns; over two years, they earn 18.4% greater returns; and measured by financial statements, they improve margins and ROA, increase payout and leverage, reduce balance sheet assets, and sell their targets more frequently and for higher prices. Hedge funds follow five cohesive activist strategies: they (1) improve business strategies, (2) advise on mergers and acquisitions, (3) demand that target firms sell themselves, (4) optimize capital structures, and (5) fix corporate governance. These results imply that skilled hedge funds can add significant value as corporate partners.

Introduction If you ask CEOs about their greatest fears, they might start talking about activist hedge funds. “If someone with a 203 area code calls you, call them back very carefully,” said public relations maven Joele Frank at a 2008 panel on activist hedge funds.1,2 Apparently, Greenwich, CT can be very dangerous. As the hedge funds located there and across the country redesign corporate governance, they have kicked up a storm of controversy. In the past few years, Nelson Peltz purchased Wendy’s, Carl Icahn pushed Yahoo into Microsoft’s arms, and William Ackman encouraged Target to divest its real estate. To top it all off, activists have even targeted other hedge funds, as when Carl Icahn sued over Steel Partners’ plans to go public.3 While there is no shortage of bold claims, hard evidence is tougher to come by. Do activist hedge funds truly threaten corporate health? To the contrary, using hand-collected data on 718 activist investments, I find that activist hedge funds add substantial value to their socalled victims—measured either by stock market returns or accounting performance. I improve on the existing literature, which fails to distinguish activist investing from astute stock picking, to show that hedge funds actually catalyze their targets’ outperformance. In particular, I test whether activism adds value beyond stock picking by comparing activist investments to the same hedge funds’ passive investments. In addition, I describe the most common activist strategies. Judging by short-term returns, investors applaud an activist hedge fund’s decision to target a company. Moreover, investors prefer these activist investments to passive investments. Over 61 days, controlling for the differences in target firms, activist investments outperform passive investments by 3.8%, and investments seeking board seats outperform other activist
1

While the term “activist” may draw to mind images of Gandhi, I use it to refer to a better dressed, better paid, and much more self-interested cohort: aggressive investors who attempt to influence their targets. Throughout this thesis, I will use the terms “activist” and “active” interchangeably. 2 “Discussing the Perils of Activist Investors,” The ew York Times [New York], 3 April 2008. 3 “Icahn and BofA Take Steel Partners to Court,” The ew York Times [New York], 7 February 2007.

-1-

I also find that these returns have declined over time. returns have declined over time. Further. and investments seeking board seats outperform other activist investments by 18. increase their margins by 2. stock prices jump significantly. divest unprofitable assets. increase their payout ratios by 2. activists follow five cohesive strategies: they (1) improve business strategies. (2) advise on mergers and acquisitions.6% more than passive targets—most likely by shedding unprofitable divisions. Except for activism seeking to fix corporate governance. but they also sell for higher prices. Again. and increase their leverage by 0. activist targets increase their return on assets (“ROA”) by 3. not only do activist targets sell themselves more frequently. Activist targets also reduce their assets by 4. except for corporate governance activism. Activism earns these excess returns by improving its targets—they increase profitability. and (5) fix corporate governance. investors react positively to activism even when it contains no stock picking signal.9% more. all strategies generate substantial long-term returns. investors reward all activist strategies. Each strategy targets different firms and improves them in distinct ways.5% more. and sell themselves more frequently and for higher prices. Controlling for the differences in target firm characteristics. activist targets outperform passive targets by 18.2%. In improving target firms. controlling for the differences in target firms. -2- . optimize their capital structures. When activists change their intentions from passive to active. Also consistent with short-term results.investments by 3. (4) optimize capital structures.0% more than passive targets.6%. In addition. Finally. except for corporate governance.7% more. Long-term returns also indicate that hedge fund activism enriches shareholders. (3) demand that target firms sell themselves.4%. While I find evidence that activism yields significant excess returns. every strategy earns significant excess returns. Over two years.

that they should watch over it with the same anxious vigilance with which the partners in a private copartnery frequently watch over their own. The same November 17. Cayne squeezed in 13 rounds of golf. Section VI describes the fundamental improvements in activist targets. 5 4 -3- . Turning to the analysis. it cannot well be expected.The new evidence presented in this paper suggests that. titled “Wall Street Bosses Spending Too Much Time on the Golf Course. Section I reviews the literature. both management teams and shareholders alike should welcome their insights. shareholders want management to invest as much as possible in projects whose internal rates of return exceed the firm’s weighted average cost of capital (“WACC”) (Koller et al. generally because their incentives differ.”4 In modern terms: CEOs often roll up their sleeves not for auditing spreadsheets but for dealing cards. Section II explains my methodology. Section VII concludes. Section I: Background For hundreds of years. Adam Smith warned. In August 2007. This paper is organized as follows. see Berle and Means (1932) and Jensen and Meckling (1976). Section IV shows that activists earn excess short-term returns. instead of avoiding hedge funds’ calls. activist hedge funds add significant value to their targets—judged by either stock prices or financial statements. In 1776. Section V presents similar evidence for long-term returns. Bear Stearns CEO Jimmy Cayne spent 10 days playing bridge out of state.” also reported that in June alone. CEOs have eroded the wealth of nations. far from economic hitmen. 2007 Wall Street Journal Article. “The directors of such [joint-stock] companies. while errant hedge funds brought his firm to its knees. it details the firms that activists target and the strategies that they use. In general.5 Agency costs arise whenever management (the agent) takes actions which harm shareholders (the principal). being the managers rather of other people’s money than of their own. Thus. 2005). however. Section III describes hedge fund activism. But CEOs paid based For a more modern treatment.

8 See Rock (1992). For the problems with options themselves. So while shareholders might want the firm to disgorge unproductive assets. shareholder activist Joseph Stilwell publicly wagered $25. -4- . CEOs paid based on ROA might forgo projects with attractive IRRs if those projects would drag down a firm’s overall returns. With their voting rights. it also increases management’s compensation. And even if managers wanted to maximize shareholder value. and job security. Black (1992) and Pound (1992) argue that shareholder activism could improve corporate performance. large shareholders might combat misguided management teams. indiscriminately purchasing assets. Parrino et al. SEC Schedule 13D. the literature strongly disagrees. 7 Predential Bancorp.6 Further. (2003) suggest that stockholders may simply prefer to walk away. But on the whole. they just might not know any better.000 that CEO Thomas Vento could not define return on equity on a per share basis. which offer their own perverse incentives. 6 For the problems with performance targets. Black (1998). Bainbridge (2006).on earnings growth might increase earnings by investing in projects with internal rates of return (“IRRs”) greater than zero but below the firm’s WACC. pride. Due to the free rider problem.8 Bebchuk (2008) discusses the costs that activists must endure. While doing so harms shareholders. managers only earn those options by meeting performance targets. see also Hall and Murphy (2003). the six-fold increase in large US firms’ market capitalization led directly to a six-fold increase in CEO pay. for example. 2007 annual meeting. activists’ personal benefits do not justify the necessary costs. At Prudential Bancorp’s February 9. While some herald stock options as an incentive alignment panacea. Filed 2 November 2008. management might harbor other intentions. Gabaix and Landier (2008) find that from 1980 through 2003. Similarly.7 For these reasons. The first problem is that too few activists play the game. and Kahan and Rock (2006). see Yermack (1995) and Bebchuk and Fried (2003). Vento failed. Jensen (1986) argues that managers may grow the company beyond its optimal size.

2007). not much. They enjoy many structural advantages that distinguish them from prior activists. and Walkling (1996). Malatesta. (5) they can take can take longterm. Wallis. activists sometimes induce improvements in corporate governance. drawing on prior legal scholarship. I elaborate on these points. see Wahal (1996). if ever. and Del Guercio. Institutions achieve the effects on firm performance that one might expect from this level of effort—namely. hedge funds are different. In particular. and Woidtke (2008). hedge funds have the incentives to engage in activism—(1) they earn pay-for-performance bonuses. For other critiques.” this thesis argues that hedge funds succeed where others have failed. induce long-term changes in performance. especially for public pension funds. such as those mutual funds face. Gillan and Starks (2007). Nelson. and (3) they can avoid ERISA regulation—and they have the weapons to engage in activism—(4) they can make concentrated bets. For other failures. see Lipton and Rosenblum (1991). their wooden arrows make little dent in the corporate fortress. Carleton. and “block purchasers. Romano (2001).9 The empirical record agrees. see Bebchuck (2005.10 In a literature review. and Weisbach (1998). “A small number of American institutional investors spend a trivial amount of money on overt activism efforts. and Del Gurcio. Malatesta. including diversification requirements and insider trading regulations. Johnson and Shackell (1997). (2) they suffer fewer conflicts of interest. Black (1998) concludes. See Black (1990) and Romano (1993).” While the early literature focuses on mutual funds. and (6) they can use derivatives. regulatory constraints. I argue why activist hedge funds might outperform their activist peers.Even when activists do take up arms. and political constraints. Gillan and Starks (2000). First. and Woidtke (2008). 10 For corporate governance successes. Karpoff. Wallis. (A) Hedge Funds Are Different As many a pitchbook eagerly points out. I document empirical evidence that activist hedge funds do improve corporate value. Second. Karpoff (2001). and Walkling (1996). pension funds. illiquid positions. Barber (2007). Below. Black (1998). Davis and Kim (2008). Karpoff. but they rarely. 9 -5- . Most institutional investors suffer conflicts of interest.

et.First. mutual fund managers might placate a company to win its pension business (Davis and Kim 2007). Fifth. 12 See Coffee (1991). In contrast. (2007) and Hu and Black (2007). 13 See Christoffersen. mutual funds cannot hold more than 10% of any company’s stock or concentrate more than 5% of their own portfolio in a given name (Klein and Zur 2008). Indeed. hedge funds can take illiquid positions to weather a long fight.11 Second. 11 -6- . al.13 Or hedge funds Golec (1992) and Deli (2002) find that only 6-7% of mutual fund companies offer their managers performance bonuses. In contrast. Fourth. mutual fund managers rarely earn performance bonuses. hedge funds earn pay-for-performance bonuses. hedge funds suffer fewer conflicts of interest than other institutional investors. restrained by subchapter M of the Internal Revenue Code and the 1940 Investment Company Act. which restrict the risks other institutional investors can stomach. often as a negotiating ploy. Third. mutual funds must maintain constant liquidity to satisfy redemption requests (Klein and Zur 2008). Whereas hedge funds sign investors into lengthy lock-up agreements. and public pension fund managers might flatter a CEO to win contributions for a re-election campaign (Kahan and Rock 2006). hedge funds might acquire voting rights without taking on economic exposure. For example. hedge funds can make concentrated bets to fight entrenched management teams. Bhide (1993). activists might threaten to purchase the entire target company. Kahan and Rock (2006) report that 97% of mutual funds’ annual compensation ignores investment gains. And finally. Through stock lending and derivatives. hedge fund managers often earn up to 20% or more of their firm’s annual return (Brav et al. and Aragon (2007). Similarly. which incentivize them to wage successful activist campaigns. hedge funds can avoid ERISA or “prudent man” regulations.12 The threat of liability may scare these potential activists from taking the risk. hedge funds use exotic strategies. Hedge funds do not face the 1940 Investment Company Act’s pay restrictions. 2008).

might use these derivatives expressly to gain economic exposure. Boyson and Moradian (2008). mutual funds must follow the Investment Company Act’s short-selling and leveraging restrictions (Klein and Zur 2008). find that the Hermes U. To date.14 Using a case study approach. Brav et al. They find strong evidence for short-term abnormal returns and mild evidence for long-term improvement. these findings might simply mark great passive investing—choosing companies primed to improve beyond the market’s expectations. To mitigate this problem. 14 -7- . (2008). for instance. First. Several recent papers focus on hedge fund activism. does not mean it actually used activism to earn those returns. positive excess returns. and Klein and Zur (2008) study large crosssections of hedge fund activism. (B) Similar Studies Do ot Distinguish Activism From Stock Picking While activists proclaim their successes in myriad press releases. Greenwood and Schor (2008). they provide no evidence that activists have done anything more than simply pick undervalued companies poised to outperform expectations. nonetheless. and Klein and Zur (2008) study hedge fund activism. they justify its requisite costs. Clifford fails to control for Becht et al. But none of these studies distinguishes activism from passive investing in disguise.K. In contrast. And simply because an activist target improves its margins does not mean the activist did anything to catalyze the change. (2008). But this study offers a flawed dichotomy. to test whether activism adds value beyond stock picking. we must compare activist investments to passive investments. Simply because an activist earns abnormal returns. Clifford (2008) compares activist investments to passive investments. Rather than activism. Puustinen (2007). Puustinen (2007). Thus. the literature continues to compare apples to pairs—hedge funds’ activist investments to a non-hedge fund benchmark. (2006). Becht et al. More similar to this study. Focus Fund’s activism earns large. Brav et al. by increasing activism’s potential gains. it is not so easy to assess their claims. Boyson and Mooradian (2008). Clifford (2008).

In addition to ignoring stock picking. Second. Well-known passive investors. Further. Most importantly. First. -8- . like improving corporate governance. But these filings differ across other crucial dimensions. Clifford also defines activism too broadly. an investor might file a Schedule 13D to convey information to other investors. they disagree that activists improve their targets. While certain activist strategies. In contrast. might benefit from leaking their best ideas to the public. might generate large returns. astute investors might prefer to file Schedule 13Ds to protect themselves from potential litigation.15 This paper eliminates these problems. might simply represent the market preferring fresh ideas to pickled ones. activism might outperform simply by targeting more attractive firms. he calls all Schedule 13D filings activist investments.1% simply hold the stock for “investment purposes. Greenwood and Schor (2008) present some evidence on heterogeneity. others. The larger reaction around Schedule 13D filings. 48. In particular. prior studies barely discus heterogeneity within activism. which he calls active. like selling the target firm. these strategies might affect companies in conflicting ways that cancel each other out. In my sample of 942 Schedule 13Ds. Second. to Schedule 13G filings. And third. for example.” Instead of capturing the active-passive distinction. may not. hedge funds must file Schedule 13Ds within 10 days of building their position but need not file Schedule 13Gs for an entire year. Clifford’s study might measure some other confounding variable. which he calls passive. I analyze activists’ strategies in detail. even passive investors holding at least 20% in a company must file a Schedule 13D. In particular. For example. therefore. even though many explicitly renounce activist intentions. business strategy activism might reduce leverage.differences in target firms. instead. These legally savvy investors might also have the best ideas. while capital structure activism might increase leverage. Clifford compares Schedule 13D filings. they propose that activists simply put firms 15 There are several possible confounds. previous studies improperly conclude that fundamentals change very little. Without doing so. Sweren (2008) documents abnormal returns when hedge funds leak their best ideas. for instance. Without testing for such heterogeneity. leaving little aggregate effect. so Clifford’s results might reflect investors’ preference for these concentrated bets.

But this includes inexperienced hedge funds. In summary. Expanding the activist sample to include these novices resembles defining an athlete as anyone who has ever -9- . Greenwood and Schor (2008) argue that hedge funds’ poor performance during the 2008 Credit Crunch proves that they rely on selling their targets. might explain the results. I improve on the literature in several ways. I correct for this bias. But by asymmetrically removing successful sales. and not activists’ reliance on selling their targets. (3) control for differences in target firms. in contrast to the existing literature. Prior scholars call “activist” any hedge fund which has ever filed an activist Schedule 13D. This structural pressure. which bring down the averages. First. Section II: Experimental Design (A) Contributions to the Literature To argue that activism adds value. biasing their results downward.in play. and (2) it largely ignores heterogeneity. (2) compare activist Schedule 13D investments to passive Schedule 13D investments. (4) measure short-term abnormal returns around events which occur after the hedge fund has disclosed its position. (1) it fails to distinguish this outperformance from stock picking. Additionally. activist hedge funds earn negligible long-term excess returns. and not failed sales. levered institutions like hedge funds unwound dramatically. In Section V. This section elaborates on these frontiers. This is the first paper to: (1) use a comprehensive sample of experienced hedge funds. I study hedge funds that repeatedly pursue activism. But this overlooks the crisis’s main symptom. and (6) fully document cross-sectional variation in activists’ strategies. while the literature documents some activist outperformance. The authors show that. (5) analyze board seats. excluding sales. Grenwood and Schor have imprecisely measured activists’ non-sale returns.

Fifth. the first time it expresses activist intentions. While Clifford (2008) also compares activist returns to passive returns. I expand on this method below. By selecting only hedge funds known for activism. Third. launching proxy fights. to disentangle activism from stock picking ability. after rounding up two NY Yankees and five hundred physical education teachers. Second. Because 80% of all investments using an instrument seek board seats. I resolve this bias.owned a baseball glove. or suing the target. I improve his methodology by (A) limiting my focus to Schedule 13Ds and (B) recognizing that many Schedule 13Ds espouse passive intentions. I expand on this method below. providing private financing. I consider five instruments: seeking board seats. I consider them most thoroughly. prior scholars conclude that no one can consistently throw a 90 mph fastball. By following each investment’s SEC trail. I distinguish activist instruments. this method more cleanly isolates the activist signal from the stock picking signal. I control for variation in target firm characteristics like undervaluation and underperformance. . I document short-term abnormal returns around events that take place after the initial Schedule 13D filing. I compare hedge funds’ active Schedule 13D investments to the same hedge funds’ passive Schedule 13D investments. like selling the firm. offering to buy the target. Board seats distinguish activism from stock picking by providing tangible proof that the hedge fund seeks to influence its target. from activist strategies. like seeking board seats. I also expand on his study by more thoroughly examining changes in target firms’ accounting performance. I find subsequent filings in which a hedge fund changes its intentions—for example. Because the hedge fund has already disclosed a passive interest in the target firm. an activist uses instruments to implement its strategies.10 - . This improves on Clifford’s (2008) study by dismissing the alternative hypothesis that activists target more attractive firms. Fourth.

First. a second layer of stock picking to peel back. And second. perhaps because different strategies cancel each other out. While Puustinen (2007) and Brav et al. To eliminate this bias. On the other hand. we must demonstrate that (1) shareholders enjoy greater returns than they otherwise would and (2) target firms improve their accounting performance more than they otherwise would. I add evidence that strategies differ over the long-run.11 - . to the extent activist investments outperform passive investments. these findings cannot stand alone. if activist targets became better companies without earning excess returns. Both activist and passive investments benefit from hedge funds’ superior stock picking abilities. prior studies have observed little change in target firms’ fundamental accounting performance. . capital structure. It is probably a necessary condition. By comparing activist targets to passive targets. M&A. 16 Thus. Because activism costs more than passive investing. Thus. too. only by understanding their strategies can we replicate activist hedge funds’ ability to improve firms. however. (2008) analyze short-term returns by strategy. and corporate governance.Sixth. (B) Testing For Activism To prove that hedge fund activism adds value. activism must add value. But unlike the Beatles. Hedge funds might pick more attractive firms for activist investments than for passive investments.16 There is. This is important for two reasons. activist outperformance is a sufficient condition to show that activism adds value. sale. then the market must have already expected those improvements—and the activist probably did not cause the change. then hedge fund activism would simply be a penname for astute stock picking. but only activist targets benefit from hedge funds’ activist abilities. I come closer to measuring how these activist targets otherwise would perform than does any prior paper. If activism earned excess returns without improving target firms. I identify five cohesive activist strategies: business strategy. hedge funds should demand greater returns. I also investigate heterogeneity in fundamental improvements.

com/. that is. if we observe that board seat investments outperform other activist investments. the list suffers from little if any bias in describing the historical record.I control for differences in target firms—in particular. Boyson and Mooradian (2008). rather than assembling a random sample of a broad universe. I find no evidence of survivorship bias 17 Puustinen (2007). Ken and Robyn deserve full credit. I consider the 48 hedge funds which the shareholder activism advisory firm 13D Monitor defines as most active. Brav et al. Thus. . To the extent I had time to emerge from my dorm room at any point during the month of January. If activist investments still outperform passive investments. and Klein and Zur (2008) all use news searches to form an initial hedge fund list. 17 In particular. Clifford (2008). (2008) limit their sample to hedge funds with at least two activist investments.13dmonitor. seeking board seats helps activists influence the firm’s decisions. my sample generalizes not to every ex-Goldman Sachs trader who files a Schedule 13D. (2008). Because it is comprehensive. (C) Collecting Data Unlike previous studies. I analyze board seats to better distinguish activist investments from passive investments. whereas a fine line separates activism from passive investing.12 - . Brav et al. board seats must add value. I study only hedge funds which focus on activism. differences in undervaluation and in underperformance. As before. At worst. a much thicker line separates board seat investments from other investments. I am completely indebted to Ken Squire and Robyn Rosenblatt for allowing me access to this database and for putting up with my annoying questions. which include hedge funds that have made very few activist investments. the hedge fund has opened a dialogue with management. Thus. they compile a comprehensive list that fits a narrow criteria. 18 13D Monitor sought to include every fund that repeatedly uses activism. the activist might get its way. we can conclude that activism adds value. Even if it never wins a seat. Unlike the mere rhetoric accompanying many filings. Additionally. Thus. 18 13D Monitor maintains a subscription-based database of Schedule 13D filings at http://www. controlling for target firm variation. even hedge funds with only one prior activist investment enter the sample. At best. the activist might threaten a proxy fight as a bargaining chip. but rather to experienced activists.

First. filers must declare why they acquired the shares. proxy fights. M&A.9% exhibit negative average activist abnormal returns. even if the sample were biased towards hedge funds whose activist investments perform quite well. of which I classified 393 as active and 325 as passive. 19 .19 As evidence against backfill bias. Of the 393 activist investments. to focus on the current activist climate. Section 13(d) of the 1934 Exchange Act requires investors to file with the SEC when they amass large stakes with activist intentions. 213 sought board seats. I collected information on all 942 investments that these hedge funds made. From these 942 investments. I collected the entire activist history for each investment. several hedge funds in my sample performed quite poorly. hedge funds must file a Schedule 13D within 10 days of acquiring at least 5% in any public company. To further dissect these investments. I narrowed the sample. Without exception. or lawsuits—or any activist strategy—business strategy. 21 While activists rarely pierce the 5% threshold in large firms.21 In Item 4. this sample paints an accurate picture. Using 13D Monitor’s online database. capital structure. buyout offers. I excluded filings before January 1. it would not necessarily be biased towards hedge funds whose activist investments outperform their passive investments. I turned to the Edgar database. I excluded all firms which declared bankruptcy before the hedge fund invested. private financings. I classify an investment as activist if the investor announces any activist instrument—board seats.13 - . This narrowed my sample to 718 investments. I compared my sample to two sources: Puustinen (2007)’s sample and hedge fund consulting firm Hedge Fund Solutions’s database. Thus. 2000. Second. or corporate governance. sale. 20 Of the 46 firms with long-term return data.000 individual SEC filings. 13D Monitor’s sample provides a fitting laboratory. these hedge funds may be the biggest culprits.when comparing my sample to those in similar papers. the most frequent funds in each source overlapped with my sample. By reading over 6. In particular. In fact. activists tend to focus on targets small enough that they can establish large positions. if activism does simply represent disguised passive investing.20 And finally. 23. which houses SEC filings dating back to 1994.

bid up its price. I find almost no such instances of 20%+ passive positions. Sweren (2008) documents excess returns after hedge funds file Schedule 13Fs. of a company’s stock may file a Schedule 13G with the SEC..4% of passive Schedule 13Ds. which implies that these funds do differentiate passive from activist investments.22 But sometimes these firms will instead file a Schedule 13D. And finally. but less than 20%.24 After collecting information on these investments. So why might a hedge fund file a Schedule 13D rather than a Schedule 13G? First. Second. While this might bias my Schedule 13D sample towards large passive firms. using the more rigorous Schedule 13D safeguards the hedge fund against liability.g.14 - . 24 For example. e. . I calculate the target firm’s return in excess of 22 Passive investors acquiring at least 20%.” Should we believe them? First. I found no evidence that the hedge funds engaged any attempts to change the target firms. the SEC and target firms use the courts to enforce truthful disclosures. Passive institutional investors who acquire at least 5%.23 Second. each hedge fund in my sample announces either passive or activist intentions at least once. They may wish to communicate their investment thesis so that other investors. Montgomery Medical Ventures. hedge funds subsequently announce activist intentions. in searching Bloomberg news for certain randomly chosen passive investments. SEC vs. Clifford (2008) argues that hedge funds exercise extreme caution to avoid filing any misleading disclosures. Fourth. This devotion to updating their goals indicates that hedge funds take care to convey honest information. The strongest evidence for this proposition comes from the hedge funds themselves: many filings announce that they are holding the stock for “investment purposes only. Third. LP (1996) and Ronson Corp v.Because this is the first paper to suggest that many Schedule 13D filings might harbor passive intentions. funds might take advantage of Schedule 13D’s immediate disclosure. 23 See. I find that after 32. Steel Partners II (2005). I pulled share price histories from the Center for Research in Security Prices (CRSP). citing conversations with hedge fund managers. however. upon realizing the firm’s true value. some elaboration is in order. must file a Schedule 13D.

25 . where the firm-specific component (denoted abnormal) is the difference between the firm’s value and the median value of its industry peers. their activist efforts target less profitable firms than do their passive Tobin’s q equals the firm’s market value divided by its replacement value—measured as the market value of debt plus equity divided by the book value of debt plus equity.the market model measured against the CRSP value-weighted NYSE/Amex/Nasdaq index. A low Tobin’s q implies undervaluation. I use two types of data on target fundamentals. I could only obtain Compustat data dating back to January 1. I calculate these variables by hand using firm-level data from Compustat. First. The analysis excludes firms for which CRSP lacks data—most likely small. Unfortunately. and abnormal leverage. I calculate the firm-specific component (denoted abnormal q) as the percentage difference between the firm’s Tobin’s q and the median of its industry peers.15 - . First. While unreported. with low Tobin’s q. abnormal return on equity (“ROE”). Both Bloomberg and Compustat were missing data for several firms. I use accounting data pulled from Bloomberg. In general. 2004. presumably smaller firms. following Ang and Chen (2006) and Puustinen (2007). This might bias my findings. I use abnormal q to proxy for undervaluation. I describe the firms which activist hedge funds target. using the GICS industry classification. Second. all results are robust to the use of the Fama French factors.25 While hedge funds target relatively profitable firms overall. they target undervalued firms. For example. this section elaborates on the differences between activism and passive investing. I describe the event study method more fully in Appendix A. unlisted firms. I calculate industry-adjusted statistics to better analyze a target firm’s idiosyncratic performance. I also calculate abnormal margins. Section III: Hedge Fund Activism Before turning to the results.

*.4 1.3 12. Column 2 compares activist investments seeking board seats to other activist investments.4 0. Table D-1 adds more information. Column 1 shows that hedge funds go activist in undervalued and underperforming firms.3 1.4 1.6% more undervalued. active investments seeking board seats.9 8.1 8.3 -2. Column 1 compares activist investments to passive investments.2 24.0 -5.1 13.0 1.0 -1.1 -6.5 *** *** *** *** *** *** ** * efforts. sale.6 10. and corporate governance.7 1.5 -0.8 14. I discuss heterogeneity among activists’ strategies.9 * * *** * ** * ** Board 636.1 146 Diff -255.4 -1. the average activist target is 16.1 9.0 -1.7 -11. In Appendix D. and 1% levels.1 3.6 3. activist targets are 29. Relative to passive targets. All variables refer to the trailing twelve months prior to the Schedule 13D.0 0.16 - . Second.0 16. we should first describe the ex-ante differences in the two samples.7 12.5 492 Diff 102. (1) Activism Activist Market Cap Tobin's q Abnormal q EBITDA Margin Abnormal Margin Return on Assets Return on Equity Abnormal ROE Leverage Abnormal Leverage Payout Ratio Cash/Assets N= 757.8 11.0 0.5 5.2 9. and *** refer to statistical significance at the 10%. Measured by Abnormal q. they go active in the .5 20.4 -0.7 12. Table I reports summary statistics for active investments.5 23. capital structure. I describe five main categories: business strategy.2 19.4 22.6 9. 5%.0 9.7 -5.3 2.7 0. and active investments not seeking board seats. (A) Activists Target Undervalued Firms Before we can measure activist investments against a passive benchmark.7 -0.2 -1. M&A. All data is winsorized at the 10% level.7 7.2 15.5 -22.0 8.7 22.1 -0.1 1.3 303 Passive 654.0 1.8 10.68).Table I Target Firm Characteristics The table reports the characteristics of hedge funds' targets.6 157 (2) Board Seats Other 891.4 17.0 -4. **.9 3.0 -2.1 -29.1 16.7 -0.6 -16. relative to the median firm in its industry. passive investments.4 22.1 -4.0% more undervalued (t = -3. including definitions. While hedge funds target overall profitable firms.3 16.7 17.2 -11.

1% lower ROA (t = 2. Relative to passive targets. and Klein and Zur (2008).59).34). .2 lower Tobin’s q (t = 2. and 6.” he includes some passive investments.1% lower margins (t = 1. with 5.9% more cash per dollar in assets (t = 1. The average activist target has 3.28 This might be explained by his methodology. Column 2 shows that within activist investments. 1.7% higher ROE than the median firm in its industry.least profitable ones. From the hedge funds’ SEC filings. I categorized the investments into five broad strategies: (1) “Business Strategy” activism. While these results resemble most prior studies.42). I calculate ROE as EBITDA divided by the book value of equity.49). 28 See Boyson and Mooradian (2008). 27 I calculate payout ratio as common dividends divided by net income.27 Overall. (B) Activists Use Different Strategies Activists tailor their strategy to each target. (2008). This evidence refutes the alternate hypothesis that activists seek board seats to associate themselves with successful firms. however.17 - . those seeking board seats are even more undervalued.49) and 0. Compared to other activist investments.4% lower ROE (t = 1. Consistent with the literature. (2) “M&A” activism.0% higher margins and 0.7% lower payout ratios (t = 0. which seeks to maximize value through extraordinary transactions. I calculate ROA as EBITDA divided by book value of assets. with 0. like acquisitions or spinoffs. these passive investments might lift the observed profitability. Clifford (2008) finds that active targets are more profitable than passive targets. Relative to passive targets.49). 2.13). they have 1. 26 I calculate margins as EBITDA divided by sales.71). and 2. Brav et al. which attempts to improve the firm’s competitive strategy and capital allocation.26 Activist targets also hoard assets. it seems they seek board seats to improve underperforming firms. they have 1.47). hedge funds seek board seats in the most troubled companies. They are also less profitable. the table provides moderate evidence that hedge funds try to unlock value in underperforming target firms. Instead.0% lower ROA (t = 1.0% lower ROE (t = 2. because he considers all Schedule 13Ds to be “active.1% lower margins (t = 4.

This way.24 *** 2.14 0. and (5) “Corporate Governance” activism.78 -0. The dependent variable is a dummy equal to one if the hedge fund targets the firm with a particular activist strategy. and *** represent statistical significance at the 10%. which tries to influence leverage or payout policies.72 -0.14 0.22).61 -1.02 -0.21 *** -5.13 0. This supports the hypothesis that business strategy and sale activism seek mutually exclusive goals—where one builds long-term businesses.02 -3.43 ** 1.43 -2.60 *** -6.75 547 0. 5%.34 1. and Pseudo-R 2 0.07 0.33) and lower margins (t = -1.41 ** 2.62 0.66 -4.25 547 * 1. -0.79 * 1.00 0. which addresses the firm’s governance policies. Market Cap Tobin's q Growth Margin Leverage Cash/Assets Constant No. While correlations among the strategies vary from 15% to 37%.06 9.76 *** 547 t -stat -1.07 1. at 7%.19 * 1.33 1.13).20 0. Bus Strat t -stat Coeff.18 - . In my sample.78 -2. Table II shows a logit regression that uses firm characteristics to predict the hedge fund’s strategy. *.31 0.89 0.04 Sale Coeff. the table reports the liklihood that activists target firms with various characteristics. active or passive.43 0. business strategy and sale activism correlate the least. I elaborate on these categories by excerpting from SEC filings.61 * -1.30 1. Table II Logit Analysis By Strategy Using a logit regression. the other quickly sells them.17 1.06 *** M&A t -stat Coeff. If activist strategies truly differ. we should expect each strategy to target firms with particular weaknesses.06 -0. which demands that the firm sell itself.03 Cap Struc t -stat Coeff.05 0.13 10. I distinguish between activism which improves companies through capital allocation and activism which improves companies through corporate transactions.85 -0.11 -0. All data is winsorized at the 10% level. I improve on Brav et al. Business strategy activism targets poorly performing companies with lower sales growth (t = -1.22 0.04 -0.41 0.57 -2.51 -1. 0. **.42 * -1.35 1. Targets are also larger (t = 3. and 1% levels.07 -0. 0.78 -15.76 -0.14 547 3.04 ** -1.78 0.29 *** -2.43 -2.71 0.03 -56. (4) “Capital Structure” activism.50 -0.39 0.46 547 0.06 Corp Gov t -stat Coeff.51 -0.04 *** 2. 0.42 1. hedge funds most frequently use M&A and Sale activism (130 times each) and least frequently use business strategy activism (75 times).33 -1.06 *** -5.67 *** -6.03 * -21.21 4. (2008) by separating what they call “business strategy” into business strategy and M&A. The sample population includes all hedge fund investments.(3) “Sale” activism.22 0. In Appendix B.19 -1.01 .

and *** indicate statistical significance at the 10%.51).29 Corporate governance activism targets firms with excess cash (t = 1.01 1. They also have low existing leverage (t = -1.9 1.59)—probably to nurse businesses back to profitability—while M&A activism nominates directors least frequently (z = 0.02 perhaps because large companies struggle to allocate capital among their many divisions.2 1.29 0.60 0. perhaps because they are easier to sell in whole.61) with lots of cash on hand (t = 2.89 2.7 * 316 z -stat 2. . *. business strategy activism wins board seats most 29 I measure leverage as long-term debt divided by total assets. 5%. M&A and sale activism target undervalued companies (t = -0.92 0. Similarly.33 -0. perhaps because larger firms have underperforming divisions rife for spinoffs. which can increase their payout.Table III Board Seats By Strategy The table present the logistic regression odds ratios that each strategy (1) seeks board seats and (2) wins board seats.06).04). (1) Board Logit Odds Bus Strategy M&A Advice Sale Cap Structure Corp Gov No.1 *** 1. and Pseudo-R 2 2.78).1 0. The dependent variable is a dummy set equal to 1 if the activist sought or won a board seat.04 Odds 1.99 0.43). Sale activism targets smaller companies (t = -1.5 1.89)—probably because the hedge funds anticipate exiting after quick flips. M&A activism targets large companies (t = 2. and 1% levels. Column 1 and Column 2 present odds ratios from logistic regressions that predict which strategies seek board seats and which strategies win board seats. Targets enjoy very high margins (t = 2. Business strategy activism nominates directors most frequently (z = 2.40 1.59 0.89 and t = -1. Table III describes the frequency with which each strategy seeks board seats. likely following Jensen’s (1986) hypothesis that misaligned managers hoard assets.3 155 (2) Win Logit z -stat 0.61 0.3 1.41).19 - . Capital structure activism targets healthy firms.3 1. **.9 ** 1.

they drive them back down. In this section. Investors may believe that hedge funds can access better information or conduct better analysis without adding any activist value. while M&A and sale activism succeed least often (z = 0. I use a dummy variable regression to compare activist filings to . I now turn to the first research question. I argue that activism does add value beyond simple stock picking. once shareholders discover hedge funds’ intentions. however. shareholders vote with their wallets. we should observe immediate excess returns. these price swings presage actual changes in firms’ values. to short-term speculators. Second. not the moral hazard problem—that is. activists target undervalued firms with room to improve. I compare activist filings to passive filings.29). the market favors activist investments—and in particular. like hedge funds seeking to sell the business. Having described activists’ methods. In summary.often (z = 0. Particularly in efficient markets. the filings may signal target firms’ already good prospects without actually signaling that hedge funds will in fact improve those prospects. which convey no explicit activist intentions. targets prefer long-term partners. in the short-term. Thus. First. these short-term returns do not themselves prove that activism adds value. Sweren (2008) documents short-term abnormal returns around hedge fund Schedule 13F filings. Depending on the strategy they want to use. Does this activism increase returns? Section IV: Short-Term Stock Returns Everyday in the stock market. They may reflect hedge funds solving the adverse selection problem. and with bad news.20 - . activist investments which seek board seats. with good news.92). they choose different targets. if activism adds value. This makes sense. I find strong evidence that. like hedge funds advising on business strategy. In fact. they bid up prices. Absent efficient markets.33 and z = -0. using three tests.

7% *** 67.+2) 2. Panel A compares activist investments to passive investments. I investigate when activists add the most value—what target firm characteristics.8% *** 12. I also investigate heterogeneity across these returns.6% *** 342 (-2.1% *** 5. Panel B compares active investments seeking board seats to active investments that do not. activist instruments.6% *** 4.9% *** 2. I find evidence that the market favors investments which take active steps.8% *** 45 No Board Seats (-30.+30) 10. Passive Active (-30.7% *** 80. or activist strategies does the market judge most effective? Undervalued. and *** indicate statistical significance at the 10%.9% *** 45 (-2. and 1% levels. I estimate parameters over the 200-day window prior to the Schedule 13D filing. like seeking board seats.7% *** 79.1% * (-2.5% *** 3.4% *** 5.9% *** 149 (-2. Panel A: Active vs. Finally. not simply to stock picking.7% * 5. I identify events which isolate the activist signal from the passive investing signal. 5%.+30) 1.+2) 7.5% *** 149 Difference (-30.21 - .7% * 2. all strategies enjoy abnormal returns. *.6% *** 77.+30) 3. **.+30) Median Average % Positive N= 12. I report p -values for medians using the Wilcoxon Signed Rank Test.+2) 4.0% *** Panel B: Board Seats vs.5% *** 17. the stock market responds favorably to activism.4% *** 340 Difference (-30. (A) Activism Exceeds Stock Picking Table IV reports the market’s reaction to activism.9% * .+2) 1.+30) Median Average % Positive N= 11.9% *** 80. In Appendix D. activist investments outperform—particularly when the hedge fund seeks board seats. I calculate excess returns over 61. I note that these returns have declined over time. In particular.+30) 8.3% *** (-2. Third. controlling for target firm undervaluation and underperformance.6% *** 57. Other Active Investments Board Seats (-30.4% *** 194 (-2. Except for corporate governance activism.+2) 3.8% *** 8. underperforming target firms experience the largest excess returns. Again. Thus. I find significant excess returns.7% *** 88.passive filings. Table D-2 shows more details.and 5-day windows relative to the market model Table IV Short-Term Returns by Investment Type The table presents average daily returns around the schedule 13D filing in excess of the market model using the CRSP value-weighted index.9% *** 194 Passive (-30.2% *** 8.8% *** 12.+2) 4.6% *** 77.

The results are consistent with Boyson and Mooradian (2008). To test whether activism differs from stock picking.9% and 4. On filing day and the following day.In the Short-Run.22 - 13 D 20 5 D D ay s ay s ay s ay s ay s .37). active positions jump by 4.4%. For active (blue) and passive (red) hedge fund investments. 80.6% over 30 days. The dashed lines divide active positions into those in which the hedge fund sought board seats (long dashes) and those in which the hedge fund did not (short dashes). Over 61 days. Brav et al.91 and z = 3. Panel A in Table IV compares activist filings to passive filings. the abnormal returns trend up to 13. activism outperforms by 8. active positions jump by 4. Afterward. and Klein and Zur (2008). Figure 1 plots the average excess return.4% of active positions and 57. Active Investments Outperform Passive Over a 61-Day Event Window 20% Board 15% Active NoBoard 10% 5% Passive 0% -5% Fi lin g D ay s ay s ay s D 25 30 ay s ay s ay s D D ay s D D D D D -5 10 15 D -3 0 -2 5 -2 0 -1 0 -1 5 Figure 1. Over 61 days. the chart plots the average excess return around the schedule 13D filing. Both active and passive investments experience positive and significant excess returns surrounding the filing date.3% (t = 3.5% and passive positions by 1. from 30 days prior to the Schedule 13D filing date to 30 days afterward. Clifford (2008). using the CRSP value-weighted index. (2008). Before the filing. This .1%.96). from 30 days prior to the filing until 30 days afterward. between 10 days to 1 day prior.6% of passive positions show positive returns (z = 8.6% and passive positions by 2.

controlling for the target firm’s undervaluation and underperformance. The regression uses Tobin’s q to proxy for undervaluation and margins to 30 In Appendix A. . Overall.0% Active Passive Board No Board Figure 2.In the Short-Run. and no board investments.0% 2. I regresses the 5-day abnormal return on activism and board seat dummy variables. In other words. Activist targets might differ from passive targets in a way that makes them more likely to outperform.0% 0. Active Investments Outperform Passive Controlling For Undervaluation and Underperformance 10. Within active investments. I elaborate on these dummy variable regressions. far from a controlled experiment. those seeking board seats earn significantly higher returns. controlling for the possibility that activist investments might be more attractive ex-ante.30 Figure 2 shows the excess returns to an average firm which a hedge fund targets for an active. board.0% 8. these comparisons do not prove that activism adds value. Over 61-days. passive.” I use a dummy variable regression to compare activist to passive investments. passive.1% (t = 1. it seems activism add value.33). measured by a dummy variable regression. supports Clifford (2008)’s results. these investments outperform by 5.0% 4. hedge funds might simply call their best ideas “activist.0% 6. board. controlling for Tobin's q and margins.23 - . Still. The chart shows the average 5-day returns to active. or no board investment.

4% *** 77. I calculate abnormal returns around three events: (1) having previously filed a passive 13D.9% *** 60. I also conducted basic news searches.+2) 1. this passive signal is of a small magnitude. it might convey some additional passive signal. and 1% levels. 31 32 In Appendix D.32 In addition. .2% *** 66. If a hedge fund increased its exposure in any subsequent filing. **. however. the first time a hedge fund wins board seats.3% *** 73.6% *** 53 (-2.2% 109 proxy for underperformance. I report p -values for medians using the Wilcoxon Signed Rank Test. (1) First Action (-30. Thus. Further. I read each investment’s entire SEC trail. 5%.4% *** 66. and (3) having already announced its intention to seek board seats. the only new information that these subsequent filings convey is the activist’s intention to improve the target firms. Because the activist has already disclosed a 5% stake in the target firm.2% *** 4.5% *** 3. and board investments outperform other activist investments by 3. *. I isolate the activist signal in subsequent filings. (2) having previously filed an activist 13D.71).+2) 0.2% *** 6.1% *** 48 (3) First Board Win (-30.31 The results are robust to other models. because activist hedge funds cannot fully control whether they win board seats after they decide to seek them. Activist investments outperform passive investments by 3.24 - .3% 0.8% * 53. But given that the hedge fund has already accumulated 5% of the company. I estimate parameters over the 200-day window prior to the event.5% *** 8.0% *** 53 (2) First Board Seat (-30.8% (t = 4. the first time a hedge fund seeks board seats.78).+2) 2.Table V Abnormal Returns After 13D Filing Date The table presents average returns around events that occur after the initial Schedule 13D filing.4% *** 110 (-2. this critique does not apply to incidents where activist hedge funds win board seats. activism adds value beyond stock picking. I calculate returns in excess of the market model using the CRSP value-weighted index.+30) Median Average % Positive N= 7. filings in which the activist wins seats convey little information about hedge funds’ ability to pick stocks—only their ability to add value as activists. (B) Purely Active Filings Show Excess Returns In addition to comparing returns around Schedule 13D filings.+30) 5. To isolate these events. Table D-3 presents the regression results. and *** indicate statistical significance at the 10%.2% (t = 1.4% ** 48 (-2.+30) 6. the first time a hedge fund announces activist intentions.3% *** 6.

2% (z = 3. Over 61-days. (C) Returns Differ By Strategy Just as important as average abnormal returns is the cross-sectional variation of those returns. 73. 60. Table V reports the 61.82). having previously filed an activist 13D.33 Figure 3 shows the returns over time.4%. the market reacts favorably to the activist signal. and 6. and having previously filed board seat 13D. 13 D . On average. the green line shows the hedge fund's first time winning a seat.38. (2) what instruments make an activist most effective. 6. The table shows abnormal returns around events after the initial Schedule 13D. and (3) what strategies add the most value.3%. the firms outperform by 8. the red line shows the hedge fund's first time seeking a board seat. Target funds earn positive and significant excess returns surrounding each event date.and 5-day abnormal returns surrounding these events. not just the stock picking signal.76. respectively.6%. Using a regression approach.Purely Active Filings Outperform After Initial Schedule 13D 10% 8% 6% 4% 2% 0% -2% ay s ay s ay s ay s ay s Fi lin g ay s ay s D ay s D ay s D ay s D ay s D ay s 30 D D D D D D D FirstActivist SeekBoard WinBoard -3 0 -2 5 -2 0 -1 5 -1 0 -5 5 15 20 Figure 3. The dependent variable is the 5-day abnormal return 33 In Appendix D. the blue line shows the hedge fund's first activist intention.9%. Table D-4 reports more details.25 - 10 25 .69 and z = 3. z = 2.4% of target firms enjoy positive returns around each event (z = 3. respectively. Table VI analyzes (1) what characteristics make a firm most susceptible to improvement.80). and 66. and z = 3. Having previously filed a passive 13D.70. z = 1. Thus.

intercepts are suppressed . -6. the logarithm of market capitalization.26 32 (2) Activist Instruments t -stat -2. and (3) dummy variables for the various activist strategies.53 -0.49 *** 6. and *** indicate statistical significance at the 10%. For interpretation purposes. sale. Column 3 regresses abnormal returns on dummy variables for those strategies. M&A.07 0.38 *** *** * ** ** *** * *** * surrounding the schedule 13D filing.30 0.56 0.87 3. For all activist investments.26 4.34 112 (3) Activist Strategies t -stat -1.51 -2. All accounting variables are winsorized at the 10% level. and in Column 2 and Column 3.41 2. (2008).75 Coeff. -1.92 -1.22 0.77 -1.17 1.26 - . including winning board seats.64 6. (2) dummy variables for various activist instruments. or launching an additional proxy fight. capital structure.11 Coeff.47 -2.63 7.31 1. t -stat 6.65 0. 5%. Column 2 regresses abnormal returns on dummy variables for those instruments. Column 1 regresses abnormal returns on target firm fundamentals. I adapt the methodology in Brav et al. and the firm’s debt-toassets ratio. I include as regressors (1) the firm’s abnormal q.Table VI Cross-Sectional Variation of Short-Term Returns The dependent variable is the 5-day abnormal return around the Schedule 13D date.55 0.73 1.87 -0. including business strategy. In columns 2 and 3. Thus.00 -0. And for all investments using activist strategies.68 1. the firm’s industry-adjusted margins.08 120 0.99 1. **.25 2. we can interpret each dummy coefficient as the . providing private financing.46 0.98 -0. seeking board seats.85 0.22 -0. *.93 ** -1.30 17.23 ** -2. and R 2 Target Fundamentals Coeff.58 -0. and 1% levels. I suppress the intercept because of the dummy variables’ full span. (1) Dependent Variable: Abnormal Returns Constant ln(MV) Abnormal q Abnormal Margin Leverage BoardWin Board Buyout Financing Proxy Bus Strategy M&A Sale Cap Structure Corp Governance Other No. For all investments using activist instruments. offering to buy the target firm. I express all non-dummy covariates as the deviation from the sample mean.53 -0.60 3. All non-dummy covariates are expressed as the deviation from the sample average values.75 4.69 4. and corporate governance.

partial return for an average target firm when the hedge fund uses that dummy variable.35 Often these buyouts often push targets into another acquirer’s arms—making them a formidable weapon in the activist’s arsenal. not only do activists target undervalued and underperforming firms.7% + 1. (2008). Thus. activists earn the highest returns when targeting firms with low abnormal q (t = -1.99). Both Column 2 and Column 3 include dummy variables which are not mutually exclusive. these buyout offers generate large returns for shareholders.6%). we expect the total abnormal return to be 4. but they earn the highest returns while doing so. Hedge funds that win board seats generate 6.7% additional returns (t = 0. the market seems to think that. the findings presented here identify an overlooked subtlety: rather than simply agitating for change.30). . for 11. Still. if a hedge fund uses both Business Strategy and M&A activism.22). stock market returns no longer proxy for the hedge fund’s return. which depends on what happens after the buyout. hedge funds put their own money on the line. the market reaction indicates activists can improve underperforming firms. I describe more fully this dummy variable regression.3% (= 2. The market also responds very positively to hedge funds that offer to buy the entire firm.0% total excess returns. Consistent with my earlier findings. Whereas prior literature has noted hedge funds’ tendency to drive returns by putting firms in play.27 - . In particular.75).87) and with low industry-adjusted margins (t = -2. This is what we should expect.3% (t = 1. Thus.34 Turning first to target firm fundamentals. as documented in Section III. through board seats. too.9% (t = 4. For example. See Greenwood and Schor (2007) and Brav et al. if activists do add value. driving target prices up 17. I also find that activist instruments yield excess returns. they’re more likely to succeed at firms with more room to improve. 34 35 In Appendix A. But when the hedge fund does purchase the target. seeking board seats generates average abnormal returns of 4. hedge funds can add value to target firms.

The bars show the 90% confidence interval.73).28 - . not just target firms’ fundamentals.31) and 2. these results contradict Boyson and Mooradian (2008).9% (t = 1. Selling the Firm Generates Highest Returns Controlling for Undervaluation and Underperformance 10% 5% 0% -5% Business Strategy M&A Sale Capital Structure Corporate Governance Figure 4. This thoughtful variation provides further evidence that shareholders expect activism to generate . the coefficients reflect activism’s effects. I find that capital structure activism generates positive returns. Figure 4 shows the average return to each activist strategy. This is due to their longer time frame.4% excess returns (t = 4. Indeed. Additionally.53). The chart shows the average 5-day abnormal return for each strategy.03). which implies that activism can improve its targets. this strategy has become less successful over time. when I include investments from 2000 through 2003 by substituting regular data for the industry-adjusted data. Sale activism generates 7. Business strategy and capital structure activism also generate significant abnormal returns of 2. (2008). While consistent with Brav et al.7% (t = 1. Perhaps due to improving governance.In the Short-Run. who find that corporate governance activism generates large returns. because both Column 2 and Column 3 control for abnormal q and abnormal margin. defined as the regression coefficients. These strategies generate statistically different returns (F = 4.

as measured by the coefficients in a dummy variable regression that controls for Tobin's q and margins. I find that short-term activist returns have declined over time. returns have declined. the blue lines plots the average activist returns to Schedule 13Ds filed each year (left axis). The blue line shows the average activist return.9%. and the red line shows the average passive return. the average firm earned only 5. value. the average firm earned 23. (D) Returns Have Decreased Over Time Finally. The line graphs (left axis) show show the average 5-day returns by 13D year. and constants are suppressed. Consistent with Puustinen (2007). I reject the hypothesis that annual returns equal each other (F = 2. but in 2008. The blue bar (right axis) shows the number of active Schedule 13Ds and the red bar (right axis) shows the number of passive Schedule 13Ds filed in my sample each year.Short-Term Returns Decline Over Time As Investments Increase in umber 25% 20% 15% 10% 5% 0% -5% 20 04 20 00 20 01 20 02 20 03 20 05 20 06 20 07 20 08 180 150 120 90 Active Passive 60 30 0 Active Filings Passive Filings Figure 5. in 2001. The bar graph shows the number of Schedule 13Ds filed each year (right axis). if shareholders were purchasing simply on expectations of superior passive investing. In Figure 5. controlling for Tobin’s q and margins. All non-dummy covariates are expressed as the deviation from the sample mean. To the contrary. they would not parse a hedge fund’s press releases to discern its intentions.29 - . I measure how these returns have changed over time.1%. . The results are robust to other models. I regress the 5-day abnormal returns against dummy variables for each filing year.14).

In summary. in fact. putting money to work in increasingly less attractive situations. activists may simply have lost their edge. I find substantial evidence that activism adds value. Thus. activism represents wasted energy at best and value destruction at worst. Second. adjusting for undervaluation and underperformance. Thus. As evidence against this hypothesis. “In the short run. I find no evidence that activists have lost their ability to improve target firms. Benjamin Graham (1973) writes. This section builds on that work to examine whether the weighing machine agrees. I cannot reject the null hypothesis that the coefficients all equal each other (F = 0. Finally. In the long run it is a weighing machine. passive returns (red line) have not demonstrably improved. outside observers might confuse true activism for passive investing. even if the hedge funds have already disclosed passive investments in the same targets. as activism becomes more common. After all. Activism aimed at undervalued firms and attempts to sell target firms earn the highest returns. the market is a voting machine.This pattern suggests three interpretations. investors applaud hedge funds’ activist intentions. hedge funds have shed their most aggressive tactics. First. when hedge funds actually do intervene.” Using what Graham might consider an exit poll. however. neither corporate shareholders nor hedge funds’ clients benefit. Section IV argued that investors believe activism adds value. if activism fails to generate long-term excess returns. as companies have learned to accept activism. shareholders might begin to expect it. . In Section VI. investors are less surprised. Activist investments outperform passive investments.30 - . however. though. Section V: Long-Term Stock Returns In the Intelligent Investor. Recently. these returns have declined. In addition.48).

we should expect these investments to continue to outperform as the activism persists. In addition. I find strong evidence that activist investments.If hedge fund activism improves its targets beyond the market’s pre-targeting expectations. then abnormal returns might simply mark great passive investing. the market should bid stock prices to a level reflecting the expected probability that the activist will succeed (Brav et al. Even excluding the 61-days surrounding the filing. If this were not the case—if the market price ever reached a level where further activism would not generate further abnormal returns—then the hedge fund would exit its position. 2008). Second. I investigate when activists add the most value—what target firm characteristics. At every point in time. not activist ability. I compare activist filings to passive filings. controlling for undervaluation and underperformance. In this section. activist investments outperform passive investments—particularly when the hedge fund seeks board seats. however. activist investments earn significantly higher returns. if hedge funds could persistently pick winning stocks without improving them. the market should continue to reward the effort. Hedge funds . outperform passive investments. I conclude that activists add value in the long-run. First. the efficient markets hypothesis proves crucial to the argument. likely because activism has rejuvenated them. In particular. Because success is never certain. underperforming target firms earn the highest long-term returns.31 - . Again. activism appears to add long-term value. As in the short-term. Instead. I compare activist filings to passive filings. activist instruments. Thus. as the activist continues to wage a campaign. or activist strategies prove most effective? As in the short-run. I also investigate heterogeneity across these returns. undervalued. particularly those which seek board seats. then target firms should outperform their risk-adjusted benchmarks. using two tests.

any observed difference between active and passive investments is similarly understated. Thus. no 36 Unlike activist investments. (2008) analyze hedge fund holding period returns—buying when hedge funds buy and selling when hedge funds sell—but this methodology (1) introduces error because we cannot perfectly measure these holding periods and (2) conflates hedge funds’ market timing ability with their activist ability. whether positive or negative. we can isolate a hedge fund’s intentions.32 - . this method should bias observed abnormal returns downward. In particular. I analyze long-term returns along arbitrary windows. Long time horizons also complicate how we categorize activism. To most purely match the hedge fund’s behavior to the relevant investment period. their abnormal returns should trend towards zero. I distill my sample as follows: passive investments include all instances where a hedge fund never amends its passive Schedule 13D.5% of investments. I observe that these returns have declined over time. Thus. I find that hedge funds significantly change strategies in 79. but over 2 years. is likely understated. improve targets’ business strategies. hedge funds revise their strategies numerous times. I still find that active investments both generate positive abnormal returns and outperform passive investments. because this downward bias affects passive investments less than it affects activist investments. To the extent activist hedge funds adopt different time frames for different investments. . Finally. once target firms realize their potential.earn excess returns when they sell target firms. Instead. Before turning to the analysis. Further. activist investments include all instances where a hedge fund announces an activist instrument or an activist strategy within 180 days of its Schedule 13D. or optimize targets’ capital structures. to better measure the reward to corporate shareholders (at the expense of the returns to hedge fund investors). passive investments have no catalyst. Brav et al. Over 5 or even 61 days. The major obstacle to analyzing long-term returns is picking a time period. it is worth elaborating on the methods I will use. any observed effect. however. they have no holding period sensitivity.36 Even in spite of this.

(A) Activism Exceeds Stock Picking Table VII measures long-term returns.8% *** (+1. Over the entire 25-month period.8% * 11.5% 53. and *** indicate statistical significance at the 10%. activist investments earn 10.33 - . *.8% * board seat investments include all activist investments where a hedge fund never seeks board seats.1% *** 13.08). Panel B compares activist investments seeking board seats to activist investments that do not.9% * Panel B: Board Seats vs. Figure 6 plots the excess returns over time.87) while .5% *** 61.1% *** 21.+24) 15.+24) 13.8% *** 58.6% ** (+1. **.+24) 7.1% ** Passive (+1. After the short-term reaction subsides.+24) Median Average % Positive 28.0% *** 14. From one month before the Schedule 13D until one month after.3% *** 21.4% *** No Board Seats (-1.3% *** 60.9% * (+1. 5%.+24) 10.7% 17.1% of passive investments show positive returns (z = 5.4% * 3.+24) 12. I report p -values for medians using the Wilcoxon Signed Rank Test. from one month after the Schedule 13D until 12 months after.2% *** (+1. Panel A compares activist investments to passive investments. I calculate excess returns (1) from one month before the Schedule 13D until 24 months after and (2) from one month after the Schedule 13D until 24 months after. 66. Table D-5 shows more details.7% 55. Consistent with short-term results.3% *** 66. and board seat investments include all activist investments where a hedge fund seeks a board seat within 180 days of its Schedule 13D.0% ** Difference (-1.70 and z = 2.9% *** (+1. both activist and passive investments earn positive and significant excess returns. and 1% levels. I estimate parameters over the 36 months prior to the Schedule 13D. In Appendix D.+24) 5.2% ** 16. Panel A: Active vs.3% 1.+24) 9.+24) 6.1% *** (-1.3% excess returns and passive investments earn 2.8% ** 13.2% *** 75. activist investments flourish while passive investments falter: activist investments increase an additional 8.+24) Median Average % Positive 20.Table VII Long-Term Returns by Investment Type The table presents average monthly returns around the Schedule 13D filing in excess of the market model using the CRSP value-weighted index.+24) 10.7% *** Difference (-1. Passive Active (-1.+24) 18.5% (z = 3. Other Active Investments Board Seats (-1.6% 7.8%.8% of activist investments and 55.6% *** 66.2% *** 31.

Active Investments Outperform Passive Over 25-month Event Window 40% Board 30% 20% 10% Passive 0% -10% -1 13 M D on F t 1 ilinh M g on th 3 M on th s 6 M on th s 9 M on th s 12 M on th s 15 M on th s 18 M on th s 21 M on th s 24 M on th s Active NoBoard Figure 6. This makes sense. activist investments earn 13.55) while passive investments earn 1. the strategy is not investible because certain investments reveal their activism after the event date. (2008). so should returns. these changes play out slowly over time—and so should returns. In contrast. the market continues to receive positive information.16). Brav et al. By estimating target firms’ factor loadings . These results are consistent with Boyson and Mooradian (2008). from 1 month prior to the filing until 24 months afterward.5% excess returns (z = 5. the chart plots the average excess return around the schedule 13D filing. As passive investing reveals itself all at once. to the extent activists do improve their targets. It’s also worth noting that this activist momentum is not a free lunch. Over a longer window.In the Long-Run. The dashed lines divide active positions into those in which hedge funds sought board seats (long dashes) and those in which hedge funds did not (short dashes).5% (z = 0.34 - . First. The increasing returns imply that over two years. activism might carry more risk than passive investing. from one month after the Schedule 13D until 24 months after. For active (blue) and passive (red) hedge fund investments. Additionally. and Klein and Zur (2008).4% (z = -0.29). passive investors’ subsequent actions matter very little. passive investments fall 0.

I compare activist returns to passive returns. Panel B shows that board seat activism outperforms other activism by 16.9% comes after the short-term reaction (t = 1. Panel A in Table VII shows that over 25 months. controlling for Tobin's q and margins. 11. I use a dummy variable regression to control for the possibility that hedge funds might go activist in more attractive targets. measured by a dummy variable regression.0% 30. passive.57) and by 13. board.0% 0.0% 10.0% 20. I regress the 25-month return on dummy variables for activism and board seat activism.9% over the full period (t = 1. Overall.61). before the hedge funds launch their activism.31). Of this outperformance. What looks like excess risk-adjusted returns relative to the stocks’ prior risk might actually represent nothing more than benchmark returns relative to the stocks’ current risk. . To test whether activist hedge funds do more than pick stocks.0% Active Passive Board No Board Figure 7. and no board investments.In the Long-Run. activism outperforms passive investing by 17.28). To more cleanly distinguish activism from stock picking.8% after the short-term reaction (t = 1.6% (t = 2.35 - . activism seems to add value. I underestimate the target firms’ risk and thus their benchmark returns. The chart shows the average 25-month returns to active. Active Investments Outperform Passive Controlling For Undervaluation and Underperformance 40.

(B) Returns Differ By Strategy As with short-term returns.71 -0.91 -10. 5%. activism adds value beyond stock picking. All non-dummy covariates are expressed as the deviation from the sample average values.03 Coeff.47 -1. **. activists earn larger returns from undervalued targets and underperforming targets.63 -0. and *** indicate statistical significance at the 10%. intercepts are suppressed .08 * *** ** controlling for Tobin’s q and margins. *. In columns 2 and 3.21 (3) Activist Strategies Coeff.68 39.14 ** 0.4% (t = 2.93 0. For all activist investments.87 -48.45 -0.12 -0.11 ** -1.70 0. Column 2 regresses abnormal returns on dummy variables for those instruments.36 - . Activism outperforms passive investing by 18. And for all investments using activist strategies.10 * -0.67 0.59).17). -16.31 -1.38 0.91 0.16 -1.86 -15. -23.42 0. 8.14 -20.04 243 t -stat 1.Table VIII Cross-Sectional Variation of Long-Term Returns The dependent variable is the 25-month abnormal return around the Schedule 13D date.84 -51.90 4.04 0. In Table VIII. I regress 25-month abnormal returns on (1) target firm fundamentals.37 The results are robust to other models. I investigate the conditions that make activism most effective. Table D-6 shows the underlying regression.48 13.88 -0.93 -8. and R 2 Target Fundamentals Coeff.62 12.46 -0.29 101 t -stat -1. Consistent with short-term 37 In Appendix D.88 0.6% (t = 1.94 -8.60 0.82 -14.31 3. Thus.14 0. Turning first to target firm fundamentals. and (3) activist strategies. (2) activist instruments.56 -0. Column 3 regresses abnormal returns on dummy variables for those strategies. . and board investments outperform other activism by 18.65 -1.72 * 1. (1) Dependent Variable: Abnormal Returns Constant ln(MV) Abnormal q Abnormal Margin Leverage BoardWin Board Buyout Financing Proxy Law Bus Strategy M&A Sale Cap Structure Corp Governance Other No.16 -1. For all investments using activist instruments.01 90 (2) Activist Instruments t -stat *** -2. Column 1 regresses abnormal returns on target firm fundamentals. and 1% levels. All accounting variables are winsorized at the 10% level.

investments that seek board seats earn 39. the longer period shows a larger coefficient on winning board seats.88). Over 25-months. the effect becomes positive. making the investment a success. After removing withdrawn board seats 38 Part of this increase is due to reduced precision in the covariates.37 - . Consistent with short-run expectations.9% (t = -1.65) and low abnormal margins (t = -1. As expected. this coefficient is biased because the regression excludes investments initiated from 2000 through 2003.31).40) to investment returns.9% excess returns (t = 3. This result contradicts short-term returns—where shareholders applaud the news that a hedge fund has won a board contest. But this contradiction is not very robust. unprofitable target firms. Alternatively. of all failed board contests in my sample. In fact. activist hedge funds might distract management from running the firm. Proxy fights and lawsuits generate moderately significant negative returns of -14. activists perform best when firms have low abnormal q (t = -1. they ensure that management hears their message.4% (t = 0. within investments that seek board seats. First. board seats continue to drive value in the long-run. When I increase the sample by substituting Tobin’s q for abnormal q and substituting margin for abnormal margin. this coefficient is biased by hedge funds that withdraw their requests after management cooperates. this finding strongly supports the hypothesis that activists can add value.7% (t = -1.38 Second. . the hedge fund might be running into a brick wall—a stubborn management team that refuses to cooperate with value-increasing propositions. respectively. far exceeding any other activist instrument.12). To corroborate that sample size plays a role.16) and -48. The evidence supports the hypothesis that activists do their best work turning around unloved. by seeking board seats.3% of cases the hedge fund voluntarily withdraws its request. in unreported results I run the new regression for the 2004-2008 period and for the 2000-2008 period. those which win board seats actually underperform those which do not win board seats by 20. winning adds 6.0% (t = -1. With respect to activist instruments. in 28. In these hostile situations.91). Contrary to expectations.returns.

they should differ in fundamental performance.38 - . multiple activist strategies earn positive returns.57). not necessarily returns. In fact. As in the short-run. Also consistent with short-run expectations. the effect of winning becomes positive. Winning board seats increases returns by 7. The bars show the 90% confidence interval.20).5% (t = 0. dropping the board contest once management acquiesces to their demands.46). defined as the regression coefficients.Many Strategies Generate High Returns Controlling for Undervaluation and Underperformance 40% 20% 0% -20% -40% Business Strategy M&A Sale Capital Structure Corporate Governance Figure 8. In addition to activist instruments. if strategies differ. When broadening the sample by substituting Tobin’s q for abnormal q and margin for abnormal margin. however.2% (t = 0. different strategies’ returns are statistically indistinguishable from one another (F = 0.28). corporate governance activism outperforms by 9.9% (t = -0.9% excess returns (t = 0. The chart shows the average 25-day abnormal return for each strategy. . from the original model. earning 13.9% excess returns (t = 0. it appears activists use board seats as a bargaining chip. corporate governance activism underperforms the market by 8.67) and 12. investments which change the target’s business strategy and capital structure perform very well. respectively. This does not itself disprove the hypothesis that activists follow different strategies.71). this effect disappears. Thus. Figure 8 shows the average return per strategy. Instead.

Especially if activists try to sell wounded companies. successful sales earn positive returns (9. The long dashes show only investments which were subsequently sold. The chart plots the average excess return around the schedule 13D filing. This result is most likely due to failed sales. from 1 month prior to the filing until 24 months afterward. I remove both successful and failed sales.48). I run the same regression with a dummy variable set to one if the company was sold.3% with t = 0.Activism Adds Value Beyond Selling Companies But Selling Generates Very High Returns 40% Sold Firms All Firms 30% 20% All Firms ex. those remaining unsold will dramatically underperform. The short dashes excludes any investment which was sold as well as any investment in which the activist sought a sale but failed. The chart plots active investments only. The solid line includes all active investments. generates no long-term returns. who asymmetrically remove only successful sales. Sale 10% 0% -1 13 M D on F th 1 ilin M g on th 3 M on th s M on th s M on th s on th s on th s on th s on th s M 21 24 on th s M M M 6 9 12 15 Figure 9. sale activism. Most shockingly. To test Greenwood and Schor’s (2008) hypothesis that activists add value exclusively by putting firms in play. this better measures . To counter this effect. I re-run the event study excluding those firms which the hedge funds try to sell.39 - 18 M . By eliminating the bias from failed sales. Unlike Greenwood and Schor. in unreported results. which generated the highest short-term returns. As expected.

The results are robust to other models. To the contrary. and the red line plots the average passive return. The blue line plots the average active return. (C) Returns Have Decreased Over Time Finally. In stark contrast to Greenwood and Schor.3% returns over 25 months (t = 5. This result differs from Greenwood and Schor both because I exclude failed sales and because my sample includes more experienced hedge funds. This pattern suggests the same three interpretations: (1) investors might . as measured by the coefficients in a dummy variable regression controlling for Tobin's q and margins. I regress the 25-month abnormal returns against dummy variables for each filing year. the average investment earned 23.56). Figure 10 plots the results over time. controlling for Tobin’s q and margins. I reject the hypothesis that annual returns equal each other (F = 2. I find that activism still generates significant excess returns. The blue bar (right axis) shows the number of active Schedule 13Ds and the red bar (right axis) shows the number of passive Schedule 13Ds filed in my sample each year. I measure how these returns have changed over time. The line graphs (left axis) show show the average 25-month return by year. these targets still generate 22. Figure 9 shows that while they underperform firms which were sold. returns have declined over time. in 2001.1%.6%.79). and in 2007 the average investment earned only 7.Long-Term Returns Decline Over Time As Investments Increase in umber 175% 125% 75% 25% -25% 20 04 20 06 20 01 20 00 20 02 20 03 20 05 20 07 20 08 180 135 90 45 0 Active Passive Active Filings Passive Filings Figure 10. hedge funds’ ability to add value beyond selling the target.40 - .

this section strengthens earlier results. Both relative to passive targets and in absolute terms. even adjusting for undervaluation and underperformance. By adding further evidence that activism adds value. Third. activist target firms shrink dramatically.27). activists optimize the capital structure. While passive target firms actually grow. Second. Fourth. activist targets increase their margins. I document four non-mutually exclusive ways that activist hedge funds improve target firms. activists reduce firm size. however. I cannot reject the hypothesis that passive returns equal each other in every year (F = 0. In summary. Perhaps more importantly. it also describes the levers activists pull to earn those returns. activist investments outperform passive investments. we now turn our attention to target firm fundamentals. return on assets. activists improve profitability. and return on equity. Throughout. these returns have declined. Over the long-run. Recently. . often by disgorging cash. activists sell target firms. Relative to passive targets. (2) activism might begin to look more like passive investing. Activism aimed at undervalued and underperforming firms earns the greatest excess returns. however.41 - . I find further evidence that activism adds value. First. activist targets sell themselves more frequently and for higher premiums. As evidence against the second hypothesis. Section VI: Target Firm Fundamentals Having observed that activism earns greater returns than stock picking might imply.increasingly expect activism. and (3) activists might have lost their edge. I also find evidence that activists use board seats to build long-term value—consistent with the evidence presented in Section III that business strategy activism most frequently seeks board seats while M&A activism least frequently seeks board seats. Activist targets increase their leverage and payout ratios.

for instance. relative to passive targets. excess returns increase over two years. activists target wellperforming firms. the 39 As noted in Section III. To the contrary. If the market had predicted that target firms would improve. In addition to arguments about target firms’ characteristics. Three factors mitigate this stock picking hypothesis. but relative to the median firms in their industries. a hedge fund’s portfolio might regularly outpace its peers in revenue growth. I further debunk the stock picking hypothesis by comparing hedge funds’ active targets to the same hedge funds’ passive targets. activists target undervalued companies. excess returns should not rise over time but rather should equal zero. no efficient markets hypothesis governs accounting performance. 40 I find significantly negative correlations between initial levels and subsequent improvements.40 Second. As discussed earlier.39 To the extent certain companies can improve more than others. Thus. not in their ability to equalize those fundamentals. The companies expected to outperform should not be undervalued but rather overvalued to account for yet-unrealized profits. then hedge fund activism must cause any outperformance by the activist group. activist targets are poor performers. If active and passive targets come from the same underlying pools. The beauty of financial markets lies in their ability to correctly price target firms’ fundamentals. First. we must distinguish activist hedge funds’ ability to improve their targets’ performance from activist hedge funds’ ability to pick stocks that everyone expects will increase profitability. Thus. by simply investing in stocks with high P/E ratios.Before turning to the data. it shouldn’t be well-performing firms but rather underperforming firms with more room to improve. a cautionary note is in order. . activist targets are good performers. And third. there is no reason that even a passive hedge fund should not be able to consistently pick stocks whose profitability increases more than does the aggregate market.42 - . Thus. hedge funds don’t seem to simply pick stocks that everyone expects will increase profitability. Identifying firms which outperform by accounting standards presents a more difficult task than identifying those which outperform stock market indices. though.

73 Passive 2 yr -3.13 No Board Seats 1 yr 2 yr -1.41 6.02 8. .38 -2.87 -8. I still observe that activist targets outperform.15 *** 16. from the date of the first Schedule 13D filing. *. **.97 2.59 -0. For example.71 Difference 1 yr 2 yr 1. All data from Bloomberg. and 1% levels. this should handicap activist targets’ ability to outperform.84 -4. in target firms' trailing twelve months EBITDA divided by sales (Margin) . Just as bad.and 2-year increases.21 0.28 8. Because I nonetheless observe outperformance. CEOs paid based on earnings growth might increase earnings by investing in projects with IRRs greater than zero but less than the firm’s WACC.20 -1.95 0.29 1 yr -1.08 1.76 2. (A) Activism Improves Profitability By improving target firms’ capital allocation and strategic positioning.48 2 yr 0. firm value. hedge funds go active in worse performing companies. Additionally. 5%.99 *** Panel B: Board Seats vs.43 - .05 1.27 2. All data is winsorized at the 10% level. hedge funds increase cash flow—and thus. Management might fail to do so for many reasons. For the reasons discussed above.73 2.44 0.54 0. EBITDA divided by total assets (ROA) .58 pools differ. and *** indicate statistical significance at the 10%. While this might give activist investments more room to improve.11 *** 5.97 1.70 Difference 1 yr 2 yr 0.92 -0.11 *** 7.39 -2.21 *** 4. hedge funds go active in more undervalued companies. CEOs paid based on ROA might forgo projects with attractive IRRs if they drag down a firm’s overall returns. and EBITDA divided by book value of equity (ROE) . after controlling for this effect.Table IX Profitability Improvement By Investment Type The table shows the 1. Panel A: Active vs. Passive Active 1 yr Margin ROA ROE -0. this does not appear to be a problem.80 ** 0.86 -1.33 2. Other Active Investments Board Seats 1 yr 2 yr Margin ROA ROE 0. Most noticeably.

Active Investments Become More Profitable Controlling For Initial Levels Return on Assets
3.0% 2.0% 1.0% 0.0% -1.0% -2.0% Active Passive Board No Board 2.0% 1.0% 0.0% -1.0% -2.0% -3.0% Active Passive Board No Board

EBITDA Margins

Figure 11. The chart shows the average 2-year improvement in ROA and in margins for active, passive, board, and no board investments, controlling for Tobin's q and for the initial levels of each variable.

Table IX measures increased profitability in three ways: EBITDA margins, ROA, and ROE. The former two measures are particularly appropriate because they isolate core operational performance.41 Panel A shows that, while passive investments falter, activist targets improve their profitability in absolute terms. Further, over two years, activist targets improve their margins by 4.1% more than passive targets (t = 4.26), their ROA by 5.2% more (t = 4.49), and their ROE by 17.0% more (t = 4.07). Panel B shows similar, but less significant results for board seats; they improve margins and ROA. While they have little effect on ROE, this is because board seats decrease target firms’ leverage. Figure 11 dismisses the alternate hypothesis that this outperformance comes from stock picking. I regress the 2-year increase in ROA and in margins on dummy variables set to one for activist investments or board seat investments. I control for the Tobins’ q and for the initial value of each accounting metric. I express all non-dummy covariates as the deviation from the sample
41

ROE includes capital structure decisions. It equals ROA divided by one minus the debt-to-assets ratio.

- 44 -

mean. We interpret the constant term as the profitability increase to an average passive target, and we interpret the dummy coefficient as the difference for an average active target.42 While the average passive firm becomes less profitable, with ROA falling by 1.2% (t = -1.70) and margins falling by 2.5% (t = -3.71), the average active firm becomes more profitable. The average active target increases its ROA by 3.0% more than the average passive target (t = 3.05), and the average active target increases its margins by 2.9% more (t = 2.76). Board seats show but less significant similar results. Overall, activists increase firms’ profitability. (B) Activism Optimizes Capital Structure By optimizing a firm’s capital structure, activists increase the cash flow to shareholders. First, by increasing leverage, and thus interest expense, activists can reduce taxes. Second, by disgorging cash otherwise earning at best middling returns, activists can increase the present discounted value of the firm’s equity cash flows. Management might fail to optimize the capital structure for many reasons. For example, because CEOs depend on the firms for their jobs, they might take more caution than shareholders might like before leveraging the firm. Table X tracks changes in target firms’ capital structures. Leverage refers to the debt-toassets ratio and payout measures dividends as a percent of net income. In addition to the average change in each ratio, the table also shows the percent of firms which increased these ratios. Panel A shows that activist targets increase their leverage and payout ratios more than passive targets. The results are much stronger over one-year horizons than two-year horizons. This makes sense; unlike profitability measures, which ideally increase over the years, these capital structure ratios fluctuate. Leverage naturally declines over time as firms pay down their debt, and payout ratios might bounce around with special dividends. Over one year, activist targets increase their leverage by 1.1% more than passive targets (t = 1.65) and increase their payout ratio by 4.5%
42

In Appendix D, Table D-7 contains the regression results.

- 45 -

Table X Capital Structure Changes By Investment Type
The table shows the 1- and 2-year increases, from the date of the first Schedule 13D filing, in trailing twelve months debtto-assets (Leverage) and dividends divided by net income (Payout) . The chart shows both the mean difference in the ratio as well as the percent of firms which increased their ratios. All data from Bloomberg. All data is winsorized at the 10% level. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels.

Panel A: Active Investments Lever Capital Structure More Than Passive Investments Active 1 yr Leverage % Increase Payout % Increase 0.64 43.6% 5.37 32.6% 2 yr 1.51 45.3% 0.53 18.5% 1 yr -0.45 35.4% 0.91 12.7% Passive 2 yr 1.56 42.2% 10.00 16.7% Difference 1 yr 2 yr 1.09 ** 8.2% 4.46 ** 19.9% *** -0.05 3.1% -9.47 ** 1.9%

Panel B: Board Investments Lever Capital Structure Less Than Non Board Investments Board Seats 1 yr 2 yr Leverage % Increase Payout % Increase -0.75 33.8% 7.36 26.7% 0.72 40.0% 0.76 17.6% No Board Seats 1 yr 2 yr 2.08 58.1% 3.58 40.0% 2.86 55.0% -0.75 20.0% Difference 1 yr 2 yr -2.83 ** -2.15 -24.2% ** -15.0% 3.78 -13.3% 1.51 -2.4%

more (t = 2.08). In addition, 8.2% more active firms increased their leverage than did passive firms (t = 1.25), and 19.9% more active firms increased their payout ratios than did passive firms (t = 2.38). Thus, activist targets lever up more than their passive peers. Panel B shows that investments seeking board seats take on 2.8% less leverage (t = 2.21) than other activist investments. Two mutually exclusive reasons might account for this observation. Most consistent with results elsewhere, activists might seek board seats to provide long-term strategic advice. In Section III, I showed that activists most frequently take board seats to advance business strategy goals and least frequently to advance M&A goals. These activists might focus their efforts on building value before taking on additional financial risk. Alternatively, activists might seek board seats when they anticipate management will otherwise dismiss their proposals. These stubborn, entrenched management teams might resist leverage.

- 46 -

5% more than passive firms (t = 1.43 While the average passive firm marginally increases its payout by 2. Figure 12 shows that. Table D-8 shows more detail.Active Investments Increase Payout Controlling For Initial Levels and Initial Cash Payout Ratio 3. and no board investments. Thus. 43 In Appendix D.7% more than passive firms (t = 0.3% (t = 1.0% Active Passive Board No Board Figure 12. I control for Tobin’s q.0% 0. Thus.0% 1. The chart shows the average 1-year change in payout ratio and in the debt-to-assets ratio for active.0% Active Passive Board No Board -1.98).0% 0. passive. larger firms lead to greater compensation and greater job security.4% (t = -0. (C) Activism Reduces Firm Size Activist shareholders also add value by counteracting management’s harmful tendency to build empires. by divesting unprofitable assets activists can improve future returns. Payout increases by 2. .0% Leverage 2. For CEOs. To grow the firm.79) and decreases its leverage by 0. initial levels of each variable. and the initial level of cash as a percentage of total assets.83). once again. it appears activism earns some of its gains by disgorging cash and leveraging the capital structure.0% 1. and leverage increases by 0. board. they might purchase assets which yield suboptimal returns. and initial cash holdings.0% 2. these results are due to activism.26). not to stock picking. the initial value of each accounting metric. controlling for Tobin's q . the average active firm increases both payout and leverage.47 - .

Panel A: Active Investments Shrink More Than Passive Investments Active 1 yr Assets Sales Cash -4. and *** indicate statistical significance at the 10%. 5%.1% Table XI examines this hypothesis.0% 1.4% 0.4% -8.1% 2 yr -6.1% 1.4% 24. and cash.5% 5.0% Difference 1 yr 2 yr -5. but on the other hand. **. The inconsistent evidence likely reflects two conflicting forces: on one hand.23) and by 6. *. and 1% levels.04).0% 35. activist targets reduce their sales by 4.6% more than other activist investments (t = -2.11).5% 0. While active targets also appear to reduce their cash by 8. On average.36).4% 22.7% 30. measured by assets.3% 50.0% Panel B: Board Investments Shrink More Than Non-Board Investments Board Seats 1 yr 2 yr Assets Sales Cash -7.6% -3. active targets reduce their assets by 4.2% 1 yr 0. expressed as a percent change from their values in the year prior to activism. Panel B shows that board seat investments shrink the most.0% more over two years (t = 1.6% 0. sales. sales. they appear to increase their cash by 11.8% more than passive targets over one year (t = -3.5% Passive 2 yr 0. active target firms actually shrink.80) over . Each variable is expressed as a percentage of its value in the year prior to the Schedule 13D filing. activist target firms are shrinking.5% ** 5.2% Difference 1 yr 2 yr -4.3% 20.8% *** -6.and 2-year increases. Similarly.9% No Board Seats 1 yr 2 yr -1.0% *** -8.6% more over two years (t = -2.Table XI Agency Cost Reductions By Investment Type The table shows the 1.4% more (t = 0.2% more than passive firms (t = -2. All data from Bloomberg.2% *** -1.4% -22. from the date of the first Schedule 13D filing.3% more than passive firms over one year (t = 1.6% *** -10.1% *** -5. All data is winsorized at the 10% level. adjusted for core inflation.0% -9.0% -4.89) over one year and by 1.3% 11. Investments where activists seek board seats shrink their assets by 5. in inflation-adjusted firm size. It shows assets. they are also increasing profitability.58) over two years. and cash.1% 27. Panel A shows that while passive target firms grow.8% -2.6% 26.48 - .6% *** -4.

shrinking 4. To isolate activism from stock picking. Taken together with the earlier evidence that activist firms increase profitability. . controlling for Tobin's q and for the initial firm size. As Jensen (1986) argues.81).4% more (t = -1.8% (t = 1.34).0% 0.1% more (t = -2. Similarly. They also shrink their sales by 5. effectively reduce firm size.44 While passive targets actually increase assets by 2. activist investments. board. passive.5% 0. this decreases agency costs.0% more (t = 1. but board seat investments decrease assets by 3. Table D-9 shows more details.0% -2.6% more (t = -2.0% 2.8% (t = 0.04) over one year and by 8. 44 In Appendix D. rather than unintentionally shrink into irrelevance. the activist investments seeking board seats grow the least. The chart shows the average 2-year change in assets and sales.7%.49 - .5% more (t = 1.Active Investments Reduce Assets Controlling For Initial Assets Asset Growth 4. for active.97) over two years. active investments decrease assets by 1.0% 4. and no board investments. one year and by 10.0% -4.0% Active Passive Board No Board 6.0% 1. Thus.5% 3.8%. shrinking 4.12). The changes in sales show similar patterns. particularly those with board seats. Figure 13 controls for differences Tobin’s q and in firm size. while all investments grow their sales.0% Active Passive Board No Board Sales Growth Figure 13.49). it appears that activist investments purposefully pare down.55) over two years. adjusted for inflation. active investments without board seats actually increase assets by 0.

3% 16.7% *** 18.3% 31. and second.8% 21 24 mo 30.8% 47 Passive 6 mo 0. activists both put their targets into play and market them better than do passive investors.3% higher premiums (t = 1. activists add value by selling their targets to acquirers who values them more highly.4% 24 24 mo 17. **. But For Higher Prices Board Seats 6 mo % Sold Avg. activists sell 9. hedge funds’ activist targets sell themselves more frequently and for higher prices. Thus. activist investors sell 13.93) and on average sell them for 18.50 - . even if it benefits shareholders. I measure the total buy-and-hold gross return from one month before the Schedule 13D until the sale closes.9% 14 No Board Seats 6 mo 19. it measures the percent of targets which sell themselves. Executive managers might fail to sell the firm. and 1% levels. while passive investors sell none.2% 44.9% 33 Difference 6 mo -17. Panel A: Active Investments Sold More Frequently Active 6 mo % Sold Avg.9% 39.2% *** 24 mo 13. *. it measures the average premium those firms fetched. Premium umber Sold 9.3% * Panel B: Investments With Board Seats Sold Less Frequently.0% 3 24 mo 10.5% 9 Difference 6 mo 9.1% ** (D) Activism Sells Target Firms Finally.2% 21.2% 20. Table XII measures how well activists sell their targets.8% *** 26. Premium umber Sold 2.64). Panel B separates active investments seeking baord seats from those which do not. Average premium is calculated from one month before the hedge fund files a Schedule 13D until the date on which the sale closes. Panel B shows that when activists seek . First. Panel A shows that.7% more targets (t = 4.0% *** 27.0% 0 24 mo 4.Table XII Target Firm Sales By Activism Type The table shows the percent of all firms which sold themselves and the average premiums they received. Panel A separates active investments from passive investments.1% *** 24 mo -19. and *** indicate statistical significance at the 10%. because doing so imperils their job security. 5%.4% 57. Over two years. Within the first 6 months. relative to their passive investments. To capture the premium attributable to the activist.2% of their targets.

91). they sell firms less frequently but for much higher prices. In particular. having built a strong business. I find strong evidence that different strategies work in different ways. Table D-12. Over two years. they’re less likely to sell—in particular. I run cross-sectional regressions like those in Section IV and Section V. rather than sheer luck. In total. When doing so. To describe these strategies. I describe the differences among the various activist strategies. Again.78) 45 In Appendix D. To facilitate interpretation. this result is consistent with the interpretation that activists use board seats to build long-term businesses. we should expect activists’ professed strategies to predict target firms’ improvements with reasonable accuracy. For all investments which use activist strategies. Thus. In general. Table D-10. 19. (E) Improvements Differ by Strategy If hedge fund activism.8% less likely over two years (t = 3. Table D-11.45 Business strategy activism builds profitable businesses. When applicable. But when they do sell. This section tests that hypothesis. activism unlocks substantial value by selling the target firm.board seats.0% (t = -1. as described earlier in this section. activists also reduce target firms’ payout ratios by 10. all non-dummy covariates are expressed as the deviation from the sample mean. . Perhaps to remove distractions from building these businesses. and Table D-13 show more detail. I control for the dependent variable’s initial value. they demand a healthy premium— on average. we interpret the coefficients on each dummy variable as that strategy’s partial effect on a firm with average characteristics.9% (t = 2. relative to passive investments.77). Figure 14 shows the key results. and all constants are suppressed. I regress the accounting statistics described earlier on dummy variables for each strategy.1% larger (t = 2.51 - . ROA increases 6.71) and margins increase 2. 26. catalyzes these observed changes.10).3% (t = 0.

74 6% 0% -6% Business Strategy M&A Sale Capital Structure Corporate Governance Panel B: Capital Structure Activism Increases Payout 20% F = 1.11 5% 0% -5% -10% Business Strategy M&A Sale Capital Structure Corporate Governance .39 10% 0% -10% -20% Business Strategy M&A Sale Capital Structure Corporate Governance Panel C: Corp Gov & Sale Activism Decrease Assets 10% F = 1.Panel A: Business Strategy Activism Improves ROA 12% F = 2.52 - .

73). For all activist investments which use activist strategies. it earns on average a 24. . When hedge funds fail to sell these firms. the blue bars (left axis) show the sale odds ratio from a logistic regression that regresses a dummy variable equal to one if the firm sold itself on strategy dummies. these dependent variables are regressed on dummy variables for each strategy. Presumably due to increased profitability and lower payout ratios. All non-dummy covariates are expressed as the deviation from the sample mean. and leverage by 1. they often seek divestitures. Where applicable.6% premium (t = 2. When a hedge fund demands that a target firm sell itself.53 - . regressions control for the dependent variable's initial level.77) and sales increase 4.6% (t = 1. M&A activism sells firms more frequently.2% (t = -0. and asset growth (Panel C). M&A and sale activism effectively push target firms to sell themselves.04). Panel B.74).5% over two years (t = -1.3 times more likely to sell themselves as those which do not (t = 1.14). and all constants are suppressed. target firms’ assets increase 3.Panel D: M&A Activism Sells Firms Most Frequently Corporate Governance Activism Earns Highest Premiums 250% 50% 200% Sale Odds Ratio Sale Odds (left axis) 150% Sale Premium (right axis) 40% Averarge Premium 30% 100% 20% 50% 10% 0% Bus Strat M&A Sale CapStruc CorpGov 0% Figure 14. reducing total assets by 4. In Panel D. These charts plot the results of cross-sectional regressions on various accounting metrics: ROA (Panel A). I also show cross-sectional variation of firm sale statistics (Panel D). those in which the hedge fund uses a M&A-related strategy are 2. The red outlines (right axis) show a linear regression that regresses the average premium on the strategy dummies. payout ratios (Panel B).2% (t = 0. and sale activism sells them for higher prices.43). and Panel C show 90% confidence intervals and F statistics on the null hypothesis that all strategies generate equal improvements. Panel A. Within investments that use activist strategies.

reduce their potential agency costs. activists struggled. towards the middle of the decade. I find much more significant evidence that . I find manifold evidence that activism adds value. I find no evidence that activists’ abilities have declined over time. perhaps reflecting heterogeneity even within the strategy—while some hedge funds want the firm to leverage itself. and sell themselves more often and for richer valuations. activism adds value in the pattern we would expect given hedge funds’ stated strategies. I find no support for the hypothesis that activist skill has declined. optimize their capital structure. As Jensen (1986) argues. and towards the end of the decade. Interestingly. I conduct joint hypothesis tests on the null hypothesis that the yearly coefficients all equal one another.8% (t = -1. Target firms improve their profitability.71). Target firms increase their payout ratio by 10. (F) Activist Ability Has ot Declined In unreported results. management teams which ignore shareholder interests may seek to grow assets at the expense of profitability. Corporate governance activism reduces agency costs. Target firms reduce their assets by 4.54 - . activists regained their effectiveness. others want the firm to de-lever a precarious balance sheet.Capital structure activism increases cash to investors. Thus. I regress the changes in activist targets’ accounting measures on dummy variables for the Schedule 13D filing year.4% (t = 2. Activist hedge funds seem to take particular interest in reducing firm size when they distrust management. controlling for Tobin’s q and for the accounting variables’ initial levels. Nor do I find evidence that passive returns have changed over time.8). No other accounting measures provide any significant results. it hasn’t decreased. activists were very effective at reducing firm size. leverage actually falls by 1. Early in the decade.10). In addition. To summarize.21).0% over two years (t = -1. While I find that activists’ ability to reduce assets has changed over time (F = 2.

or fixing corporate governance. They improve profitability. By comparing activist investments to passive investments while controlling for ex-ante differences in target firms. and Klein and Zur (2008). activists revive struggling firms. Most notably. Each test strongly supports the hypothesis that activism adds value. Brav et al. in particular. selling target firms. optimizing capital structures. In particular.activist hedge funds improve target firms than do previous studies. advising on M&A transactions.55 - . it is also because I dissect these improvements by strategy.g. The new evidence presented in this paper expands the existing literature on activism. (2008). In contrast.46 While this is in part due to my unique sample. and I identify short-term events which contain only activist signals. I have made no adjustments for 46 See. and sell target firms. I study activist instruments. spin off assets. e. Activists also tailor their approach to particular firms’ needs. . Activism earns substantial short-term and long-term returns by unlocking shareholder value. I control for ex-ante differences in target firms. while prior studies have shown that activist hedge funds earn excess returns. prior studies looked at the results in aggregate—so different strategies canceled each other out. I present strong evidence they earn these returns by doing more than picking stocks.. Section VII: Conclusion This paper argues that activist outperformance represents more than stock picking in disguise. This paper has focused on the returns to corporate shareholders. sometimes at the expense of measuring the returns to hedge fund investors. focusing on improving business strategies. I use several unique tests to distinguish activism from passive investing—I compare activist investments to passive investments. Boyson and Mooradian (2008). increase payout. I isolate hedge fund activism from hedge fund investing.

. In particular. Further research should investigate whether activist hedge funds follow any predictable patterns that corporate executives might mimic. why can hedge funds unlock value beyond management’s ability even though hedge funds invest in many companies.56 - . pride. Perhaps corporate boards might increase their firms’ values by compensating their CEOs less like CEOs and more like hedge fund managers. Thus. Additionally. Further research might investigate whether activists earn greater returns in firms where managers’ interests are most poorly aligned with shareholders’ interests. capital structure. I show that agency costs detract from a firm’s value. In a broader context. this paper presents several lessons that might call corporate boards to activism themselves. and corporate governance. I also present evidence that expert advice can turn around struggling firms.hedge fund fees. while managers devote their whole lives to just one firm? Most likely. further research might study whether it similarly benefits hedge funds’ limited partners. the CEO benefits from other factors—like compensation. perhaps corporate boards should encourage hedge funds to invest. Given hedge funds’ ability to improve shareholders’ returns. this is due to agency costs: whereas the hedge fund activists benefit from stock returns. M&A. Perhaps undervalued companies should more frequently hire outside consultants to revisit their policies on capital allocation. and job security. while I have shown that activism benefits shareholders.

Rit is the return to security i. the premiums and discounts awarded to firms based on their market capitalization and ratio of book value of equity to market value of equity (B/M). and Rmt is return to the valued-weighted market portfolio in period t. To A1 . First. over the event window. I calculate the firm’s risk.Appendix A: Statistical Methods (A) Event Studies This essay uses the event study methodology. Companies with high B/M ratios often suffer low earnings yields on their assets. Both models assume that investments with greater risk offer greater reward. Rit = α i + β i Rmt + ε it E (ε it ) = 0 Var (ε it ) = σ ε i 2 In this model. β i measures the security’s correlation with the overall market. The market model assumes we can measure risk by a firm’s correlation with the overall market. It also assumes asset returns follow a normal distribution (MacKinlay 1997). Then. Small firms tend to be more risky and participate more frequently and for longer periods of time in earnings downturns. α i represents the idiosyncratic risk associated with a specific company. The error term ε it has an expected mean of zero. I compare observed returns to what they would have been without that event. I calculate these “counterfactual” returns using two popular models: the market model (Fama 1970) and the Fama French three factor model (Fama and French 1992). and the variance of ε it is assumed to be homoskedastic with a normal distribution. To determine how certain events affect security returns. I assume the firm performed like other firms with similar risk levels. Fama and French (1992) recommend augmenting the market model to include more possible risks—in particular. over a measurement window.

T-2 T-1 T0 T1 Estimation Window Event Window 1 See Puustinen (2007). Although these models are similar. T0 represents the event day. where SMB represents the difference in returns on portfolios of small and big market capitalization equities. First. They propose a three factor model: Rit = α i + β i Rmt + si SMBt + hi HMLt + υ it E (υ it ) = 0 Var (υ it ) = σ ε i 2 As before.1 Thus. I follow two steps. Fama and French suggest running time-series regressions to determine the premiums and discounts associated with market capitalizations and B/M ratios. I run times-series regressions over an estimation window prior to the event. T1 represents the end of the event window. In the diagram below. α i is the intercept term. and for monthly returns I run regressions over the 36 month window ending one month before the event. To test for robustness. I check my results against the Fama French three factor model. T-2 and T-1 represent the start and end of the estimation period. most activism studies use the simpler market model. and HML is the difference between returns of portfolios separated by B/M ratios (Fama and French 1992). Boyson and Mooradian (2008). for daily returns I run regressions over the 200 day window ending 10 days before the event. In this study. A2 . Rit is the return on security i and Rmt is the return on the value-weighted index at time t. (2008). respectively. and Klein and Zur (2008). To estimate the parameters. Brav et al. I report results using the market model. The two new terms represent the Fama French factors.incorporate these factors into the market model.

CAR( t1 . t 2 . where N is the number of events. t 2 ) = . A3 . To test returns over more than one day or one month. Under the null hypothesis. where 200 represents estimation period’s length. and . and Mamingi 1997). To test for statistical significance. Laplante. Abnormal returns simply equal actual returns minus expected returns . such that approaches zero (Dasgupta. where and . I measure abnormal returns. This equation holds when the estimation period length T is large. abnormal returns have conditional mean zero and conditional variance . I use the coefficient estimates from the time-series regression to estimate counterfactual returns over the event window. I calculate abnormal returns (CARs) over the event window. the paper will calculate the variance of the mean CAR . This paper tests whether the mean CARs among a portfolio of activist investments differ significantly from zero.Second. For any two times t1 .

First. I interpret as the average partial return to a target firm whose covariates equal the sample averages in which the activist hedge fund uses strategy j. For a passive target firm whose covariates equal sample averages. To do so. I express all non-dummy covariates as the deviation from the mean. A4 . I conclude that activism generates significantly greater returns than passive investing. if firm i’s covariates equal the sample averages. and the X ij ’s represent other covariates. An activist target firm whose covariates equal sample ^ greater returns. If γ is significantly greater than zero. For example. I use dummy variable regressions to analyze whether investments which I consider active enjoy greater returns or accounting outperformance than other investments. I use dummy variable regressions to measure the average partial returns on various binary variables—most frequently. returns equal averages earns . I use a regression of the form where Y i is the return to investment i.(B) Dummy Variable Regressions This paper also makes extensive use of dummy variable regressions to analyze binary variables like an activist’s strategy. binary variables for various activist strategies. I use a regression of the form where Y i and the X ij ’s have the same interpretations. Second. Because I have suppressed the constant term. I express all nondummy covariates as the deviation from the mean.

and if a hedge fund targets firm i with strategy 2 and strategy 3. and . then all the all the ’s equal zero. so we expect the return to be A5 . equal one. except and ’s equal zero.

energetic individual like Keith. The first section below presents a typical letter announcing that a hedge fund will seek board seats. My offer to Motorola stills stands. respectively. who has 145 million reasons to spend his time working toward the spin-off being accomplished. capital structure activism. It is true that Sandy Warner. SEC Schedule 13D. In a political election when constituents believe their representatives’ performance was inadequate. Investments Seeking Board Seats In a letter to Motorola. head of the Nominating Committee called me and offered seats to two of my Nominees if I would drop the proxy fight. Carl Icahn wrote1: You stated during today’s conference call. “we discussed Board Nominees with Carl Icahn and we proposed two nominees and he declined. If as you have stated. then what possible reason is there for not putting Keith Meister on the Board. After all.” Again this is only partially true. Warner then replied that the Board did not “know” Meister. We disagree. Warner what does one have to do to qualify — lose $37 billion dollars? Mr. how much can he eat at the Board meetings? On a positive side. we all want to benefit the stockholders of Motorola. Warner that I would gladly accept this offer if the Board would also accept Keith Meister. The next five sections describe business strategy activism.” I asked Mr. You have stated to the press that our request for information about what steps the Board actually took to correct the problem at Motorola is an unnecessary distraction. However. M&A activism.Appendix B: Excerpted Hedge Fund Letters To elaborate on the distinctions among activist strategies. may well make this promise come true in a timely fashion. Mr. having a highly intelligent.. filed 27 March 2008. Why should it be different in Corporate America? I do however agree with you that this proxy fight is a distraction that Motorola at this junction can ill afford. they are certainly not denied information as to whether their representative acted in a grossly negligent fashion. B1 . My answer was that Meister would fly anywhere at any time to meet the Board so they could “know” him (I did mention that the situation at Motorola is too serious for the Board to remain a country club). you failed to mention in your conference call that I told Mr. sale activism. Warner replied summarily to this offer that Meister did not “qualify. this section excerpts various hedge fund disclosures. We ask 1 Motorola Inc. and corporate governance activism.

Furthermore. This happens all the time and we would support such extra compensation for BAX executives in the event of a successful exit.. filed 20 April 2005. Brinks Co.. Elliot Associates wrote2: Elliott is puzzled by the pace of cost-cutting and restructuring actions taken by the Company over the past two years. Any concerns about losing key people during a sale process can be addressed with meaningful divestiture incentive bonuses. put egos aside and let’s get on with the urgent business at hand. profitable company fit to navigate the intense competition of the future.. whether that results in one or more transactions or a shut-down/liquidation. and add new independent members to the Board of Directors. will yield immense value to shareholders by creating a leaner. Whereas we believe a sense of urgency is appropriate. which includes a qualified investment banker with a board-approved mandate to approach qualified prospective buyers. we believe that by waiting until 2006 to divest BAX in the hope it will earn its cost of capital would be making a decision to trade the certainty of the currently robust M&A and freight markets for the uncertainty of BAX’s hypothetical ability to earn its cost of capital. are achieved via an organized process. M&A Activism In a letter to Brinks Co. if executed properly. Waiting for the buyers to come will likely result in no transaction. we believe. We suggest Pier 1 close down a significant number of stores.the Board meet with Meister. SEC Schedule 13D. We believe BCO is currently in a brief cyclical window of robustness in world freight markets and strong earnings for BAX. coupled with low-interest rates and a vibrant M&A market. 2 3 Pier 1 Imports Inc. It is imperative that the Company immediately focus on executing cost-cutting and restructuring initiatives in order to stem its current cash bleed. Business Strategy Activism In a letter to Pier 1 Imports. lease part or all of the corporate headquarters. the Company has reacted at a glacial pace to the intensified competitive landscape that has developed in the US home furnishing retail industry over the past five to ten years. The depressed market capitalization of BCO far outweighs any benefit of holding BAX. “right-size” the corporate and divisional SG&A. B2 . MMI Investments wrote3: We understand that BAX operations might be further improved. SEC Schedule 13D. but conclude that this is insufficient grounds to retain them. filed 9 April 2007. Most divestitures today. Elliott believes that these initiatives.

00 per share in cash. SEC Schedule 13D. 2000 from Newcastle Partners. moon and stars all align. Ikon Office Solutions. Inc. at minimum. we invite the Board to contact either of the undersigned to directly address such concerns in a productive manner. We would welcome the opportunity to discuss this or present further details to the Board. The Board's rejection of the offer demonstrates the Board's lack of dedication to pursuing the best interests of the stockholders of the Company and maximizing stockholder value. it appears that the Board spent their time preparing a press release attempting to belittle the ALL CASH offer. Instead of giving any consideration to the offer. Instead. The offer was summarily rejected with no indication that the Board consulted with an investment banking firm or relied on a "fairness opinion" to justify the rejection. SEC Schedule 13D. a phone call to the Interested Parties to address any questions or concerns of the Company's Board of Directors. Ltd. If there are concerns with any of the standard conditions contained in the offer. and a highly favorable environment in which to do so.. for $18. we were disappointed to see that you failed to return any of the phone calls made to you by the Interested Parties in connection with such offer.P. we believe that the offer warranted. as described in such letter. We also recognize and commend the 4 5 Gehl Co. B3 .It is rare in life or business that the sun. Newcastle Partners wrote4: We were surprised to learn by press release of the rejection by Gehl Company (the “Company) of the offer made in a letter to you dated December 22. the "Interested Parties") to acquire the Company. Sale Activism: In a letter to Gehl Co. Capital Structure Activism In a letter to Ikon Office Solutions. We conclude that it is crucial for the future of BCO that a buyer is sought promptly.. Steel Partners wrote5: We commend management for its efforts that have resulted in improved operating margins and a simplified capital structure. clearly indicates to us that the offer was not given serious consideration by the Board. L. and CIC Equity Partners. which rejection was announced in a press release only a few hours after it was delivered to you. The Board's rejection of the offer. We believe that BCO has such an occasion at this moment: a win-win transaction for exiting an undesirable subsidiary and funding a legacy liability. filed 20 March 2001. or in any way seek any additional information about the offer. Common sense aside. (collectively. filed 29 June 2007.75 per share. given that the offer represented a 67% premium to the Company's then current market price of $10..

SEC Schedule 13D. We therefore recommend that the Company capitalize on its operating momentum and utilize the strength of its balance sheet to pursue a public recapitalization at $17." One can only pity the poor student who suffers the indignity of attaching your name to his academic record. It would also be accretive immediately. the Company should be in a position to self tender for at least $850 million of common stock. With the Company's excess cash and a conservative debt financing package. in addition to lowering the Company's cost of capital and enhancing its EPS growth. in the course of our investigation. Importantly.) Irik. filed 14 February 2005. we believe this refinancing is highly achievable in a short period of time. fees for your cronies and to insulate you from the numerous lawsuits that you personally face due to your prior alleged fabrications. This represents a 20% premium to the 20-day closing average of the shares. we believe that the Company's common stock continues to trade below its intrinsic value and your balance sheet remains highly inefficient. Third Point Capital wrote6: Sadly. that at Cornell University there is an "Irik Sevin Scholarship. A review of your record reveals years of value destruction and strategic blunders which have led us to dub you one of the most dangerous and incompetent executives in America. Corporate Governance Activism In a letter to Steel Gas Partners. but is said with some authority. misstatements and broken promises. Based on our discussions with several leading investment banks. Steel would commit not to tender into this offer.Company for expanding the Company's existing share repurchase program. It seems that Star Gas can only serve as your personal "honey pot" from which to extract salary for yourself and family members. your ineptitude is not limited to your failure to communicate with bond and unit holders. B4 . I have known you personally for many years and thus what I am about to say may seem harsh.50 a share. and more so in the long term. The proposed recapitalization would provide immediate liquidity at a meaningful premium to the current share price for your shareholders who elect to tender. (I was amused to learn. As a long term shareholder. However. this recapitalization would result in a strong and flexible balance sheet that would support the Company's long term strategic vision. or approximately 39% of its outstanding shares at the aforementioned price per share. the junior subordinated units that you hold are completely out of the money and hold little potential for receiving any future value. We believe it will be well received by the shareholders who desire liquidity. at this point. It is time for you to step down from your role as CEO and director so that you can do what you do best: retreat to your waterfront mansion in the Hamptons where you can play tennis and hobnob with 6 Star Gas Partners LP.

B5 . The matter of repairing the mess you have created should be left to professional management and those that have an economic stake in the outcome.your fellow socialites.

For all hedge funds with at least three activist observations and three passive observations. activist investments outperformed. Turning first to short-term returns. it is worth investigating whether certain outlier funds unduly influence the aggregate outcomes. and (3) target firm improvements. Most Funds' Active Investments Exceed Passive Investments Results ot Driven By Outliers 40% Cannell 30% A ctive Exces s Return Ramius Icahn Elliot Shamrock FEF Stilwell Third Ave Riley SAC Pershing Southeastern 20% Newcastle Sandell JANA Third Point Harbinger 10% Greenlight Discovery DE Shaw Wynnefield Barington 0% -10% Nierenberg -20% -20% -10% 0% 10% 20% Passive Excess Return 30% 40% Figure C1. I analyze hedge fund heterogeneity in (1) short-term returns. This does not appear to be the case. For each hedge fund with at least three activist investments and three passive investments. from 30 days prior to each investment until 30 days after. Especially because 13D Monitor selectively chose the hedge funds for my sample.Appendix C: Hedge Fund Heterogeneity Because it requires substantial skill to improve a firm beyond the management team’s ability (or willingness). we might expect substantial variation hedge funds’ activist success. In this section. I look for outliers in any particular fund’s activist or passive investments. (2) long-term returns. I find no evidence that outliers have distorted my results. the chart plots the fund's activist return aginst its passive return. For funds above the 45° line. For all funds above In Short-Run. Figure C1 plots active returns against passive returns. C1 .

Most Funds' Active Investments Exceed Passive Investments Results ot Driven By Outliers 120% SCSF 100% 80% A ctive Excess Return Steel SAC 60% 40% 20% 0% -20% -40% -40% Barington Nierenberg Southeastern Wynnefield FEF JANA Ramius Elliot Riley Icahn Cannell ValueAct Third Point Discovery -20% 0% 20% 40% 60% Passive Excess Return 80% 100% 120% Figure C2. For 73. The greater the distance from the 45° line. If above the 45° line. In C2 . from one month before each investment until 24 months after. the more the activist investments outperformed (or underperformed).In Long-Run. I use the same method to next analyze long-term returns. the short-term outliers have disappeared. activist investments outperformed. Coupled with the strong median outperformance. active investments outperformed passive investments. this adds comfort that my prior results were not skewed by outliers. Further. For each hedge fund with at least three activist investments and three passive investments. Figure C2 plots hedge funds’ active excess returns against passive excess return. the 45° line. Cannell Capital’s high flying short-run returns have fallen back to earth. and both Nierenberg and SAC migrate above the 45° line as their activist investments outperform their passive investments over two years. from one month prior to the investment until 24 months after.9% of the 23 funds. For funds above the 45° line. the chart plots the fund's activist return aginst its passive return. I find few outliers. Interestingly. a fund’s activist investments outperformed its passive investments. active investments outperform passive investments.

fact, for 76.5% of the 17 funds, activist investments outperformed passive investments over the period. This median outperformance adds more comfort that my prior results were not skewed by outliers. While SCSF appears to be a positive outlier, its 109.4% activist return is driven by only four investments. Thus, few outliers coupled with strong median outperformance give no reason to question the earlier results. Finally, I turn to heterogeneity in target firm improvements. Figure A3 repeats the above analysis for changes in ROA (Panel A), leverage (Panel B), and assets (Panel C). Due to a scarcity of observations, I analyze the change in ROA for any fund with at least two activist and two passive observations, but I analyze the change in leverage and the change in assets for any fund with at least three activist and three passive observations. Turning first to profitability, as expected, in 63.6% of hedge funds, their activist targets improve ROA, and in 90.9% of investments, their passive targets reduce ROA. For 81.8% of hedge funds, activist target firms’ ROA increase more than passive target firms’ ROA. While firms above the line show much more variance than those below the line, the distribution gives no reason to suspect that outlier funds have skewed the results, especially given the dramatic median outperformance. Leverage shows a similar pattern. As expected, in 57.1% of hedge funds, their activist targets increase leverage, and in 57.1% of hedge funds, their passive targets reduce leverage. For 71.4% of hedge funds, activist target firms’ leverage increases more than passive target firms’ leverage. Both due to the strong median performance, and because there appear to be no clear outliers, I find no reason to reject the long-term results discussed earlier. Finally, analyzing the change in assets lends further strength to my earlier conclusions. As expected, activist investments grow more slowly than passive investments; for 69.6% of hedge funds, activist targets reduce total assets, and for 65.2% of hedge funds, their passive

C3

Panel A: 82% of Hedge Funds Improve ROA
4%
Steel

2%
JANA

ValueAct Third Point

Ramius Elliot

Southeastern

Active ROA Increase

0%
Greenlight

Wynnefield

-2%

Icahn Farallon

-4%

-6%

-8% -8%

-6%

-4%

-2% 0% Passive ROA Increase

2%

4%

Panel B: 71% of Hedge Funds Increase Leverage
4%
Wynnefield Steel ValueAct

3% Active Leverage Increase

2%
Ramius

1%
Icahn

0%

Ramius

JANA

-1%

-2% -2%

-1%

0% 1% 2% Passive Leverage Increase

3%

4%

C4

Panel C: 74% of Hedge Funds Reduce Assets
20%

Greenlight

10% Active Asset Growth

Nierenberg SAC ValueAct Third Point

0%

Cannell Barington SCSF Ramius Wynnefield

JANA

Breeden Icahn

Stilwell Southeastern Highland Riley Discovery

FEF

Harbinger

Elliot

-10%

Steel Farallon

-20% -20%

-10%

0% Passive Asset Growth

10%

20%

Figure C3. For various hedge funds, the chart plots changes in ROA, leverage, and assets for activist and passive investments in Panel A, Panel B, and Panel C, respectively. Panel A includes all hedge funds with at least two activist and passive observations. Panel B and Panel C, which had more observations, include all hedge funds with at least three activist and passive observations.

targets increase total assets. In a sample of 23 hedge funds, in 73.9% of funds, activist targets shrink more than passive targets. Thus, I find strong evidence to support the earlier conclusions that hedge funds reduce assets in their activist investments. To give more details about the funds comprising my sample, Table C-1 describes each fund’s investments by strategy and by instrument. To summarize, despite using a sample comprised of hand-picked hedge funds, I find no evidence that any particular hedge funds have skewed my results. To the contrary, examining returns and improvements on a fund-by-fund level has only strengthened my earlier conclusions.

C5

0% 0.1% 20.0% 16.0% 0.3% 11.7% 13.7% 9.0% 0.3% 14.5% 53.0% 0.0% 13.1% 33.0% 0.9% 0.7% 0.4% 16.0% 40.3% 41.7% 18.7% 27.5% 23.3% 18.2% 50.7% 16.5% 20.3% 0.7% 10.2% 25.1% 23.6% 11.9% 16.7% 39.0% 11.6% 10.0% 0.7% 13.8% 13.6% 52.0% 20.0% 30.9% 0.7% 7.0% 0.4% 8.0% 0.0% 0.3% 0.0% 56.3% 12.7% 8.1% 5.3% 71.7% 2.0% 16.2% 17.0% 20.1% 11.0% 0.0% 7.7% 20. and what percentage of that fund's investments fall into each instrument category.3% 12.0% 9.2% 22.0% 35.0% 17.0% 15.9% 24.7% 33.3% 16.7% 0.7% 40.0% 21.0% 5.3% 16.2% 70.0% 18.0% 0.7% 33.3% 0.3% 33.0% 0.4% 38.0% 9.7% 66.0% 17.2% 25.4% 4.0% 0.0% 0.7% 9.0% 80.4% 0.3% 14.3% 33.0% Seek Board 37.0% 0.0% 22.7% 13.2% 22.5% 34.9% 28.5% 45.3% 2.7% 0.0% 21.3% 16.3% 0.4% 8.8% 56.8% 4.6% 27.4% 37.1% 20.1% 38.0% 44.0% 21.9% 10.0% 10.0% 0.0% 0.5% 5.0% 27.7% 8.0% 40.3% 50.7% 66.0% 0.0% 7.7% 4.4% 11.Table C-1 Hedge Fund Sample This table describes the hedge funds in my sample.0% 40.2% 100.0% 0.1% 10.3% 0.0% 33.2% 5.0% 0.5% 20.0% 0.0% 0.3% 0.7% Win Board 23.0% 6.0% 33.4% 4.0% 42.0% 0.5% 14.7% 5.7% 8.0% 0.0% 22.1% 5.0% 13.0% 17.0% 33.3% 14.5% 35.5% 30.0% 0.5% 2.0% 33.0% 16.9% 2.0% 55.0% 10.3% 66.7% 8.7% 62.0% 12.0% 44.0% 60.0% 11.3% 0.0% 0.9% 2.1% 30.0% 0.7% 0.0% 12.0% 15.1% 16.0% 0.8% 12.8% 33.1% 13.1% 11.0% 46.4% Any Instrument 37.3% 47.4% 21.1% 7.1% 22.3% 8.0% 8.0% 18.0% 18.0% 25.3% 11.1% 0.0% 33.6% 29.2% 5.0% 0.7% 0.0% 5.3% 70.7% 0.3% 5.2% 33.0% 12.3% 6.2% 0.0% 26.4% 0.0% 42.7% 42.0% 13.6% 0.9% 38.0% 0.1% 0.4% 4.3% 51.8% Any Strategy 33.1% 0.8% 13.0% 22.3% 16.0% 13.0% 0.8% 11.1% 25.1% 0.7% 0.4% 7.8% 4.0% 11.7% 0.0% 0.0% 0.9% 4.1% 20. what percentage of that fund's investments fall into each strategy category.0% 7.3% 0.5% 2.3% 38.0% 11.0% 8.5% 14.0% 50.6% 22.2% 15.3% 55.0% 37.4% 0.5% 25.6% 71.8% 88.4% 11.0% 8.3% 21.1% 4.3% 25.0% 0.3% 0.0% 14.1% 0.2% 29.1% 0.8% Lawsuit 3.6% 60.0% 11.4% 7.0% 8.0% 16.0% 0.7% 8.8% 21.0% 0.2% 66.3% 15.0% 0.0% 0.0% 0.0% 0.7% 3.4% 15.4% 45.3% 14.9% 14.0% 0.3% 22.5% 50.0% 33.0% 14.0% 40.7% 27.8% 24.0% 7.2% 70.9% 29.0% 5.8% 47.7% 29.0% 8.0% 0.4% 0.0% 44.0% 0.0% 6.4% 16.3% 13.0% 60.2% 0.9% 35.6% 33.5% 55.7% 0.4% 25.2% 100.8% 0.4% 37.1% 0.7% 20.5% 28.3% 6.0% 4.4% Activist Strategies Capital Corp Sale Structure Governance 19.3% 25.9% 20.0% 0.1% 0.5% 14.1% 0.7% 2.4% 24.6% 0.7% 22.0% 14.0% 0.0% 26.1% 16.2% 2.1% 17.0% 11.4% 56.1% 33.3% 57.2% 40.0% 0.7% 20.0% 9.0% 11.3% 0.1% 0.6% 20.0% 42.0% 80.0% 14.0% 5.3% 33.5% 7.0% 22.0% 40.5% 45.1% 30.7% Activist Instruments Proxy Buyout Fight 9.0% 16.8% 10.7% 71.0% 0. the table shows how many filings each fund made over the time period.0% 13.1% 44. Total Filings Steel Partners ValueAct Capital Ramius Capital Carl Icahn Third Point Farallon Wynnefield Capital Harbinger Capital Riley Invetment Mgmt Financial Edge Fund Blum Capital Barington Capital Canell Capital JANA Partners SCSF Equities Shamrock Activist Fund Discovery Group Elliot Associates Joseph Stilwell Newcastle Partners Southeastern Asset Mgmt MMI Investments SAC Capital Highland Capital Nierenberg Investment Third Avenue Mgmt DE Shaw Sandell Asset Mgmt Clinton Group Costa Brava David M Knott Greenlight Capital Cresecendo Breeden Capital Pershing Square Blue Harbor New Mountain Vantage Relational Investors Total 51 49 39 37 37 35 29 24 24 23 22 20 20 18 18 17 15 15 15 14 14 13 13 12 12 12 11 10 9 9 9 9 8 7 7 6 5 5 718 Business Strategy 3.7% 0.0% 17.0% 0.0% 20.9% 26.7% 4.8% 55.0% 77.2% 46.0% 22.4% 0.3% 53.9% 6.3% 22.0% 3.0% 0.0% 0.2% 0.0% 0.4% 20.0% 0.0% 14.5% 16.0% 5.1% 0. For each fund.3% 100.1% 40.7% 16.1% 54.0% 16.4% 59.0% 33.0% 60.0% 0.1% 21.0% 0.0% 5.3% 29.0% 0.3% 37.0% 29.1% 7.5% 18.1% 24.3% 70.5% 50.3% 0.2% 5.0% .0% 76.9% 14.4% 22.0% 0.

Appendix D: Unreported Regressions and Additional Analysis .

1 0. Tobin's q is the ratio of market value of equity plus long-term debt over book-value of equity plus long-term debt.55 1.2 Median 329.5 -0.3 19.3 -0. Column 3 and Column 6 report the differences in means and associated t -statistics.4 22.6 14.6 3. Growth is the firm's one-year sales growth.9 8.2 9.2 2.8 19. medians.4 22. and standard deviations.1 0.1 20. All data is winsorized at the 10% level.0 2.4 18.2 15.7 Mean 102. Return on Assets is the firm's EBITDA divided by its assets.5 7.0 7.3 3.22 0.36 0.3 13.40 1.4 9.6 -16.0 8.7 22.6 9.7 -0.9 1.9 8.1 1.0 10. Capex/Assets is the firm's capital expenditures divided by total assets.8 -0.9 1.2 -1.1 -0.8 0.0 -2.7 0. Passive Targets (1) Active Mean Market Cap Tobin's q Abnormal q P/E Ratio Growth Margin Abnormal Margin Return on Assets Return on Equity Abnormal ROE Leverage Abnormal Leverage Payout Ratio Dividend Yield Cash/Assets Capex/Assets R&D/Assets 757. Abnormal Margin is the firm's margin less its industry median margin. Abnormal ROE is the firm's ROE less its industry median ROE.8 -0.7 -0.9 0.0 0.5 StDev 814.8 0.8 0.9 81.49 0.5 1.6 Mean 654.89 1.8 16.8 11.6 0.5 16.0 0.9 1. All variables refer to the trailing twelve months prior to the Schedule 13D and are from Bloomberg or Compustat.7 20.84 1.4 9.9 StDev 888.5 -0.68 1.1 3.5 4.8 18.4 1.8 1.00 0. and 1% levels. Dividend Yield is the firm's total dividends divided by market capitalizatoin.0 17.47 3.1 -29.1 1. Panel B separates activist investments which seek board seats from activist investments which do not.0 6. Cash/Assets is the firm's cash and cash equivalents divded by total assets.3 * * *** * * ** * ** (3) Difference t -stat 1.7 -0.9 18.3 19. Margin is the firm's EBITDA divided by sales. Column 1.0 6.4 16.9 10. Leverage is the firm's debt-to-assets ratio.4 -4.3 4. Payout Ratio is the firm's common dividends divided by net income.1 15.5 -1.0 7.8 9.6 34. and Column 5 report means.3 1.7 12. and *** refer to statistical significance at the 10%. Market Cap is market capitalization. Abnormal Leverage is the firm's leverage less its industry median.23 1. Panel A: Active vs.9 (2) Passive Median 270.34 0.8 3.71 1.3 2.63 1.2 34. Column 2.8 2.1 9.9 16.0 16.0 1.8 -0. **.1 0.1 16.3 -34.1 12.30 * * .0 0.5 16.7 8.3 8.9 8.8 1. and R&D/Assets is the firm's R&D expense divided by total assets. *. Abnormal q is the target firm's Tobin's q as a percent of its industry's median Tobin's q . Column 4. 5%.0 0. Return on Equity is the firm's EBITDA divided by the book value of equity.7 1.9 19.2 15.9 3.49 2.6 15.2 -1.0 0.1 0.42 0.4 -1.8 74.5 0. P/E Ratio is the firm's trailing twelve months price-to-earnings ratio.1 1.5 20. Panel A separates activist investments from passive investments.Table D-1 Characteristics of Target Companies The table reports the characteristics of hedge funds' targets.

3 (5) No Board Seats Median 472.0 -0.13 3.5 5.54 4.0 0.7 -11.5 16.0 Median 253.9 17.0 0.1 8.5 -22.0 -5.54 2.3 0.0 1.8 16.10 1.0 6.47 1.6 -0.7 7.10 0.7 72.8 16.4 17.0 StDev 926.07 2.59 0.0 0.4 0.0 -0.8 7.7 17.9 -1.0 -5.0 0.6 16.7 0.2 6.6 7.3 0.8 6.0 0.3 15.25 0.3 *** *** (6) Difference t -stat 2.4 18.4 -0.7 0.1 -4.8 11.95 *** *** *** *** *** ** .7 17.8 1.5 23.5 17.8 1.1 17.6 8.0 25. Active Investments Which Do Not (4) Board Seats Mean Market Cap Tobin's q Abnormal q P/E Ratio Growth Margin Abnormal Margin Return on Assets Return on Equity Abnormal ROE Leverage Abnormal Leverage Payout Ratio Dividend Yield Cash/Assets Capex/Assets R&D/Assets 636.5 3.9 -5.0 1.9 10.5 1.1 0.2 1.5 0.7 -2.6 16.0 0.0 17.3 1.0 17.21 0.6 16.3 -2.2 19.2 -11.66 0.0 -1.7 9.85 0.9 7.4 8.8 36.6 15.9 -0.2 19.8 StDev 837.0 9.49 2.4 0.5 8.5 13.3 20.1 5.6 Mean -255.5 18.1 6.6 18.8 0.93 2.2 24.4 4.1 4.0 0.1 -6.4 1.1 -0.7 12.5 2.8 10.8 76.6 Mean 891.0 9.0 6.3 1.6 5.5 13.9 14.3 0.2 1.4 33.9 1.5 -29.8 2.0 -4.35 0.Panel B: Active Investments Seeking Board Seats vs.44 2.2 1.2 -42.8 2.7 -0.

8% 0.66 7.6% 11.0% *** 9.7% *** 79. z-statistic N= -13.1% *** 10.75 194 (-30. *.+30) -12.1% 31. **.6% 5.90 5.99 149 (+0.4% 53.50 8.+2) -5.8% 5.+30) 5% 25% Median 75% 95% Average % Positive Patel z-statistic Non par.Table D-2 Short-Term Returns by Investment Type The table presents average daily excess returns around the Schedule 13D filing for various event windows. Panel A: Active Investments Perform Better Than Passive Investments Active (-30.4% 10.6% -3.98 3.9% 5.+2) -3.0% *** 15.9% *** 9.0% *** 2.+30) -33. I report p -values for medians using the Wilcoxon Signed Rank Test.1% 2.7% 0.89 5.7% 21.+30) (-2.40 4.7% 1.0% -0.4% *** 10.3% -10.3% 5.98 45 (+0.2% *** 14.2% *** 4.2% * .4% *** 9.7% 2.+2) Difference (+0.8% 25.3% *** 23.2% -2.+2) -6.1% 4.1% * 2.8% 6.6% 19.+30) 8.8% 4.9% *** 3.09 45 (+0.3% 7.7% *** 67.+30) (-2.9% 4.6% *** 71.00 6.0% -0.5% *** 67.5% 0.8% *** 25.91 194 (-2.+2) -5.+2) (+0.1% 0.41 3.+30) 1. and *** indicate statistical significance at the 10%.+2) Difference (+0.3% 35.7% *** 80.5% *** 1.7% *** 57.4% 51.3% 9.16 3.2% -0.77 8.9% 1.2% *** 2.7% 38. and 1% levels.8% *** 24.70 149 (-30.+30) -24.6% *** 77.+2) -7.+2) -7.+30) -23.7% *** 73.6% 5.00 3.6% 5.6% *** 57.6% 40.76 194 (+0.9% *** 5.1% *** 8.9% *** 16.3% *** 4.6% *** 19.7% *** 88.3% *** 3.8% *** 64.+30) -13.2% 8.8% *** 4.6% 21.1% *** 7.9% * 3.15 149 No Board Seats (-2.3% 2.5% *** 29. estimating parameters over the 200 day window prior to the Schedule 13D.+30) -8.6% 17.7% * 2. Panel A compares active investments to passive investments.1% 7.8% 59.5% 1.8% *** 16.9% *** 80.2% -5.76 340 (+0.6% *** 16.+2) -3.48 45 (-2.1% *** 8.+30) 5% 25% Median 75% 95% Average % Positive Patel z-statistic Non par.9% 1.6% 11.5% 0.5% *** 69. Panel B further splits active investments into those which seek board seats and those which do not.6% *** 75.2% 44.0% *** 14.1% 2.81 149 (+0.6% *** 4.9% 10. 5%.8% 4.16 7.26 340 (-30.04 3. z-statistic N= -21.6% *** 77.8% 7.9% *** 7.0% 16.4% *** 10.61 194 (+0.1% ** 4.0% *** 11.04 7.4% *** 3.+2) -7.5% *** 13.5% 30.1% 12.8% *** 9.7% -1.7% 7.9% 4.4% 4.8% *** 7.2% 12.5% 12.95 340 Passive (+0.37 342 (-2. I measure returns in excess of the market model using the value-weighted CRSP NYSE/Amex/Nasdaq index.39 45 (-30.7% 3.8% *** Panel B: Active Investments Seeking Board Seats Perform Better Than Active Investments That Do Not Board Seats (-30.3% 0.9% *** 6.0% 29.27 6.82 5.8% *** 17.6% *** 77.6% *** 8.1% 3.+2) (+0.7% * 3.

38 *** -2.56 *** 3.93 ** -1.97 * -1. All nondummy variables are winsorized at the 10% level and are expressed as the deviation from the sample average values.03 -0.75 *** 4. **. 3.06 (4) Board Seats Coeff. and *** indicate statistical significance at the 10%.13 -0.58 3.55 *** 6.24 122 0.45 -2. t -stat 5.45 1. Column 2 and Column 4 regress abnormal returns from exclusively active investments on Board .23 2. Column 1 and Column 3 regress all abnormal returns on Active .69 -2.Table D-3 Short-Term Returns By Type.14 -0. and R Coeff. Panel A controls for abnormal q and abnormal margin and Panel B controls for Tobin's q and margin. t -stat 2.46 -0.09 ** -1. and R Activism Coeff. a dummy set to one for board investments.45 *** 5.02 -0.37 0.96 -0.15 -0. a dummy set to one for activist investments. t -stat 5.09 3.65 *** 6. Controlling For Undervaluatoin The dependent variable is the abnormal return during the 5 days around the Schedule 13D filing date.15 290 Activism t -stat *** 5. 5%. and 1% levels.15 ** 1.07 -1.07 ** -2.23 ** -2.78 441 0.41 3. *. Panel A: Industry-Adjusted Variables (1) Dependent Variable: Abnormal Returns Constant Abnormal q Abnormal Margin Active Board 2 No.07 . All dummies refer only to information contained in the initial schedule 13D filing.11 Panel B: Non-Adjusted Variables (3) Dependent Variable: Abnormal Returns Constant Tobin's q Margin Active Board 2 No.71 162 0.07 (2) Board Seats Coeff.

80 3.5% *** 20.82 110 (3) First Board Win (-2.76 53 (-2.5% *** 2.7% 4.+30) -23.5% -2.4% -4. estimating parameters over the 200 day window prior to the event. for various event windows.7% -3.1% 10.38 3. Column 2 analyzes schedule 13D/A filings or news stories in which.+30) 5% 25% Median 75% 95% Average % Positive Patel z-statistic Non par.2% 2.2% 0.3% 2.81 3.2% *** 66. 5%.0% 26.6% 0.1% *** 4.2% 0.7% 13. and *** indicate statistical significance at the 10%.5% *** 77. (1) First Action (-30.7% 25.2% 5.9% 6.2% *** 20. having already announced its intention to seek board seats.66 53 (+0.63 1.5% *** 51.2% -2.1% *** 69.21 2.5% *** 11.2% 1.6% * 10.3% 0.3% 2.5% 2. a hedge fund first seeks board seats. the hedge fund first wins board seats.4% *** 3.6% *** 3.95 4.7% 21.3% 20.41 0.4% *** 77.6% *** 17.69 1. in excess of the value-weighted market.+2) -3.70 48 (2) First Board Seat (-2. Column 3 analyzes schedule 13D/A filings or news stories in which.1% -0.71 4.49 53 (+0.0% 46.+2) -9.05 109 (+0.2% 1.9% -2.+2) -3.8% 2.3% *** 71.01 48 (+0.1% -6.3% 8.+30) -29.6% 3.3% 7.8% 2.3% 6.8% * 53.75 2.+2) -5.4% ** 2.0% *** 5.Table D-4 Abnormal Returns After 13D Filing Date The table presents average returns after the initial Schedule 13D filing.7% *** 8.9% *** 60.4% 41.3% *** 66. I measure returns in excess of market model using the value-weighted CRSP NYSE/Amex/Nasdaq index.6% 8.21 53 (-30.+30) -18.9% 7.7% *** 3.3% *** 18.4% -1. **. I report p -values for medians using the Wilcoxon Signed Rank Test.3% *** 73.5% *** 6.34 3. z-statistic N= -30.9% 5.6% *** 4.5% 2.83 2.76 110 .+2) -2. having already announced activist intentions.6% 0.8% ** 61.5% 5.56 48 (-30.8% *** 9.0% 0.+2) -7.01 48 (+0.1% 34.5% -1.7% *** 2.3% 4.8% 0.1% *** 4.78 109 (+0.+30) -13.1% 5.2% *** 10.+30) -12. a hedge fund first announces activist intentions.3% 6.7% 6. *. Column 1 analyzes schedule 13D/A filings in which.8% *** 4. and 1% levels.4% *** 66.7% 35.2% -5. having already filed a passive 13D.

+6) (+1.9% * 5.5% 11.03 121 (+1.2% *** 60.3% 21.5% 6.29 0.3% 5.84 1.+24) -153.8% * 17. and *** indicate statistical significance at the 10%.87 112 (-1.2% 47.11 112 (+1.3% 45.2% -0.1% 7.2% 3.+24) -132.3% *** 66.+24) (-1.9% -11.5% -11.3% -31.06 119 (-1. I report p -values for medians using the Wilcoxon Signed Rank Test. Panel A compares active investments to passive investments.7% -127.+24) -75. *. z-statistic N= -91.7% 4.2% *** 7.8% -24. and 1% levels.08 2.14 119 (+1.2% -18.5% *** 61.6% *** 66.+6) (+1.0% -8.1% 64.16 16.+24) Difference (+1.6% ** 5.2% *** 4.+6) -37.2% 125.39 238 (+1.58 239 13.8% -0.45 115 No Board Seats (+1.8% 1.32 118 (+1.5% *** 56.+12) -53.7% 31.6% 175.6% 18.+12) (+1.57 0. **.5% 177. Panel A: Active Investments Perform Better Than Paassive Investments Active (-1.1% 20.1% 13.8% ** 55.1% 149.5% 53.0% -0.+12) (+1.5% -8.89 1.88 2.2% * 36.+24) Difference (+1.9% *** 3.1% 9.87 239 9.70 244 (+1.48 217 (+1.0% -10.6% 40.67 2.4% *** 5.2% 21.7% *** 0.2% -0.0% 7.6% 50.Table D-5 Long-Term Returns by Investment Type The table presents average monthly excess returns around the Schedule 13D filing for various event windows.48 6.7% 150.0% ** 3.5% -0.4% ** 8.4% 20.+24) 5% 25% Median 75% 95% Average % Positive Patel z-statistic Non par.9% *** 56.4% * 49.+24) -144.5% 1.2% 3.3% *** 38. z-statistic N= -132.+24) 5% 25% Median 75% 95% Average % Positive Patel z-statistic Non par.5% *** 62.15 4.2% ** 4.3% *** 60.4% 58. 5%.4% * 1.8% *** 8.8% -13. Panel B further splits active investments into those which seek board seats and those which do not.9% * Panel B: Active Investments Seeking Board Seats Perform Better Than Active Investments That Do Not Board Seats (-1.+24) -154.2% *** 19.3% 7.7% * 13.1% *** 57.2% * 10.6% 4.1% *** 22.4% *** 12.8% 6.+12) (+1.8% ** 16.3% 10.3% *** 58.7% 15.+6) -48.9% 59.8% * .4% 8.19 3.+12) -106.2% -12. estimating parameters over the 36 month window prior to the Schedule 13D.8% 14.7% -20.4% 49.3% *** 2.8% -20.7% 55.3% 8.6% 11.5% 75.6% -19.8% *** 58.9% ** 11.6% 4.9% *** 4.7% 7.29 218 (-1.0% 19.6% 95.9% 0. I measure returns in excess of the market model using the value-weighted CRSP NYSE/Amex/Nasdaq index.43 5.54 216 Passive (+1.+6) -55.1% ** 1.4% -24.19 1.0% *** 56.5% ** 2.97 2.0% 2.3% 54.+24) -77.1% 30.1% *** 5.8% ** 54.6% 2.5% 111.1% 2.+24) 15.+12) -89.87 2.2% *** 75.2% 28.3% 159.+6) -58.3% 172.3% *** 53.5% -30.55 3.08 227 (+1.1% *** 38.30 112 (+1.1% 0.4% 116.9% 114.

04 .17 0.Table D-6 Long Term Returns By Type.63 * 197 (4) Board Seats t -stat 1.01 ** 4.32 251 t -stat 0.35 ** 388 t -stat 0.03 Coeff.33 ** -11.01 2.80 0. All nondummy variables are winsorized at the 10% level and are expressed as the deviation from the sample average values.24 0.01 1.36 -2.11 2. a dummy set to one for activist investments.67 -9. **.03 Panel B: Non-Adjusted Variables (3) Dependent Variable: Abnormal Returns Constant Tobin's q Margin Active Board 2 No. and 1% levels. 6.79 18.59 -1.90 -1. *. 7.04 Coeff.90 ** 18. Controlling For Undervaluatoin The dependent variable is the abnormal return during the 25 months around the Schedule 13D filing date.75 -1.45 150 (2) Board Seats t -stat 1.87 -0.00 -1.52 -1. 15. Panel A controls for abnormal q and abnormal margin and Panel B controls for Tobin's q and margin.07 * -0. and R Activism Coeff. Panel A: Industry-Adjusted Variables (1) Dependent Variable: Abnormal Returns Constant Abnormal q Abnormal Margin Active Board 2 No.21 * -1. Column 1 and Column 3 regress all abnormal returns on Active .40 ** -0.60 -7. 4.86 -1.81 -14. 5%.45 0.50 * -0. and R Activism Coeff.59 0. Column 2 and Column 4 regress abnormal returns from exclusively active investments on Board . and *** indicate statistical significance at the 10%. a dummy set to one for board investments.25 1.

All regressions control for Tobin's q and for the initial ROA (Panel A) or EBITDA margins (Panel B).15 ** -1.Table D-7 Profitability Improvements. t -stat -2.70 1.16 Coeff.51 -0.05 0. **.23 0.55 -1.73 -0.20 *** -2. and R Activism Coeff. and R Activism Coeff.24 -0.16 38 0.03 *** 3. t -stat -1.53 *** -3.34 0.46 0. Column 2 and Column 4 regress improvements for activist targets on Board . All nondummy variables are winsorized at the 10% level and are expressed as the deviation from the sample average values.32 2.20 1.87 *** 2.82 40 (2) Board Seats t -stat 0.12 ** -2. and *** indicate statistical significance at the 10%.23 0.16 0. 5%.89 1.16 3. a dummy equal to one for board seat investments. and 1% levels. Controlling For Stock Picking The dependent variable is the two-year increase in ROA (Panel A) and in EBITDA margins (Panel B).13 (4) Board Seats Coeff. a dummy equal to one for activist investments.05 108 0. *. 1.24 ** 0.71 0.31 *** -4.06 -0. t -stat -0.26 Panel B: Margins (3) Dependent Variable: Abnormal Returns Constant Tobin's q Initial Level Active Board 2 No. Column 1 and Column 3 regress improvements for all firms on Active .76 0.76 105 0.10 . Panel A: Return On Assets (1) Dependent Variable: Abnormal Returns Constant Tobin's q Initial Level Active Board 2 No.24 -0.61 2.

43 * -0. and 1% levels. Column 1 and Column 3 regress improvements for all firms on Active .04 Coeff. Panel A: Payout (1) Dependent Variable: Abnormal Returns Constant Tobin's q Initial Level Cash Active Board 2 No.05 0.79 -0.86 0.08 ** 92 (4) Board Seats t -stat 1.56 0.02 1.26 0.06 . Column 2 and Column 4 regress improvements for activist targets on Board . and the initial cash divded by initial assets.36 0.08 -0. All regressions control for Tobin's q. 4.50 1.07 -2.Table D-8 Capital Structure Changes. and *** indicate statistical significance at the 10%.12 ** 10.05 *** -2. the initial payout ratio (Panel A) or leverage (Panel B). 1. 2.20 -1. 5%.16 * 10.30 -0.37 -0.21 1.78 -1. Controlling For Stock Picking The dependent variable is the two-year increase in payout ratio (Panel A) and in the debt-to-assets ratio (Panel B). and R Activism Coeff.48 102 t -stat 1.39 0. All nondummy variables are winsorized at the 10% level and are expressed as the deviation from the sample average values.70 0. *.24 * -0. t -stat -0.25 ** -1.05 ** -4.75 42 (2) Board Seats t -stat 1.98 222 0.07 Coeff. a dummy equal to one for board seat investments.83 0.68 -3.59 -0.11 -1.53 0. a dummy equal to one for activist investments.45 * 2.54 * -1.65 0.19 0. and R Activism Coeff. **.99 2.11 -0.07 Panel B: Leverage (3) Dependent Variable: Abnormal Returns Constant Tobin's q Initial Level Cash Active Board 2 No.83 0.

09 Panel B: Sales (3) Dependent Variable: Abnormal Returns Constant Tobin's q Initial Level Active Board 2 No. All nondummy variables are winsorized at the 10% level and are expressed as the deviation from the sample average values.85 0.90 *** 3.63 494 0. t -stat 0.49 256 0. t -stat 4.34 5. a dummy equal to one for board seat investments. t -stat 4.19 -0. **. a dummy equal to one for activist investments.30 244 0.66 0.00 ** -1. *.23 -1.00 0.Table D-9 Firm Size Reductions.06 (2) Board Seats Coeff.75 *** 6.67 5. Panel A: Assets (1) Dependent Variable: Abnormal Returns Constant Tobin's q Initial Level Active Board 2 No.77 ** 1.79 -0.03 6.89 -4. and 1% levels.56 ** -2.75 0. Column 1 and Column 3 regress improvements for all firms on Active .15 0.03 -4. 5%. and *** indicate statistical significance at the 10%.61 0.15 ** 2. and R Activism Coeff. Column 2 and Column 4 regress improvements for activist targets on Board . and R Activism Coeff.61 *** 4.82 -0.06 . Controlling For Undervaluatoin The dependent variable is the two-year increase in assets (Panel A) and in sales (Panel B). t -stat 2.29 *** 3.05 (4) Board Seats Coeff. All regressions control for Tobin's q and for initial assets (Panel A) or initial sales (Panel B).95 *** 3.00 0.12 522 0.00 -1.42 * -1.81 8.

and *** indicate statistical significance at the 10%.21 0.6 64 .58 -1. The data is winsorized at the 10% level.32 0.8 1. considering only the sample of investments which announce activist strategies.9 *** -0.11 1. the constants are supressed.81 -0.8 10. The data is winsorized at the 10% level. The regressions control for the initial payout ratio or the leverage as well as the initial cash divided by assets. Return on Assets Coeff Initial Value Business Strategy M&A Advice Sale Capital Structure Corporate Governance Other No.1 * -6.11 1.1 * -1.0 -1.86 0. Due to the full span of the dummy variables.4 ** 5.0 26 t -stat -2.2 1.22 0.70 2.0 21. and R 2 -0.39 -2.7 2. **. and *** indicate statistical significance at the 10%.11 1.81 -0. and 1% levels.3 0. 5%.3 2.50 EBITDA Margin Coeff -0.78 0.3 -0.47 -0.4 *** 6.7 -1.6 -3.11 0.0 ** 2.42 -0.81 2.90 -1.87 0.Table D-10 Profitability Improvement By Strategy The table regresses the two-year change in (1) Return on Assets and (2) EBITDA margin on dummy variables for the various activist strategies.4 -0.91 -0.6 -2.29 Coeff Leverage t -stat -1. **.17 1. Due to the full span of the dummy variables.71 -0. considering only the sample of investments which announce activist strategies.11 1.22 -0. *.1 1.13 0. and 1% levels.73 1.62 0.2 ** 2. and R 2 0. The regressions control for the initial ROA or the initial margin.15 0.8 * -10. 5%.4 ** 3. the constants are supressed.4 1.4 0.5 29 t -stat -2. *.35 -1.46 Table D-11 Capital Structure Changes By Strategy The table regresses the two-year change in (1) payout ratio and (2) leverage on dummy variables for the various activist strategies.12 0.0 ** -0.55 -0. Payout Ratio Coeff Initial Value Cash Business Strategy M&A Advice Sale Capital Structure Corporate Governance Other No.8 29 t -stat 0.

0 * 178 Sales t -stat -0. and R 2 0. In Column 1.57 ** ** ** ** .47 0.23 -1. *.27 ** 0.2 0.26 0.62 0. **. and 1% levels.6 3.23 ** 2. and Column 2 shows a regression that predicts the average premium those firms receive.8 40. and *** indicate statistical significance at the 10%.98 -0. 5%.26 2. *.5 * 2.7 74.0 2.06 Table D-13 Sale Efficiency By Activist Strategy Column 1 shows a logit model that predicts whether a target firm will sell itself.14 2. considering only the sample of investments which announce activist strategies.8 32 0. and 1% levels. the constants are supressed. Column 1 presents R 2 .32 0.7 -4. 5%.23 -0.3 13.9 3.77 0.04 1. Due to the full span of the dummy variables.6 15.3 12.Table D-12 Firm Size Reductions By Strategy The table regresses the two-year percentage change in (1) assets and (2) sales on dummy variables for the various activist strategies.33 1.10 1.6 ** 185 t -stat -2.21 0.14 1.10 (2) Avg Premium Coeff t -stat 7. **. The data is winsorized at the 10% level.93 0.7 24.62 0. The regressions control for the initial assets or sales.26 -0.0 ** 3.65 2.0 14.6 -4.65 0.65 1.95 0.43 0.74 -0.68 -1. and *** indicate statistical significance at the 10%. and Column 2 presents Pseudo-R 2 .11 -1.23 1. Both regressions include data only from investments which use activist strategies.6 -2. (1) Odds of Sale Odds z -stat Bus Strategy M&A Sale Cap Struc Corp Gov Other No. I measure premium as the total return to shareholders from one month before the activist first files a Schedule 13D until the sale closes.06 1.0 4.05 Coeff 0. Assets Coeff Initial Value Business Strategy M&A Advice Sale Capital Structure Corporate Governance Other No.83 0. the dependent variable is a dummy set to 1 if the firm sold itself . In Column 2 the dependent variable is the average premium for each such sale. and R 2 0.48 164 -2.

Journal of Financial Research 29: 199-216. Barber. The “Wall Street Walk’ and Shareholder Activism: Exit as a Form of Voice. Bebchuck. Anat and Pfleiderer. The Modern Corporation and Private Property. Bebchuck. Lucian and Jesse Fried.com/abstract=934712. 1998. Share Restrictions and Asset Pricing: Evidence From the Hedge Fund Industry. Bethel. Available at SSRN: http://ssrn. Bainbridge. Amar. Adolf and Gardiner Means. Bratton. 1918(R2). Stanford Graduate School of Business Research Paper No. Marco. Journal of Investing 59: 66–80. Direct Evidence on Market-Driven Acquisitions Theory. 2005. The Hidden Costs of Stock Market Liquidity. Law–Econ Research Paper No. and Stefano Rossi. Block share Purchases and Corporate Performance. Stephen. 1932. Directory Primacy and Shareholder Disempowerment. Lucian. The Journal of Finance 53: 605–34. Ang. Julia Liebeskind. Jennifer. New York: Macmillan. 2006. Monitoring the Monitor: Evaluating CalPERS’ Shareholder Activism. George. . Virginia Law Review 93: 675– 732. Bhide. Julian Franks. Michigan Law Review 89: 520–608. 2007. 05–25. and Tim Opler. 1990. 2006. Working Paper. Journal of Financial Economics 34: 31–51. 2007. 2009. 2003. Harvard Law Review 118: 835–914.com/abstract=928689. Hedge Funds and Governance Targets. Bebchuk. Returns to Shareholder Activism: Evidence From a Clinical Study of the Hermes UK Focus Fund. Paul. William. UCLA School of Law. The Case for Increasing Shareholder Power. Executive Compensation as an Agency Problem. James and Yingmei Cheng. Journal of Economic Perspectives 17: 17–92. 2006. Lucian. The Myth of the Shareholder Franchise. Journal of Financial Economics 83: 33–58. Berle. Working Paper. Black. 2008. Colin Mayer. Available at SSRN: http://ssrn. Aragon. 2007. Becht. Bernard. Brad. 1993. Shareholder Passivity Reexamined.References Admati.

Corporate Governance. Del Guercio. Available at SSRN: http://ssrn. The Influence of Institutions of Corporate Governance through Private Negotiations: Evidence from TIAA–CREF. Columbia Law Review 91: 1277–1368.upenn.wharton. Willard. The ew Palgrave Dictionary of Economics and the Law 3: 459–465. 1909. Capital Market Responses to Environment Performance in Developing Countries. Susmita. Journal of Financial Economics 85: 552–70. Diane and Jennifer Hawkins. The Journal of Finance 53: 1335–62. Shareholder Activism and Corporate Governance in the United States. Hedge Funds as Shareholder Activists from 1994– 2005. 2008. 141/2007. 1999. Journal of Financial Economics 52: 293–340. Boyson. Coffee. 2006. and Tracie Woidtke. Working Paper. The World Bank Development Research Group Working Paper No. Black. Brav. 1998. Wei Jiang. and Wei Jiang. Frank Partnoy. 1998. Benoit Laplante. Davis. Available: http://finance. 193. Do Board Members Pay Attention When Institutional Investors “Just Vote No”? CEO and Director Turnover Associated with Shareholder Activism.com/abstract=575242. John. James Nelson. Bernard.Black. Christopher. 2007.com/abstract=992739. Christoffersen. Hedge Fund Activism. and Nlandu Mamingi. and Adam Reed. Bernard. Value Creation or Destruction? Hedge Funds as Shareholder Activists. Bradley. and Firm Performance. ECGI – Finance Working Paper No.edu/~itayg/Files/Research%20Statement%20– %20Website. The Journal of Finance 63: 1729–75. Dasgupta. and Randall Thomas. Working Paper. Susan. Alon Brav. Journal of Corporate Finance 14: 323–36. Agents Watching Agents: The Promise of Institutional Investor Voice. Vote Trading and Information Aggregation. Available at SSRN: http://ssrn. Itay Goldstein. . Shareholder Activism and Price Dynamics: Evidence From Closed–End Funds. and Michael Weisbach. 2008. Liquidity Versus Control: The Institutional Investor As Corporate Monitor. Laura Wallis. 1997. 2007. Nicole and Robert Mooradian. UCLA Law Review 39: 811–893. Diane. Working Paper.pdf. Chris Gêczy. Gerald and Han Kim. Clifford. 2008. David Musto. Del Guercio. Carleton. 1992. Alon. Business Ties and Proxy Voting by Mutual Funds. Michael. The Motivation and Impact of Pension Fund Activism.

1970. Gillan. Journal of Finance 59: 2041–60. Ownership Structure. 2000. Greenwood. Benjamin. Why Has CEO Pay Increased So Much? Quarterly Journal of Economics 123: 49–100. The Intelligent Investor. Journal of Applied Corporate Finance 19: 55–73. Hall. Eugene. 2008. Olubunmi. Joseph. Xavier and Augustin Landier. Harvard Business School Working Paper. Lilli and John Pound. Stuart and Laura Starks. Cash and Corporate Control. Eugene and Kenneth French. 1992. 1973 [2003]. and Shareholder Voting: Evidence From Shareholder–Sponsored Corporate Governance Proposals. the Free Rider Problem. Eugene and Michael Jensen. Stuart and Laura Starks. Journal of Law and Economics 26: 301–25. Robin and Michael Schor. Journal of Finance 47: 697–718. Revised Edition.Deli. Grossman. Journal of Financial Studies 47: 427-65. Gillan. 1998. Information. Bell Journal of Economics 11: 42–64. 1983.com/abstract=663523. Daniel. Stuart and Laura Starks. The Journal of Finance 25: 383–417. 2004. Available at SSRN: http://ssrn. 1993. Separation of Ownership and Control. Fayele. 1992. Golec. . 2007. Fama. Working Paper. Empirical Tests of a Principal Agent Model of the Investor–Investment Advisor Relationship. The Trouble with Stock Options. and the Theory of the Corporation. Journal of Finance 58: 109–33. Journal of Financial and Quantitative Analysis 27: 81–95. 2002. Fama. Sanford and Oliver Hart. Fama. New York: Harper Collins. Efficient Capital Markets: A Review of Theory and Empirical Work. Corporate Governance Proposals and Shareholder Activism: The Role of Institutional Investors. Brian and Kevin Murphy. Takeover Bids. Gillan. The Cross-Section of Expected Stock Returns. The Evolution of Shareholder Activism in the United States. Mutual Fund Advisory Contracts: An Empirical Investigation. Journal of Financial Economics 57: 275– 305. 1980. Gabaix. A Survey of Shareholder Activism: Motivation and Empirical Evidence. Graham. Gordon. 2008. Journal of Economic Perspectives 17: 49–70. Hedge Fund Investor Activism and Takeovers.

Journal of Corporate Finance 13: 343–67. Jonathan. MacKinlay. Robert. Marilyn and Margaret Shackell. Johnson. Craig. McConnell. American Economic Review 76: 323–29. Richard Sias. Paul Malatesta.com/abstract=885365. 2006. and David Wessels. 2007. and Ralph Walkling. Karpoff. Valuation: Measuring and Managing the Value of Companies. Available at SSRN: http://ssrn. Matthew. Klein. Michael and William Meckling. Lipton. Event Studies in Economics and Finance. 2007. 1991. 2003. Tim. Karpoff. Insiders. Wall street bosses spending too much time on the golf course. A New System of Corporate Governance: The Quinquennial Election of Directors. Lynn. 2009. Institute for Law & Economics Research Paper No. and Laura Starks.Hu. Jensen. 2005. Parrino. Wall Street Journal. Ne York: Wiley. 1990. 1996. Koller. Journal of Financial Economics 42: 365–95. Shareholders’ Proposals on Executive Compensation. Journal of Financial Economics 27: 595–612. The Journal of Finance 64: 187–229. University of Pennsylvania. University of Chicago Law Review 58: 187–253.com/sbstract=121. Michael. Working Paper. The Impact of Shareholder Activism on Target Companies: A Survey of Empirical Findings. Kahan. Working Paper. 2001. 1997. . Entrepreneurial Shareholder Activism: Hedge Funds and Other Private Investors. Hedge Funds in Corporate Governance and Corporate Control. and the Decoupling of Economic and Voting Ownership: Empty and Hidden (Morphable) Ownership. Jonathan. Theory of the Firm: Managerial Behavior. Marcel and Edward Rock. 1986. Martin and Steven Rosenblum. 17 November. 1997. Agency Costs and Ownership Structure. April and Emauel Zur. Hedge Funds. 06–16. John & Sons. Marc Goedhart. Voting With Their Feet: Institutional Ownership Changes Around Forced CEO Turnover. John and Henri Servaes. Corporate Governance and Shareholder Initiatives: Empirical Evidence. Henry and Bernard Black. Jensen. Journal of Financial Economics 3: 305–60. The Journal of Economic Literature 35: 13–39. Available at SSRN: http://ssrn. 1976. The Agency Cost of Free Cash Flow: Corporate Finance and Takeovers. Journal of Financial Economics 68: 3–46. Additional Evidence on Equity Ownership and Corporate Value.

Determinants of Hedge Fund Activism and Shareholder Gains. Sunil. . Less Is More: Making Shareholder Activism a Valued Mechanism of Corporate Governance. Roberta. ature and Causes of the Wealth of ations. Pound. Journal of Financial Economics 20: 237-65. 2001. Kenneth Wiles. Proxy Contests and the Efficiency of Shareholder Oversight. and March Zenner. Romano. Romano. Helsinki. The Logic and (Uncertain) Significance of Institutional Shareholder Activism. Pekka. Wahal. Edward. Yale University. Smith. Michael. 2008. 1996. 1992. 1996. Rock. 1988. Strickland. John. 1993. Joshua. Shareholder Activism by Institutional Investors: Evidence from CalPERS. Adam. An Inquiry Into the Chicago: University of Chicago Press. Helsinki School of Economics. 1992. Public Pension Fund Activism in Corporate Governance Reconsidered. Georgetown Law Review 79: 445–406. The Wandering Eye: Do Investors Steal Ideas from 13Fs? Unpublished BA. Smith. David. Deon. Yale Journal on Regulation 18: 174–251. Journal of Applied Corporate Finance 5: 6–18. Roberta. Unpublished MA. Journal of Finance 51: 227–52. Raiders. Journal of Financial and Quantitative Analysis 31: 1–23. John. Puustinen. Pension Fund Activism and Firm Performance. A Requiem for the USA: Is Small Shareholder Monitoring Effective? Journal of Financial Economics 40: 319–38. New Haven. Columbia Law Review 79: 445–506. 1995. 1996. 1776 [1977]. Yermack. Do Corporations Award CEO Stock Options Effectively? Journal of Financial Economics 39: 237–69. Targets. and Politics: The History and Future of American Corporate Control. Sweren.Pound.

Sign up to vote on this title
UsefulNot useful