Download - Activism Adds Value
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.
- 1 -
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 so-
called 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.
- 2 -
investments by 3.2%. In addition, investors react positively to activism even when it contains no
stock picking signal. When activists change their intentions from passive to active, stock prices
jump significantly. Except for activism seeking to fix corporate governance, investors reward all
activist strategies. While I find evidence that activism yields significant excess returns, I also
find that these returns have declined over time.
Long-term returns also indicate that hedge fund activism enriches shareholders. Over two
years, controlling for the differences in target firms, activist targets outperform passive targets by
18.4%, and investments seeking board seats outperform other activist investments by 18.6%.
Again, except for corporate governance activism, all strategies generate substantial long-term
returns. Also consistent with short-term results, returns have declined over time.
Activism earns these excess returns by improving its targets—they increase profitability,
optimize their capital structures, divest unprofitable assets, and sell themselves more frequently
and for higher prices. Controlling for the differences in target firm characteristics, activist targets
increase their return on assets (“ROA”) by 3.0% more than passive targets, increase their
margins by 2.9% more, increase their payout ratios by 2.5% more, and increase their leverage by
0.7% more. Activist targets also reduce their assets by 4.6% more than passive targets—most
likely by shedding unprofitable divisions. Finally, not only do activist targets sell themselves
more frequently, but they also sell for higher prices.
In improving target firms, activists follow five cohesive 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. Each strategy
targets different firms and improves them in distinct ways. Further, except for corporate
governance, every strategy earns significant excess returns.
- 3 -
The new evidence presented in this paper suggests that, far from economic hitmen,
activist hedge funds add significant value to their targets—judged by either stock prices or
financial statements. Thus, instead of avoiding hedge funds’ calls, both management teams and
shareholders alike should welcome their insights.
This paper is organized as follows. Section I reviews the literature. Section II explains my
methodology. Section III describes hedge fund activism; it details the firms that activists target
and the strategies that they use. Turning to the analysis, Section IV shows that activists earn
excess short-term returns. Section V presents similar evidence for long-term returns. Section VI
describes the fundamental improvements in activist targets. Section VII concludes.
Section I: Background
For hundreds of years, CEOs have eroded the wealth of nations. In 1776, Adam Smith
warned, “The directors of such [joint-stock] companies, however, being the managers rather of
other people’s money than of their own, it cannot well be expected, that they should watch over
it with the same anxious vigilance with which the partners in a private copartnery frequently
watch over their own.”4 In modern terms: CEOs often roll up their sleeves not for auditing
spreadsheets but for dealing cards. In August 2007, while errant hedge funds brought his firm to
its knees, Bear Stearns CEO Jimmy Cayne spent 10 days playing bridge out of state.5
Agency costs arise whenever management (the agent) takes actions which harm
shareholders (the principal), generally because their incentives differ. In general, 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. 2005). But CEOs paid based
4 For a more modern treatment, see Berle and Means (1932) and Jensen and Meckling (1976). 5 The same November 17, 2007 Wall Street Journal Article, titled “Wall Street Bosses Spending Too Much Time on the Golf Course,” also reported that in June alone, Cayne squeezed in 13 rounds of golf.
- 4 -
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. Similarly, CEOs paid based on ROA
might forgo projects with attractive IRRs if those projects would drag down a firm’s overall
returns. While some herald stock options as an incentive alignment panacea, managers only earn
those options by meeting performance targets, which offer their own perverse incentives.6
Further, Jensen (1986) argues that managers may grow the company beyond its optimal
size, indiscriminately purchasing assets. While doing so harms shareholders, it also increases
management’s compensation, pride, and job security. Gabaix and Landier (2008) find that from
1980 through 2003, the six-fold increase in large US firms’ market capitalization led directly to a
six-fold increase in CEO pay. So while shareholders might want the firm to disgorge
unproductive assets, management might harbor other intentions.
And even if managers wanted to maximize shareholder value, they just might not know
any better. At Prudential Bancorp’s February 9, 2007 annual meeting, for example, shareholder
activist Joseph Stilwell publicly wagered $25,000 that CEO Thomas Vento could not define
return on equity on a per share basis. Vento failed.7
For these reasons, Black (1992) and Pound (1992) argue that shareholder activism could
improve corporate performance. With their voting rights, large shareholders might combat
misguided management teams. But on the whole, the literature strongly disagrees. The first
problem is that too few activists play the game. Due to the free rider problem, activists’ personal
benefits do not justify the necessary costs.8 Bebchuk (2008) discusses the costs that activists
must endure. Parrino et al. (2003) suggest that stockholders may simply prefer to walk away.
6 For the problems with performance targets, see Yermack (1995) and Bebchuk and Fried (2003). For the problems with options themselves, see also Hall and Murphy (2003). 7 Predential Bancorp, SEC Schedule 13D, Filed 2 November 2008. 8 See Rock (1992), Black (1998), Bainbridge (2006), and Kahan and Rock (2006).
- 5 -
Even when activists do take up arms, their wooden arrows make little dent in the
corporate fortress. Most institutional investors suffer conflicts of interest, such as those mutual
funds face; regulatory constraints, including diversification requirements and insider trading
regulations; and political constraints, especially for public pension funds.9 The empirical record
agrees; activists sometimes induce improvements in corporate governance, but they rarely, if
ever, induce long-term changes in performance.10 In a literature review, Black (1998) concludes,
“A small number of American institutional investors spend a trivial amount of money on overt
activism efforts. Institutions achieve the effects on firm performance that one might expect from
this level of effort—namely, not much.”
While the early literature focuses on mutual funds, pension funds, and “block
purchasers,” this thesis argues that hedge funds succeed where others have failed. First, drawing
on prior legal scholarship, I argue why activist hedge funds might outperform their activist peers.
Second, I document empirical evidence that activist hedge funds do improve corporate value.
(A) Hedge Funds Are Different
As many a pitchbook eagerly points out, hedge funds are different. They enjoy many
structural advantages that distinguish them from prior activists. In particular, hedge funds have
the incentives to engage in activism—(1) they earn pay-for-performance bonuses, (2) they suffer
fewer conflicts of interest, and (3) they can avoid ERISA regulation—and they have the weapons
to engage in activism—(4) they can make concentrated bets, (5) they can take can take long-
term, illiquid positions, and (6) they can use derivatives. Below, I elaborate on these points.
9 See Black (1990) and Romano (1993). For other critiques, see Lipton and Rosenblum (1991), Karpoff, Malatesta, and Walkling (1996), Carleton, Nelson, and Weisbach (1998), Davis and Kim (2008), and Del Gurcio, Wallis, and Woidtke (2008). 10 For corporate governance successes, see Bebchuck (2005, 2007). For other failures, see Wahal (1996), Karpoff, Malatesta, and Walkling (1996), Johnson and Shackell (1997), Black (1998), Gillan and Starks (2000), Karpoff (2001), Romano (2001), Barber (2007), Gillan and Starks (2007), and Del Guercio, Wallis, and Woidtke (2008).
- 6 -
First, hedge funds earn pay-for-performance bonuses, which incentivize them to wage
successful activist campaigns. Hedge funds do not face the 1940 Investment Company Act’s pay
restrictions. Indeed, hedge fund managers often earn up to 20% or more of their firm’s annual
return (Brav et al. 2008). In contrast, mutual fund managers rarely earn performance bonuses.11
Second, hedge funds suffer fewer conflicts of interest than other institutional investors.
For example, mutual fund managers might placate a company to win its pension business (Davis
and Kim 2007), and public pension fund managers might flatter a CEO to win contributions for a
re-election campaign (Kahan and Rock 2006).
Third, hedge funds can avoid ERISA or “prudent man” regulations, which restrict the
risks other institutional investors can stomach.12 The threat of liability may scare these potential
activists from taking the risk.
Fourth, hedge funds can make concentrated bets to fight entrenched management teams.
Similarly, activists might threaten to purchase the entire target company, often as a negotiating
ploy. In contrast, restrained by subchapter M of the Internal Revenue Code and the 1940
Investment Company Act, 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).
Fifth, hedge funds can take illiquid positions to weather a long fight. Whereas hedge
funds sign investors into lengthy lock-up agreements, mutual funds must maintain constant
liquidity to satisfy redemption requests (Klein and Zur 2008).
And finally, hedge funds use exotic strategies. Through stock lending and derivatives,
hedge funds might acquire voting rights without taking on economic exposure.13 Or hedge funds
11 Golec (1992) and Deli (2002) find that only 6-7% of mutual fund companies offer their managers performance bonuses. Kahan and Rock (2006) report that 97% of mutual funds’ annual compensation ignores investment gains. 12 See Coffee (1991), Bhide (1993), and Aragon (2007). 13 See Christoffersen, et. al. (2007) and Hu and Black (2007).
- 7 -
might use these derivatives expressly to gain economic exposure; by increasing activism’s
potential gains, they justify its requisite costs. In contrast, mutual funds must follow the
Investment Company Act’s short-selling and leveraging restrictions (Klein and Zur 2008).
(B) Similar Studies Do �ot Distinguish Activism From Stock Picking
While activists proclaim their successes in myriad press releases, it is not so easy to
assess their claims. Simply because an activist earns abnormal returns, for instance, does not
mean it actually used activism to earn those returns. And simply because an activist target
improves its margins does not mean the activist did anything to catalyze the change. Rather than
activism, these findings might simply mark great passive investing—choosing companies primed
to improve beyond the market’s expectations. Thus, to test whether activism adds value beyond
stock picking, we must compare activist investments to passive investments.
To date, nonetheless, the literature continues to compare apples to pairs—hedge funds’
activist investments to a non-hedge fund benchmark. Several recent papers focus on hedge fund
activism.14 Using a case study approach, Becht et al. find that the Hermes U.K. Focus Fund’s
activism earns large, positive excess returns. More similar to this study, Puustinen (2007),
Boyson and Moradian (2008), Brav et al. (2008), and Klein and Zur (2008) study large cross-
sections of hedge fund activism. They find strong evidence for short-term abnormal returns and
mild evidence for long-term improvement. But none of these studies distinguishes activism from
passive investing in disguise; they provide no evidence that activists have done anything more
than simply pick undervalued companies poised to outperform expectations.
To mitigate this problem, Clifford (2008) compares activist investments to passive
investments. But this study offers a flawed dichotomy. First, Clifford fails to control for
14 Becht et al. (2006), Puustinen (2007), Boyson and Mooradian (2008), Brav et al. (2008), Clifford (2008), Greenwood and Schor (2008), and Klein and Zur (2008) study hedge fund activism.
- 8 -
differences in target firms. Without doing so, activism might outperform simply by targeting
more attractive firms. Second, Clifford compares Schedule 13D filings, which he calls active, to
Schedule 13G filings, which he calls passive. But these filings differ across other crucial
dimensions. Most importantly, hedge funds must file Schedule 13Ds within 10 days of building
their position but need not file Schedule 13Gs for an entire year. The larger reaction around
Schedule 13D filings, therefore, might simply represent the market preferring fresh ideas to
pickled ones. Clifford also defines activism too broadly. In particular, he calls all Schedule 13D
filings activist investments, even though many explicitly renounce activist intentions. In my
sample of 942 Schedule 13Ds, for instance, 48.1% simply hold the stock for “investment
purposes.” Instead of capturing the active-passive distinction, Clifford’s study might measure
some other confounding variable.15 This paper eliminates these problems.
In addition to ignoring stock picking, prior studies barely discus heterogeneity within
activism. While certain activist strategies, like selling the target firm, might generate large
returns, others, like improving corporate governance, may not. Further, these strategies might
affect companies in conflicting ways that cancel each other out, leaving little aggregate effect.
For example, business strategy activism might reduce leverage, while capital structure activism
might increase leverage. Without testing for such heterogeneity, previous studies improperly
conclude that fundamentals change very little. In contrast, I analyze activists’ strategies in detail.
Greenwood and Schor (2008) present some evidence on heterogeneity. In particular, they
disagree that activists improve their targets; instead, they propose that activists simply put firms
15 There are several possible confounds. First, even passive investors holding at least 20% in a company must file a Schedule 13D, so Clifford’s results might reflect investors’ preference for these concentrated bets. Second, an investor might file a Schedule 13D to convey information to other investors. Well-known passive investors, for example, might benefit from leaking their best ideas to the public. Sweren (2008) documents abnormal returns when hedge funds leak their best ideas. And third, astute investors might prefer to file Schedule 13Ds to protect themselves from potential litigation. These legally savvy investors might also have the best ideas.
- 9 -
in play. The authors show that, excluding sales, activist hedge funds earn negligible long-term
excess returns. But by asymmetrically removing successful sales, and not failed sales, Grenwood
and Schor have imprecisely measured activists’ non-sale returns, biasing their results downward.
In Section V, I correct for this bias.
Additionally, Greenwood and Schor (2008) argue that hedge funds’ poor performance
during the 2008 Credit Crunch proves that they rely on selling their targets. But this overlooks
the crisis’s main symptom; levered institutions like hedge funds unwound dramatically. This
structural pressure, and not activists’ reliance on selling their targets, might explain the results.
In summary, while the literature documents some activist outperformance, (1) it fails to
distinguish this outperformance from stock picking, and (2) it largely ignores heterogeneity.
Section II: Experimental Design
(A) Contributions to the Literature
To argue that activism adds value, I improve on the literature in several ways. This is the
first paper to: (1) use a comprehensive sample of experienced hedge funds, (2) compare activist
Schedule 13D investments to passive Schedule 13D investments, (3) control for differences in
target firms, (4) measure short-term abnormal returns around events which occur after the hedge
fund has disclosed its position, (5) analyze board seats, and (6) fully document cross-sectional
variation in activists’ strategies. This section elaborates on these frontiers.
First, in contrast to the existing literature, I study hedge funds that repeatedly pursue
activism. Prior scholars call “activist” any hedge fund which has ever filed an activist Schedule
13D. But this includes inexperienced hedge funds, which bring down the averages. Expanding
the activist sample to include these novices resembles defining an athlete as anyone who has ever
- 10 -
owned a baseball glove; after rounding up two NY Yankees and five hundred physical education
teachers, prior scholars conclude that no one can consistently throw a 90 mph fastball. By
selecting only hedge funds known for activism, I resolve this bias.
Second, to disentangle activism from stock picking ability, I compare hedge funds’ active
Schedule 13D investments to the same hedge funds’ passive Schedule 13D investments. I expand
on this method below. While Clifford (2008) also compares activist returns to passive returns, I
improve his methodology by (A) limiting my focus to Schedule 13Ds and (B) recognizing that
many Schedule 13Ds espouse passive intentions. I also expand on his study by more thoroughly
examining changes in target firms’ accounting performance.
Third, I control for variation in target firm characteristics like undervaluation and
underperformance. I expand on this method below. This improves on Clifford’s (2008) study by
dismissing the alternative hypothesis that activists target more attractive firms.
Fourth, I document short-term abnormal returns around events that take place after the
initial Schedule 13D filing. By following each investment’s SEC trail, I find subsequent filings
in which a hedge fund changes its intentions—for example, the first time it expresses activist
intentions. Because the hedge fund has already disclosed a passive interest in the target firm, this
method more cleanly isolates the activist signal from the stock picking signal.
Fifth, I distinguish activist instruments, like seeking board seats, from activist strategies,
like selling the firm; an activist uses instruments to implement its strategies. I consider five
instruments: seeking board seats, launching proxy fights, offering to buy the target, providing
private financing, or suing the target. Because 80% of all investments using an instrument seek
board seats, I consider them most thoroughly. Board seats distinguish activism from stock
picking by providing tangible proof that the hedge fund seeks to influence its target.
- 11 -
Sixth, I identify five cohesive activist strategies: business strategy, M&A, sale, capital
structure, and corporate governance. While Puustinen (2007) and Brav et al. (2008) analyze
short-term returns by strategy, I add evidence that strategies differ over the long-run. I also
investigate heterogeneity in fundamental improvements. This is important for two reasons. First,
prior studies have observed little change in target firms’ fundamental accounting performance,
perhaps because different strategies cancel each other out. And second, only by understanding
their strategies can we replicate activist hedge funds’ ability to improve firms.
(B) Testing For Activism
To prove that hedge fund activism adds value, 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. But unlike the Beatles, these findings cannot stand
alone. If activism earned excess returns without improving target firms, then hedge fund activism
would simply be a penname for astute stock picking. On the other hand, if activist targets became
better companies without earning excess returns, then the market must have already expected
those improvements—and the activist probably did not cause the change.
By comparing activist targets to passive targets, I come closer to measuring how these
activist targets otherwise would perform than does any prior paper. Both activist and passive
investments benefit from hedge funds’ superior stock picking abilities, but only activist targets
benefit from hedge funds’ activist abilities. Thus, to the extent activist investments outperform
passive investments, activism must add value.16
There is, however, a second layer of stock picking to peel back. Hedge funds might pick
more attractive firms for activist investments than for passive investments. To eliminate this bias,
16 Thus, activist outperformance is a sufficient condition to show that activism adds value. It is probably a necessary condition, too. Because activism costs more than passive investing, hedge funds should demand greater returns.
- 12 -
I control for differences in target firms—in particular, differences in undervaluation and in
underperformance. If activist investments still outperform passive investments, we can conclude
that activism adds value.
Additionally, I analyze board seats to better distinguish activist investments from passive
investments. Unlike the mere rhetoric accompanying many filings, seeking board seats helps
activists influence the firm’s decisions. Even if it never wins a seat, the hedge fund has opened a
dialogue with management. At best, the activist might get its way. At worst, the activist might
threaten a proxy fight as a bargaining chip. Thus, whereas a fine line separates activism from
passive investing, a much thicker line separates board seat investments from other investments.
As before, controlling for target firm variation, if we observe that board seat investments
outperform other activist investments, board seats must add value.
(C) Collecting Data
Unlike previous studies, which include hedge funds that have made very few activist
investments, I study only hedge funds which focus on activism. 17 In particular, I consider the 48
hedge funds which the shareholder activism advisory firm 13D Monitor defines as most active. 18
13D Monitor sought to include every fund that repeatedly uses activism; that is, rather than
assembling a random sample of a broad universe, they compile a comprehensive list that fits a
narrow criteria. Thus, my sample generalizes not to every ex-Goldman Sachs trader who files a
Schedule 13D, but rather to experienced activists. Because it is comprehensive, the list suffers
from little if any bias in describing the historical record. I find no evidence of survivorship bias
17 Puustinen (2007), Boyson and Mooradian (2008), Brav et al. (2008), Clifford (2008), and Klein and Zur (2008) all use news searches to form an initial hedge fund list. Thus, even hedge funds with only one prior activist investment enter the sample. Brav et al. (2008) limit their sample to hedge funds with at least two activist investments. 18 13D Monitor maintains a subscription-based database of Schedule 13D filings at http://www.13dmonitor.com/. 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. To the extent I had time to emerge from my dorm room at any point during the month of January, Ken and Robyn deserve full credit.
- 13 -
when comparing my sample to those in similar papers.19 As evidence against backfill bias,
several hedge funds in my sample performed quite poorly.20 And finally, even if the sample were
biased towards hedge funds whose activist investments perform quite well, it would not
necessarily be biased towards hedge funds whose activist investments outperform their passive
investments. In fact, if activism does simply represent disguised passive investing, these hedge
funds may be the biggest culprits. Thus, 13D Monitor’s sample provides a fitting laboratory.
Using 13D Monitor’s online database, I collected information on all 942 investments that
these hedge funds made. To further dissect these investments, I turned to the Edgar database,
which houses SEC filings dating back to 1994. Section 13(d) of the 1934 Exchange Act requires
investors to file with the SEC when they amass large stakes with activist intentions. In particular,
hedge funds must file a Schedule 13D within 10 days of acquiring at least 5% in any public
company.21 In Item 4, filers must declare why they acquired the shares. By reading over 6,000
individual SEC filings, I collected the entire activist history for each investment.
From these 942 investments, I narrowed the sample. First, to focus on the current activist
climate, I excluded filings before January 1, 2000. Second, I excluded all firms which declared
bankruptcy before the hedge fund invested. This narrowed my sample to 718 investments, of
which I classified 393 as active and 325 as passive. I classify an investment as activist if the
investor announces any activist instrument—board seats, proxy fights, buyout offers, private
financings, or lawsuits—or any activist strategy—business strategy, M&A, sale, capital
structure, or corporate governance. Of the 393 activist investments, 213 sought board seats.
19 I compared my sample to two sources: Puustinen (2007)’s sample and hedge fund consulting firm Hedge Fund Solutions’s database. Without exception, the most frequent funds in each source overlapped with my sample. 20 Of the 46 firms with long-term return data, 23.9% exhibit negative average activist abnormal returns. 21 While activists rarely pierce the 5% threshold in large firms, this sample paints an accurate picture; activists tend to focus on targets small enough that they can establish large positions.
- 14 -
Because this is the first paper to suggest that many Schedule 13D filings might harbor
passive intentions, some elaboration is in order. Passive institutional investors who acquire at
least 5%, but less than 20%, of a company’s stock may file a Schedule 13G with the SEC.22 But
sometimes these firms will instead file a Schedule 13D. 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.” Should we believe them? First, the SEC and target
firms use the courts to enforce truthful disclosures.23 Second, citing conversations with hedge
fund managers, Clifford (2008) argues that hedge funds exercise extreme caution to avoid filing
any misleading disclosures. Third, each hedge fund in my sample announces either passive or
activist intentions at least once, which implies that these funds do differentiate passive from
activist investments. Fourth, I find that after 32.4% of passive Schedule 13Ds, hedge funds
subsequently announce activist intentions. This devotion to updating their goals indicates that
hedge funds take care to convey honest information. And finally, in searching Bloomberg news
for certain randomly chosen passive investments, I found no evidence that the hedge funds
engaged any attempts to change the target firms.
So why might a hedge fund file a Schedule 13D rather than a Schedule 13G? First, using
the more rigorous Schedule 13D safeguards the hedge fund against liability. Second, funds might
take advantage of Schedule 13D’s immediate disclosure. They may wish to communicate their
investment thesis so that other investors, upon realizing the firm’s true value, bid up its price.24
After collecting information on these investments, I pulled share price histories from the
Center for Research in Security Prices (CRSP). I calculate the target firm’s return in excess of
22 Passive investors acquiring at least 20%, however, must file a Schedule 13D. While this might bias my Schedule 13D sample towards large passive firms, I find almost no such instances of 20%+ passive positions. 23 See, e.g., SEC vs. Montgomery Medical Ventures, LP (1996) and Ronson Corp v. Steel Partners II (2005). 24 For example, Sweren (2008) documents excess returns after hedge funds file Schedule 13Fs.
- 15 -
the market model measured against the CRSP value-weighted NYSE/Amex/Nasdaq index. I
describe the event study method more fully in Appendix A. While unreported, all results are
robust to the use of the Fama French factors. The analysis excludes firms for which CRSP lacks
data—most likely small, unlisted firms. This might bias my findings.
I use two types of data on target fundamentals. First, I use accounting data pulled from
Bloomberg. Second, I calculate industry-adjusted statistics to better analyze a target firm’s
idiosyncratic performance. For example, following Ang and Chen (2006) and Puustinen (2007), I
use abnormal q to proxy for undervaluation. 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, using the GICS industry classification. I also calculate abnormal margins,
abnormal return on equity (“ROE”), and abnormal leverage, where the firm-specific component
(denoted abnormal) is the difference between the firm’s value and the median value of its
industry peers. I calculate these variables by hand using firm-level data from Compustat.
Unfortunately, I could only obtain Compustat data dating back to January 1, 2004. Both
Bloomberg and Compustat were missing data for several firms, presumably smaller firms.
Section III: Hedge Fund Activism
Before turning to the results, this section elaborates on the differences between activism
and passive investing. First, I describe the firms which activist hedge funds target. In general,
they target undervalued firms, with low Tobin’s q.25 While hedge funds target relatively
profitable firms overall, their activist efforts target less profitable firms than do their passive
25 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. A low Tobin’s q implies undervaluation.
- 16 -
Table I
Target Firm Characteristics
(1) (2)
Activism Board Seats
Activist Passive Diff Board Other Diff
Market Cap 757.1 654.4 102.7 * 636.0 891.4 -255.4 ***
Tobin's q 1.6 1.7 -0.1 * 1.5 1.7 -0.2 ***
Abnormal q -16.6 12.3 -29.0 *** -22.9 -11.1 -11.7
EBITDA Margin 10.9 12.1 -1.1 * 8.4 13.5 -5.1 ***
Abnormal Margin 3.0 3.2 -0.2 0.7 5.0 -4.3 ***
Return on Assets 8.5 9.4 -1.0 ** 7.4 9.5 -2.1 ***
Return on Equity 20.0 22.3 -2.4 * 17.0 23.0 -6.0 ***
Abnormal ROE 0.7 2.4 -1.7 ** 0.2 1.2 -1.0
Leverage 22.2 22.8 -0.6 19.7 24.7 -5.0 **
Abnormal Leverage 15.1 11.6 3.5 12.8 17.1 -4.3
Payout Ratio 9.1 9.8 -0.7 10.0 8.3 1.7
Cash/Assets 16.3 14.5 1.9 * 16.6 16.1 0.5
N = 303 492 - 157 146 -
The table reports the characteristics of hedge funds' targets. Column 1 compares activist investments to passive investments.
Column 2 compares activist investments seeking board seats to other activist investments. All variables refer to the trailing
twelve months prior to the Schedule 13D. All data is winsorized at the 10% level. *, **, and *** refer to statistical
significance at the 10%, 5%, and 1% levels.
efforts. Second, I discuss heterogeneity among activists’ strategies. I describe five main
categories: business strategy, M&A, sale, capital structure, and corporate governance.
(A) Activists Target Undervalued Firms
Before we can measure activist investments against a passive benchmark, we should first
describe the ex-ante differences in the two samples. Table I reports summary statistics for active
investments, passive investments, active investments seeking board seats, and active investments
not seeking board seats. In Appendix D, Table D-1 adds more information, including definitions.
Column 1 shows that hedge funds go activist in undervalued and underperforming firms.
Measured by Abnormal q, relative to the median firm in its industry, the average activist target is
16.6% more undervalued. Relative to passive targets, activist targets are 29.0% more
undervalued (t = -3.68). While hedge funds target overall profitable firms, they go active in the
- 17 -
least profitable ones. The average activist target has 3.0% higher margins and 0.7% higher ROE
than the median firm in its industry. Relative to passive targets, however, they have 1.1% lower
margins (t = 1.34), 1.0% lower ROA (t = 1.71), and 2.4% lower ROE (t = 1.49).26 Activist
targets also hoard assets. Relative to passive targets, they have 1.9% more cash per dollar in
assets (t = 1.49) and 0.7% lower payout ratios (t = 0.42).27 Overall, the table provides moderate
evidence that hedge funds try to unlock value in underperforming target firms.
While these results resemble most prior studies, Clifford (2008) finds that active targets
are more profitable than passive targets.28 This might be explained by his methodology; because
he considers all Schedule 13Ds to be “active,” he includes some passive investments. Consistent
with the literature, these passive investments might lift the observed profitability.
Column 2 shows that within activist investments, hedge funds seek board seats in the
most troubled companies. Compared to other activist investments, those seeking board seats are
even more undervalued, with 0.2 lower Tobin’s q (t = 2.47). They are also less profitable, with
5.1% lower margins (t = 4.13), 2.1% lower ROA (t = 2.49), and 6.0% lower ROE (t = 2.59). This
evidence refutes the alternate hypothesis that activists seek board seats to associate themselves
with successful firms. Instead, it seems they seek board seats to improve underperforming firms.
(B) Activists Use Different Strategies
Activists tailor their strategy to each target. From the hedge funds’ SEC filings, I
categorized the investments into five broad strategies: (1) “Business Strategy” activism, which
attempts to improve the firm’s competitive strategy and capital allocation, (2) “M&A” activism,
which seeks to maximize value through extraordinary transactions, like acquisitions or spinoffs,
26 I calculate margins as EBITDA divided by sales. I calculate ROA as EBITDA divided by book value of assets. I calculate ROE as EBITDA divided by the book value of equity. 27 I calculate payout ratio as common dividends divided by net income. 28 See Boyson and Mooradian (2008), Brav et al. (2008), and Klein and Zur (2008).
- 18 -
(3) “Sale” activism, which demands that the firm sell itself, (4) “Capital Structure” activism,
which tries to influence leverage or payout policies, and (5) “Corporate Governance” activism,
which addresses the firm’s governance policies. I improve on Brav et al. (2008) by separating
what they call “business strategy” into business strategy and M&A. This way, I distinguish
between activism which improves companies through capital allocation and activism which
improves companies through corporate transactions. In Appendix B, I elaborate on these
categories by excerpting from SEC filings.
In my sample, hedge funds most frequently use M&A and Sale activism (130 times each)
and least frequently use business strategy activism (75 times). While correlations among the
strategies vary from 15% to 37%, business strategy and sale activism correlate the least, at 7%.
This supports the hypothesis that business strategy and sale activism seek mutually exclusive
goals—where one builds long-term businesses, the other quickly sells them.
If activist strategies truly differ, we should expect each strategy to target firms with
particular weaknesses. Table II shows a logit regression that uses firm characteristics to predict
the hedge fund’s strategy. Business strategy activism targets poorly performing companies with
lower sales growth (t = -1.33) and lower margins (t = -1.22). Targets are also larger (t = 3.13),
Table II
Logit Analysis By Strategy
Bus Strat M&A Sale Cap Struc Corp Gov
Coeff. t -stat Coeff. t -stat Coeff. t -stat Coeff. t -stat Coeff. t -stat
Market Cap 0.06 *** 3.13 0.04 *** 2.78 -0.03 * -1.51 0.03 * 1.29 0.00 0.13
Tobin's q 9.19 0.42 -15.72 -0.89 -21.02 -1.04 -56.34 *** -2.35 10.14 0.50
Growth -1.43 * -1.33 0.62 0.76 -0.20 -0.22 1.21 1.24 -0.06 -0.07
Margin -2.41 -1.22 -0.11 -0.07 0.85 0.51 4.61 *** 2.61 -0.78 -0.43
Leverage 0.17 0.19 0.05 0.07 -0.33 -0.42 -1.30 * -1.41 0.31 0.39
Cash/Assets 1.02 * 1.60 1.43 ** 1.78 1.04 ** 1.66 1.57 ** 2.06 0.79 * 1.43
Constant -3.14 *** -6.78 -2.21 *** -5.75 -1.76 *** -4.71 -2.25 *** -5.14 -2.67 *** -6.46
No. and 547 0.06 547 0.04 547 0.03 547 0.06 547 0.01
Pseudo-R2
Using a logit regression, the table reports the liklihood that activists target firms with various characteristics. The dependent variable is a dummy equal to
one if the hedge fund targets the firm with a particular activist strategy. The sample population includes all hedge fund investments, active or passive. All
data is winsorized at the 10% level. *, **, and *** represent statistical significance at the 10%, 5%, and 1% levels.
- 19 -
Table III
Board Seats By Strategy
(1) (2)
Board Logit Win Logit
Odds z -stat Odds z -stat
Bus Strategy 2.1 *** 2.59 1.5 0.92
M&A Advice 1.2 0.89 1.1 0.33
Sale 1.9 ** 2.40 0.9 -0.29
Cap Structure 1.3 1.01 1.3 0.60
Corp Gov 1.7 * 1.99 1.3 0.61
No. and 316 0.04 155 0.02
Pseudo-R2
The table present the logistic regression odds ratios that each strategy (1) seeks board seats and (2) wins board seats. The
dependent variable is a dummy set equal to 1 if the activist sought or won a board seat. *, **, and *** indicate statistical
significance at the 10%, 5%, and 1% levels.
perhaps because large companies struggle to allocate capital among their many divisions. M&A
and sale activism target undervalued companies (t = -0.89 and t = -1.04). M&A activism targets
large companies (t = 2.78), perhaps because larger firms have underperforming divisions rife for
spinoffs. Sale activism targets smaller companies (t = -1.51), perhaps because they are easier to
sell in whole. Capital structure activism targets healthy firms, which can increase their payout.
Targets enjoy very high margins (t = 2.61) with lots of cash on hand (t = 2.06). They also have
low existing leverage (t = -1.41).29 Corporate governance activism targets firms with excess cash
(t = 1.43), likely following Jensen’s (1986) hypothesis that misaligned managers hoard assets.
Table III describes the frequency with which each strategy seeks board seats. Column 1
and Column 2 present odds ratios from logistic regressions that predict which strategies seek
board seats and which strategies win board seats. Business strategy activism nominates directors
most frequently (z = 2.59)—probably to nurse businesses back to profitability—while M&A
activism nominates directors least frequently (z = 0.89)—probably because the hedge funds
anticipate exiting after quick flips. Similarly, business strategy activism wins board seats most
29 I measure leverage as long-term debt divided by total assets.
- 20 -
often (z = 0.92), while M&A and sale activism succeed least often (z = 0.33 and z = -0.29). This
makes sense; targets prefer long-term partners, like hedge funds advising on business strategy, to
short-term speculators, like hedge funds seeking to sell the business.
In summary, activists target undervalued firms with room to improve. Depending on the
strategy they want to use, they choose different targets. Having described activists’ methods, I
now turn to the first research question. Does this activism increase returns?
Section IV: Short-Term Stock Returns
Everyday in the stock market, shareholders vote with their wallets; with good news, they
bid up prices, and with bad news, they drive them back down. Particularly in efficient markets,
these price swings presage actual changes in firms’ values. Thus, if activism adds value, once
shareholders discover hedge funds’ intentions, we should observe immediate excess returns.
Absent efficient markets, however, these short-term returns do not themselves prove that
activism adds value. They may reflect hedge funds solving the adverse selection problem, not the
moral hazard problem—that is, the filings may signal target firms’ already good prospects
without actually signaling that hedge funds will in fact improve those prospects. Investors may
believe that hedge funds can access better information or conduct better analysis without adding
any activist value. In fact, Sweren (2008) documents short-term abnormal returns around hedge
fund Schedule 13F filings, which convey no explicit activist intentions.
In this section, using three tests, I argue that activism does add value beyond simple stock
picking. First, I compare activist filings to passive filings. I find strong evidence that, in the
short-term, the market favors activist investments—and in particular, activist investments which
seek board seats. Second, I use a dummy variable regression to compare activist filings to
- 21 -
passive filings, controlling for target firm undervaluation and underperformance. Again, activist
investments outperform—particularly when the hedge fund seeks board seats. Third, I identify
events which isolate the activist signal from the passive investing signal. I find significant excess
returns. Thus, the stock market responds favorably to activism, not simply to stock picking.
I also investigate heterogeneity across these returns. In particular, I investigate when
activists add the most value—what target firm characteristics, activist instruments, or activist
strategies does the market judge most effective? Undervalued, underperforming target firms
experience the largest excess returns. I find evidence that the market favors investments which
take active steps, like seeking board seats. Except for corporate governance activism, all
strategies enjoy abnormal returns. Finally, I note that these returns have declined over time.
(A) Activism Exceeds Stock Picking
Table IV reports the market’s reaction to activism. In Appendix D, Table D-2 shows
more details. I calculate excess returns over 61- and 5-day windows relative to the market model
Table IV
Short-Term Returns by Investment Type
Panel A: Active vs. Passive
Active Passive Difference
(-30,+30) (-2,+2) (-30,+30) (-2,+2) (-30,+30) (-2,+2)
Median 11.8% *** 4.4% *** 3.6% *** 1.9% *** 8.2% *** 2.5% ***
Average 12.9% *** 5.7% *** 4.6% *** 2.7% *** 8.3% *** 3.0% ***
% Positive 80.4% *** 79.9% *** 57.6% *** 67.4% ***
N = 194 194 342 340
Panel B: Board Seats vs. Other Active Investments
Board Seats No Board Seats Difference
(-30,+30) (-2,+2) (-30,+30) (-2,+2) (-30,+30) (-2,+2)
Median 12.5% *** 7.8% *** 10.8% *** 4.1% *** 1.7% * 3.7% *
Average 17.7% *** 8.6% *** 12.6% *** 5.7% *** 5.1% * 2.9% *
% Positive 88.9% *** 77.8% *** 77.9% *** 80.5% ***
N = 45 45 149 149
The table presents average daily returns around the schedule 13D filing in excess of the market model using the CRSP value-weighted
index. I estimate parameters over the 200-day window prior to the Schedule 13D filing. Panel A compares activist investments to passive
investments. Panel B compares active investments seeking board seats to active investments that do not. I report p -values for medians using
the Wilcoxon Signed Rank Test. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels.
- 22 -
In the Short-Run, Active Investments Outperform Passive
Over a 61-Day Event Window
Figure 1. For active (blue) and passive (red) hedge fund investments, the chart plots the average excess return
around the schedule 13D filing, from 30 days prior to the filing until 30 days afterward. 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).
Passive
Active
Board
NoBoard
-5%
0%
5%
10%
15%
20%
-30 Days
-25 Days
-20 Days
-15 Days
-10 Days
-5 Days
13D Filing
5 Days
10 Days
15 Days
20 Days
25 Days
30 Days
using the CRSP value-weighted index. Both active and passive investments experience positive
and significant excess returns surrounding the filing date. Over 61 days, 80.4% of active
positions and 57.6% of passive positions show positive returns (z = 8.91 and z = 3.37). Figure 1
plots the average excess return, from 30 days prior to the Schedule 13D filing date to 30 days
afterward. Before the filing, between 10 days to 1 day prior, active positions jump by 4.6% and
passive positions by 2.1%. On filing day and the following day, active positions jump by 4.5%
and passive positions by 1.4%. Afterward, the abnormal returns trend up to 13.9% and 4.6% over
30 days. The results are consistent with Boyson and Mooradian (2008), Brav et al. (2008),
Clifford (2008), and Klein and Zur (2008).
To test whether activism differs from stock picking, Panel A in Table IV compares
activist filings to passive filings. Over 61 days, activism outperforms by 8.3% (t = 3.96). This
- 23 -
In the Short-Run, Active Investments Outperform Passive
Controlling For Undervaluation and Underperformance
Figure 2. The chart shows the average 5-day returns to active, passive, board, and no board
investments, controlling for Tobin's q and margins, measured by a dummy variable regression.
0.0%
2.0%
4.0%
6.0%
8.0%
10.0%
Active Passive Board No Board
supports Clifford (2008)’s results. Within active investments, those seeking board seats earn
significantly higher returns. Over 61-days, these investments outperform by 5.1% (t = 1.33).
Overall, it seems activism add value.
Still, far from a controlled experiment, these comparisons do not prove that activism adds
value. Activist targets might differ from passive targets in a way that makes them more likely to
outperform. In other words, hedge funds might simply call their best ideas “activist.” I use a
dummy variable regression to compare activist to passive investments, 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, passive, board, or no board investment. I regresses the 5-day abnormal return on activism
and board seat dummy variables, 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, I elaborate on these dummy variable regressions.
- 24 -
Table V
Abnormal Returns After 13D Filing Date
(1) (2) (3)
First Action First Board Seat First Board Win
(-30,+30) (-2,+2) (-30,+30) (-2,+2) (-30,+30) (-2,+2)
Median 7.5% *** 1.2% *** 6.3% *** 2.5% *** 5.2% *** 0.3%
Average 8.3% *** 4.4% *** 6.9% *** 3.4% *** 6.2% *** 0.8% *
% Positive 73.6% *** 66.0% *** 60.4% ** 77.1% *** 66.4% *** 53.2%
N = 53 53 48 48 110 109
The table presents average returns around events that occur after the initial Schedule 13D filing. I calculate returns in excess of the market model
using the CRSP value-weighted index. I estimate parameters over the 200-day window prior to the event. I report p -values for medians using the
Wilcoxon Signed Rank Test. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels.
proxy for underperformance.31 The results are robust to other models. Activist investments
outperform passive investments by 3.8% (t = 4.78), and board investments outperform other
activist investments by 3.2% (t = 1.71). Thus, activism adds value beyond stock picking.
(B) Purely Active Filings Show Excess Returns
In addition to comparing returns around Schedule 13D filings, I isolate the activist signal
in subsequent filings. Because the activist has already disclosed a 5% stake in the target firm, the
only new information that these subsequent filings convey is the activist’s intention to improve
the target firms.32 In addition, because activist hedge funds cannot fully control whether they win
board seats after they decide to seek them, 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.
To isolate these events, I read each investment’s entire SEC trail. I also conducted basic news
searches. I calculate abnormal returns around three events: (1) having previously filed a passive
13D, the first time a hedge fund announces activist intentions, (2) having previously filed an
activist 13D, the first time a hedge fund seeks board seats, and (3) having already announced its
intention to seek board seats, the first time a hedge fund wins board seats.
31 In Appendix D, Table D-3 presents the regression results. 32 If a hedge fund increased its exposure in any subsequent filing, however, it might convey some additional passive signal. But given that the hedge fund has already accumulated 5% of the company, this passive signal is of a small magnitude. Further, this critique does not apply to incidents where activist hedge funds win board seats.
- 25 -
Purely Active Filings Outperform
After Initial Schedule 13D
Figure 3. The table shows abnormal returns around events after the initial Schedule 13D. Having previously filed a
passive 13D, the blue line shows the hedge fund's first activist intention; having previously filed an activist 13D, the red
line shows the hedge fund's first time seeking a board seat; and having previously filed board seat 13D, the green line
shows the hedge fund's first time winning a seat.
FirstActivist
SeekBoard
WinBoard
-2%
0%
2%
4%
6%
8%
10%
-30 Days
-25 Days
-20 Days
-15 Days
-10 Days
-5 Days
13D Filing
5 Days
10 Days
15 Days
20 Days
25 Days
30 Days
Table V reports the 61- and 5-day abnormal returns surrounding these events.33 Figure 3
shows the returns over time. Target funds earn positive and significant excess returns
surrounding each event date. Over 61-days, 73.6%, 60.4%, and 66.4% of target firms enjoy
positive returns around each event (z = 3.76, z = 1.70, and z = 3.82), respectively. On average,
the firms outperform by 8.3%, 6.9%, and 6.2% (z = 3.38, z = 2.69 and z = 3.80), respectively.
Thus, the market reacts favorably to the activist signal, not just the stock picking signal.
(C) Returns Differ By Strategy
Just as important as average abnormal returns is the cross-sectional variation of those
returns. Using a regression approach, Table VI analyzes (1) what characteristics make a firm
most susceptible to improvement, (2) what instruments make an activist most effective, and (3)
what strategies add the most value. The dependent variable is the 5-day abnormal return
33 In Appendix D, Table D-4 reports more details.
- 26 -
Table VI
Cross-Sectional Variation of Short-Term Returns
(1) (2) (3)
Dependent Variable: Target Fundamentals Activist Instruments Activist Strategies
Abnormal Returns Coeff. t -stat Coeff. t -stat Coeff. t -stat
Constant 6.49 *** 6.98 - - - -
ln(MV) -0.53 -0.77 - - - -
Abnormal q -1.93 ** -1.87 -6.22 *** -2.47 -1.58 * -1.51
Abnormal Margin -0.23 ** -2.22 -0.64 *** -2.46 -0.25 ** -2.17
Leverage 0.00 -0.08 - - - -
BoardWin - - 6.69 0.99 - -
Board - - 4.30 ** 1.75 - -
Buyout - - 17.87 *** 4.30 - -
Financing - - 3.56 0.65 - -
Proxy - - 0.26 0.07 - -
Bus Strategy - - - - 2.68 * 1.31
M&A - - - - 1.63 1.26
Sale - - - - 7.41 *** 4.73
Cap Structure - - - - 2.92 * 1.53
Corp Governance - - - - -1.60 -0.85
Other - - - - 3.34 0.55
No. and R2
120 0.11 32 0.75 112 0.38
The dependent variable is the 5-day abnormal return around the Schedule 13D date. For all activist investments, Column 1 regresses
abnormal returns on target firm fundamentals. For all investments using activist instruments, Column 2 regresses abnormal returns
on dummy variables for those instruments. And for all investments using activist strategies, Column 3 regresses abnormal returns on
dummy variables for those strategies. All accounting variables are winsorized at the 10% level. All non-dummy covariates are
expressed as the deviation from the sample average values. In columns 2 and 3, intercepts are suppressed . *, **, and *** indicate
statistical significance at the 10%, 5%, and 1% levels.
surrounding the schedule 13D filing. I include as regressors (1) the firm’s abnormal q, the
logarithm of market capitalization, the firm’s industry-adjusted margins, and the firm’s debt-to-
assets ratio; (2) dummy variables for various activist instruments, including winning board seats,
seeking board seats, offering to buy the target firm, providing private financing, or launching an
additional proxy fight; and (3) dummy variables for the various activist strategies, including
business strategy, M&A, sale, capital structure, and corporate governance. For interpretation
purposes, I adapt the methodology in Brav et al. (2008). I express all non-dummy covariates as
the deviation from the sample mean, and in Column 2 and Column 3, I suppress the intercept
because of the dummy variables’ full span. Thus, we can interpret each dummy coefficient as the
- 27 -
partial return for an average target firm when the hedge fund uses that dummy variable. Both
Column 2 and Column 3 include dummy variables which are not mutually exclusive. For
example, if a hedge fund uses both Business Strategy and M&A activism, we expect the total
abnormal return to be 4.3% (= 2.7% + 1.6%).34
Turning first to target firm fundamentals, the market reaction indicates activists can
improve underperforming firms. In particular, activists earn the highest returns when targeting
firms with low abnormal q (t = -1.87) and with low industry-adjusted margins (t = -2.22). This is
what we should expect; if activists do add value, they’re more likely to succeed at firms with
more room to improve. Thus, not only do activists target undervalued and underperforming
firms, as documented in Section III, but they earn the highest returns while doing so, too.
I also find that activist instruments yield excess returns. Consistent with my earlier
findings, seeking board seats generates average abnormal returns of 4.3% (t = 1.75). Hedge funds
that win board seats generate 6.7% additional returns (t = 0.99), for 11.0% total excess returns.
Thus, the market seems to think that, through board seats, hedge funds can add value to target
firms. The market also responds very positively to hedge funds that offer to buy the entire firm,
driving target prices up 17.9% (t = 4.30). Whereas prior literature has noted hedge funds’
tendency to drive returns by putting firms in play, the findings presented here identify an
overlooked subtlety: rather than simply agitating for change, hedge funds put their own money
on the line.35 Often these buyouts often push targets into another acquirer’s arms—making them
a formidable weapon in the activist’s arsenal. But when the hedge fund does purchase the target,
stock market returns no longer proxy for the hedge fund’s return, which depends on what
happens after the buyout. Still, these buyout offers generate large returns for shareholders.
34 In Appendix A, I describe more fully this dummy variable regression. 35 See Greenwood and Schor (2007) and Brav et al. (2008).
- 28 -
In the Short-Run, Selling the Firm Generates Highest Returns
Controlling for Undervaluation and Underperformance
Figure 4. The chart shows the average 5-day abnormal return for each strategy, defined as the regression
coefficients. The bars show the 90% confidence interval.
-5%
0%
5%
10%
Business Strategy M&A Sale Capital Structure Corporate
Governance
Figure 4 shows the average return to each activist strategy. These strategies generate
statistically different returns (F = 4.03). Sale activism generates 7.4% excess returns (t = 4.73).
Business strategy and capital structure activism also generate significant abnormal returns of
2.7% (t = 1.31) and 2.9% (t = 1.53), which implies that activism can improve its targets. While
consistent with Brav et al. (2008), these results contradict Boyson and Mooradian (2008), who
find that corporate governance activism generates large returns. This is due to their longer time
frame. Indeed, when I include investments from 2000 through 2003 by substituting regular data
for the industry-adjusted data, I find that capital structure activism generates positive returns.
Perhaps due to improving governance, this strategy has become less successful over time.
Additionally, because both Column 2 and Column 3 control for abnormal q and abnormal
margin, the coefficients reflect activism’s effects, not just target firms’ fundamentals. This
thoughtful variation provides further evidence that shareholders expect activism to generate
- 29 -
Short-Term Returns Decline Over Time
As Investments Increase in �umber
Figure 5. The line graphs (left axis) show show the average 5-day returns by 13D year, as measured
by the coefficients in a dummy variable regression that controls for Tobin's q and margins. The blue
line shows the average activist return, and the red line shows the average passive return. 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.
Passive
Active
-5%
0%
5%
10%
15%
20%
25%
2000
2001
2002
2003
2004
2005
2006
2007
2008
0
30
60
90
120
150
180
Active Filings Passive Filings
value; if shareholders were purchasing simply on expectations of superior passive investing, they
would not parse a hedge fund’s press releases to discern its intentions.
(D) Returns Have Decreased Over Time
Finally, I measure how these returns have changed over time. I regress the 5-day
abnormal returns against dummy variables for each filing year, controlling for Tobin’s q and
margins. The results are robust to other models. All non-dummy covariates are expressed as the
deviation from the sample mean, and constants are suppressed. I reject the hypothesis that annual
returns equal each other (F = 2.14). To the contrary, returns have declined; in 2001, the average
firm earned 23.1%, but in 2008, the average firm earned only 5.9%.
In Figure 5, the blue lines plots the average activist returns to Schedule 13Ds filed each
year (left axis). The bar graph shows the number of Schedule 13Ds filed each year (right axis).
Consistent with Puustinen (2007), I find that short-term activist returns have declined over time.
- 30 -
This pattern suggests three interpretations. First, as activism becomes more common,
shareholders might begin to expect it. Thus, when hedge funds actually do intervene, investors
are less surprised. Second, as companies have learned to accept activism, hedge funds have shed
their most aggressive tactics. Thus, outside observers might confuse true activism for passive
investing. As evidence against this hypothesis, however, passive returns (red line) have not
demonstrably improved; in fact, I cannot reject the null hypothesis that the coefficients all equal
each other (F = 0.48). Finally, activists may simply have lost their edge, putting money to work
in increasingly less attractive situations. In Section VI, however, I find no evidence that activists
have lost their ability to improve target firms.
In summary, I find substantial evidence that activism adds value. Activist investments
outperform passive investments, adjusting for undervaluation and underperformance. In addition,
investors applaud hedge funds’ activist intentions, even if the hedge funds have already disclosed
passive investments in the same targets. Activism aimed at undervalued firms and attempts to
sell target firms earn the highest returns. Recently, though, these returns have declined.
Section V: Long-Term Stock Returns
In the Intelligent Investor, Benjamin Graham (1973) writes, “In the short run, the market
is a voting machine. In the long run it is a weighing machine.” Using what Graham might
consider an exit poll, Section IV argued that investors believe activism adds value. This section
builds on that work to examine whether the weighing machine agrees. After all, if activism fails
to generate long-term excess returns, neither corporate shareholders nor hedge funds’ clients
benefit; activism represents wasted energy at best and value destruction at worst.
- 31 -
If hedge fund activism improves its targets beyond the market’s pre-targeting
expectations, then target firms should outperform their risk-adjusted benchmarks. In addition, we
should expect these investments to continue to outperform as the activism persists. At every
point in time, the market should bid stock prices to a level reflecting the expected probability that
the activist will succeed (Brav et al. 2008). Because success is never certain, as the activist
continues to wage a campaign, the market should continue to reward the effort. 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.
As in the short-term, however, the efficient markets hypothesis proves crucial to the
argument. Instead, if hedge funds could persistently pick winning stocks without improving
them, then abnormal returns might simply mark great passive investing, not activist ability.
In this section, using two tests, I conclude that activists add value in the long-run. First, I
compare activist filings to passive filings. I find strong evidence that activist investments,
particularly those which seek board seats, outperform passive investments. Even excluding the
61-days surrounding the filing, activist investments earn significantly higher returns. Second, I
compare activist filings to passive filings, controlling for undervaluation and underperformance.
Again, activist investments outperform passive investments—particularly when the hedge fund
seeks board seats. Thus, activism appears to add long-term value.
I also investigate heterogeneity across these returns. In particular, I investigate when
activists add the most value—what target firm characteristics, activist instruments, or activist
strategies prove most effective? As in the short-run, undervalued, underperforming target firms
earn the highest long-term returns, likely because activism has rejuvenated them. Hedge funds
- 32 -
earn excess returns when they sell target firms, improve targets’ business strategies, or optimize
targets’ capital structures. Finally, I observe that these returns have declined over time.
Before turning to the analysis, however, it is worth elaborating on the methods I will use.
The major obstacle to analyzing long-term returns is picking a time period. Brav et al. (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. Instead, to better measure the reward to corporate shareholders (at the expense of
the returns to hedge fund investors), I analyze long-term returns along arbitrary windows. To the
extent activist hedge funds adopt different time frames for different investments, this method
should bias observed abnormal returns downward. In particular, once target firms realize their
potential, their abnormal returns should trend towards zero. Thus, any observed effect, whether
positive or negative, is likely understated. Further, because this downward bias affects passive
investments less than it affects activist investments, any observed difference between active and
passive investments is similarly understated.36 Even in spite of this, I still find that active
investments both generate positive abnormal returns and outperform passive investments.
Long time horizons also complicate how we categorize activism. Over 5 or even 61 days,
we can isolate a hedge fund’s intentions, but over 2 years, hedge funds revise their strategies
numerous times. I find that hedge funds significantly change strategies in 79.5% of investments.
To most purely match the hedge fund’s behavior to the relevant investment period, I distill my
sample as follows: passive investments include all instances where a hedge fund never amends
its passive Schedule 13D, activist investments include all instances where a hedge fund
announces an activist instrument or an activist strategy within 180 days of its Schedule 13D, no
36 Unlike activist investments, passive investments have no catalyst. Thus, they have no holding period sensitivity.
- 33 -
Table VII
Long-Term Returns by Investment Type
Panel A: Active vs. Passive
Active Passive Difference
(-1,+24) (+1,+24) (-1,+24) (+1,+24) (-1,+24) (+1,+24)
Median 20.1% *** 12.1% *** 10.4% * 5.3% 9.7% 6.8% *
Average 21.3% *** 13.5% *** 3.7% 1.5% 17.6% ** 11.9% *
% Positive 66.8% *** 61.1% *** 55.1% ** 53.7% ***
Panel B: Board Seats vs. Other Active Investments
Board Seats No Board Seats Difference
(-1,+24) (+1,+24) (-1,+24) (+1,+24) (-1,+24) (+1,+24)
Median 28.2% *** 18.3% *** 13.0% *** 7.6% 15.2% ** 10.8% **
Average 31.2% *** 21.6% *** 14.3% *** 7.8% *** 16.9% * 13.8% *
% Positive 75.2% *** 66.4% *** 60.9% *** 58.0% **
The table presents average monthly returns around the Schedule 13D filing in excess of the market model using the CRSP value-weighted
index. I estimate parameters over the 36 months prior to the Schedule 13D. Panel A compares activist investments to passive investments.
Panel B compares activist investments seeking board seats to activist investments that do not. I report p -values for medians using the
Wilcoxon Signed Rank Test. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels.
board seat investments include all activist investments where a hedge fund never seeks board
seats, and board seat investments include all activist investments where a hedge fund seeks a
board seat within 180 days of its Schedule 13D.
(A) Activism Exceeds Stock Picking
Table VII measures long-term returns. 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. In Appendix D, Table D-5 shows more details. Consistent with short-term
results, both activist and passive investments earn positive and significant excess returns. Over
the entire 25-month period, 66.8% of activist investments and 55.1% of passive investments
show positive returns (z = 5.70 and z = 2.08). Figure 6 plots the excess returns over time. From
one month before the Schedule 13D until one month after, activist investments earn 10.3%
excess returns and passive investments earn 2.8%. After the short-term reaction subsides, from
one month after the Schedule 13D until 12 months after, activist investments flourish while
passive investments falter: activist investments increase an additional 8.5% (z = 3.87) while
- 34 -
In the Long-Run, Active Investments Outperform Passive
Over 25-month Event Window
Figure 6. For active (blue) and passive (red) hedge fund investments, the chart plots the average
excess return around the schedule 13D filing, from 1 month prior to the filing until 24 months
afterward. 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).
Passive
Active
Board
NoBoard
-10%
0%
10%
20%
30%
40%
-1 Month
13D Filing
1 Month
3 Months
6 Months
9 Months
12 Months
15 Months
18 Months
21 Months
24 Months
passive investments fall 0.4% (z = -0.29). Over a longer window, from one month after the
Schedule 13D until 24 months after, activist investments earn 13.5% excess returns (z = 5.55)
while passive investments earn 1.5% (z = 0.16). These results are consistent with Boyson and
Mooradian (2008), Brav et al. (2008), and Klein and Zur (2008).
The increasing returns imply that over two years, the market continues to receive positive
information. This makes sense; to the extent activists do improve their targets, these changes
play out slowly over time—and so should returns. In contrast, passive investors’ subsequent
actions matter very little. As passive investing reveals itself all at once, so should returns.
It’s also worth noting that this activist momentum is not a free lunch. First, the strategy is
not investible because certain investments reveal their activism after the event date. Additionally,
activism might carry more risk than passive investing. By estimating target firms’ factor loadings
- 35 -
In the Long-Run, Active Investments Outperform Passive
Controlling For Undervaluation and Underperformance
Figure 7. The chart shows the average 25-month returns to active, passive, board, and no board
investments, controlling for Tobin's q and margins, measured by a dummy variable regression.
0.0%
10.0%
20.0%
30.0%
40.0%
Active Passive Board No Board
before the hedge funds launch their activism, I underestimate the target firms’ risk and thus their
benchmark returns. 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, I compare activist returns
to passive returns. Panel A in Table VII shows that over 25 months, activism outperforms
passive investing by 17.6% (t = 2.28). Of this outperformance, 11.9% comes after the short-term
reaction (t = 1.61). Panel B shows that board seat activism outperforms other activism by 16.9%
over the full period (t = 1.57) and by 13.8% after the short-term reaction (t = 1.31). Overall,
activism seems to add value.
To more cleanly distinguish activism from stock picking, I use a dummy variable
regression to control for the possibility that hedge funds might go activist in more attractive
targets. I regress the 25-month return on dummy variables for activism and board seat activism,
- 36 -
Table VIII
Cross-Sectional Variation of Long-Term Returns
(1) (2) (3)
Dependent Variable: Target Fundamentals Activist Instruments Activist Strategies
Abnormal Returns Coeff. t -stat Coeff. t -stat Coeff. t -stat
Constant 8.72 * 1.60 - - - -
ln(MV) 1.91 0.47 -23.63 *** -2.45 -16.10 * -1.56
Abnormal q -10.11 ** -1.65 -0.14 -0.16 -0.48 -0.42
Abnormal Margin -1.14 ** -1.88 - - - -
Leverage 0.04 0.14 - - - -
BoardWin - - -20.68 * -1.31 - -
Board - - 39.86 *** 3.12 - -
Buyout - - -15.93 -0.88 - -
Financing - - -8.82 -0.31 - -
Proxy - - -14.87 -1.16 - -
Law - - -48.01 ** -1.91 - -
Bus Strategy - - - - 13.90 0.67
M&A - - - - 4.70 0.38
Sale - - - - 0.62 0.04
Cap Structure - - - - 12.94 0.71
Corp Governance - - - - -8.84 -0.46
Other - - - - -51.29 -0.93
No. and R2
243 0.03 90 0.21 101 0.08
The dependent variable is the 25-month abnormal return around the Schedule 13D date. For all activist investments, Column 1
regresses abnormal returns on target firm fundamentals. For all investments using activist instruments, Column 2 regresses abnormal
returns on dummy variables for those instruments. And for all investments using activist strategies, Column 3 regresses abnormal
returns on dummy variables for those strategies. All accounting variables are winsorized at the 10% level. All non-dummy
covariates are expressed as the deviation from the sample average values. In columns 2 and 3, intercepts are suppressed . *, **, and
*** indicate statistical significance at the 10%, 5%, and 1% levels.
controlling for Tobin’s q and margins.37 The results are robust to other models. Activism
outperforms passive investing by 18.4% (t = 2.17), and board investments outperform other
activism by 18.6% (t = 1.59). Thus, activism adds value beyond stock picking.
(B) Returns Differ By Strategy
As with short-term returns, I investigate the conditions that make activism most effective.
In Table VIII, I regress 25-month abnormal returns on (1) target firm fundamentals, (2) activist
instruments, and (3) activist strategies. Turning first to target firm fundamentals, activists earn
larger returns from undervalued targets and underperforming targets. Consistent with short-term
37 In Appendix D, Table D-6 shows the underlying regression.
- 37 -
returns, activists perform best when firms have low abnormal q (t = -1.65) and low abnormal
margins (t = -1.88). The evidence supports the hypothesis that activists do their best work turning
around unloved, unprofitable target firms.
With respect to activist instruments, board seats continue to drive value in the long-run.
Over 25-months, investments that seek board seats earn 39.9% excess returns (t = 3.12), far
exceeding any other activist instrument. Consistent with short-run expectations, this finding
strongly supports the hypothesis that activists can add value; by seeking board seats, they ensure
that management hears their message. Proxy fights and lawsuits generate moderately significant
negative returns of -14.9% (t = -1.16) and -48.0% (t = -1.91), respectively. In these hostile
situations, activist hedge funds might distract management from running the firm. Alternatively,
the hedge fund might be running into a brick wall—a stubborn management team that refuses to
cooperate with value-increasing propositions.
Contrary to expectations, within investments that seek board seats, those which win board
seats actually underperform those which do not win board seats by 20.7% (t = -1.31). 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. First, this coefficient is biased because
the regression excludes investments initiated from 2000 through 2003. When I increase the
sample by substituting Tobin’s q for abnormal q and substituting margin for abnormal margin,
the effect becomes positive; winning adds 6.4% (t = 0.40) to investment returns.38 Second, this
coefficient is biased by hedge funds that withdraw their requests after management cooperates,
making the investment a success. In fact, of all failed board contests in my sample, in 28.3% of
cases the hedge fund voluntarily withdraws its request. After removing withdrawn board seats
38 Part of this increase is due to reduced precision in the covariates. To corroborate that sample size plays a role, in unreported results I run the new regression for the 2004-2008 period and for the 2000-2008 period. As expected, the longer period shows a larger coefficient on winning board seats.
- 38 -
Many Strategies Generate High Returns
Controlling for Undervaluation and Underperformance
Figure 8. The chart shows the average 25-day abnormal return for each strategy, defined as the regression
coefficients. The bars show the 90% confidence interval.
-40%
-20%
0%
20%
40%
Business Strategy M&A Sale Capital Structure Corporate
Governance
from the original model, the effect of winning becomes positive. Winning board seats increases
returns by 7.2% (t = 0.28). Thus, it appears activists use board seats as a bargaining chip,
dropping the board contest once management acquiesces to their demands.
In addition to activist instruments, multiple activist strategies earn positive returns. In
fact, different strategies’ returns are statistically indistinguishable from one another (F = 0.20).
This does not itself disprove the hypothesis that activists follow different strategies; if strategies
differ, they should differ in fundamental performance, not necessarily returns. Figure 8 shows the
average return per strategy. As in the short-run, investments which change the target’s business
strategy and capital structure perform very well, earning 13.9% excess returns (t = 0.67) and
12.9% excess returns (t = 0.71), respectively. Also consistent with short-run expectations,
corporate governance activism underperforms the market by 8.9% (t = -0.46). When broadening
the sample by substituting Tobin’s q for abnormal q and margin for abnormal margin, however,
this effect disappears. Instead, corporate governance activism outperforms by 9.5% (t = 0.57).
- 39 -
Activism Adds Value Beyond Selling Companies
But Selling Generates Very High Returns
Figure 9. The chart plots the average excess return around the schedule 13D filing, from 1 month prior to the filing
until 24 months afterward. The chart plots active investments only. The solid line includes all active investments.
The long dashes show only investments which were subsequently sold. The short dashes excludes any investment
which was sold as well as any investment in which the activist sought a sale but failed.
All Firms
Sold Firms
All Firms
ex. Sale
0%
10%
20%
30%
40%
-1 Month
13D Filing
1 Month
3 Months
6 Months
9 Months
12 Months
15 Months
18 Months
21 Months
24 Months
Most shockingly, sale activism, which generated the highest short-term returns, generates
no long-term returns. This result is most likely due to failed sales. Especially if activists try to
sell wounded companies, those remaining unsold will dramatically underperform. To counter
this effect, in unreported results, I run the same regression with a dummy variable set to one if
the company was sold. As expected, successful sales earn positive returns (9.3% with t = 0.48).
To test Greenwood and Schor’s (2008) hypothesis that activists add value exclusively by
putting firms in play, I re-run the event study excluding those firms which the hedge funds try to
sell. Unlike Greenwood and Schor, who asymmetrically remove only successful sales, I remove
both successful and failed sales. By eliminating the bias from failed sales, this better measures
- 40 -
Long-Term Returns Decline Over Time
As Investments Increase in �umber
Figure 10. The line graphs (left axis) show show the average 25-month return by year, as measured
by the coefficients in a dummy variable regression controlling for Tobin's q and margins. The blue
line plots the average active return, and the red line plots the average passive return. 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.
Active
Passive-25%
25%
75%
125%
175%
2000
2001
2002
2003
2004
2005
2006
2007
2008
0
45
90
135
180
Active Filings Passive Filings
hedge funds’ ability to add value beyond selling the target. In stark contrast to Greenwood and
Schor, I find that activism still generates significant excess returns. Figure 9 shows that while
they underperform firms which were sold, these targets still generate 22.3% returns over 25
months (t = 5.79). This result differs from Greenwood and Schor both because I exclude failed
sales and because my sample includes more experienced hedge funds.
(C) Returns Have Decreased Over Time
Finally, I measure how these returns have changed over time. I regress the 25-month
abnormal returns against dummy variables for each filing year, controlling for Tobin’s q and
margins. The results are robust to other models. Figure 10 plots the results over time. I reject the
hypothesis that annual returns equal each other (F = 2.56). To the contrary, returns have declined
over time; in 2001, the average investment earned 23.1%, and in 2007 the average investment
earned only 7.6%. This pattern suggests the same three interpretations: (1) investors might
- 41 -
increasingly expect activism, (2) activism might begin to look more like passive investing, and
(3) activists might have lost their edge. As evidence against the second hypothesis, however, I
cannot reject the hypothesis that passive returns equal each other in every year (F = 0.27).
In summary, I find further evidence that activism adds value. Over the long-run, activist
investments outperform passive investments, even adjusting for undervaluation and
underperformance. Activism aimed at undervalued and underperforming firms earns the greatest
excess returns. Recently, however, these returns have declined.
Section VI: Target Firm Fundamentals
Having observed that activism earns greater returns than stock picking might imply, we
now turn our attention to target firm fundamentals. By adding further evidence that activism adds
value, this section strengthens earlier results. Perhaps more importantly, it also describes the
levers activists pull to earn those returns.
I document four non-mutually exclusive ways that activist hedge funds improve target
firms. First, activists improve profitability. Both relative to passive targets and in absolute terms,
activist targets increase their margins, return on assets, and return on equity. Second, activists
optimize the capital structure, often by disgorging cash. Activist targets increase their leverage
and payout ratios. Third, activists reduce firm size. While passive target firms actually grow,
activist target firms shrink dramatically. Fourth, activists sell target firms. Relative to passive
targets, activist targets sell themselves more frequently and for higher premiums. Throughout, 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.
- 42 -
Before turning to the data, a cautionary note is in order. Identifying firms which
outperform by accounting standards presents a more difficult task than identifying those which
outperform stock market indices. The beauty of financial markets lies in their ability to correctly
price target firms’ fundamentals, not in their ability to equalize those fundamentals; no efficient
markets hypothesis governs accounting performance. Thus, 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. To the contrary, by simply investing in stocks with high P/E ratios,
for instance, a hedge fund’s portfolio might regularly outpace its peers in revenue growth. Thus,
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.
Three factors mitigate this stock picking hypothesis. First, activists target well-
performing firms.39 To the extent certain companies can improve more than others, it shouldn’t
be well-performing firms but rather underperforming firms with more room to improve.40
Second, activists target undervalued companies. The companies expected to outperform should
not be undervalued but rather overvalued to account for yet-unrealized profits. And third, excess
returns increase over two years. If the market had predicted that target firms would improve,
excess returns should not rise over time but rather should equal zero. Thus, hedge funds don’t
seem to simply pick stocks that everyone expects will increase profitability.
In addition to arguments about target firms’ characteristics, I further debunk the stock
picking hypothesis by comparing hedge funds’ active targets to the same hedge funds’ passive
targets. If active and passive targets come from the same underlying pools, then hedge fund
activism must cause any outperformance by the activist group. As discussed earlier, though, the
39 As noted in Section III, relative to passive targets, activist targets are poor performers, but relative to the median firms in their industries, activist targets are good performers. 40 I find significantly negative correlations between initial levels and subsequent improvements.
- 43 -
Table IX
Profitability Improvement By Investment Type
Panel A: Active vs. Passive
Active Passive Difference
1 yr 2 yr 1 yr 2 yr 1 yr 2 yr
Margin -0.44 0.73 -1.20 -3.38 0.76 4.11 ***
ROA 0.27 2.28 -1.84 -2.87 2.11 *** 5.15 ***
ROE 2.48 8.29 -4.73 -8.70 7.21 *** 16.99 ***
Panel B: Board Seats vs. Other Active Investments
Board Seats No Board Seats Difference
1 yr 2 yr 1 yr 2 yr 1 yr 2 yr
Margin 0.21 0.97 -1.59 -0.08 1.80 ** 1.05
ROA 0.54 2.41 -0.33 1.02 0.86 1.39
ROE 0.95 6.13 2.92 8.71 -1.97 -2.58
The table shows the 1- and 2-year increases, from the date of the first Schedule 13D filing, in target firms' trailing twelve
months EBITDA divided by sales (Margin) , EBITDA divided by total assets (ROA) , and EBITDA divided by book
value of equity (ROE) . All data from Bloomberg. All data is winsorized at the 10% level. *, **, and *** indicate
statistical significance at the 10%, 5%, and 1% levels.
pools differ. Most noticeably, hedge funds go active in more undervalued companies. For the
reasons discussed above, this should handicap activist targets’ ability to outperform. Because I
nonetheless observe outperformance, this does not appear to be a problem. Additionally, hedge
funds go active in worse performing companies. While this might give activist investments more
room to improve, after controlling for this effect, I still observe that activist targets outperform.
(A) Activism Improves Profitability
By improving target firms’ capital allocation and strategic positioning, hedge funds
increase cash flow—and thus, firm value. Management might fail to do so for many reasons. For
example, 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. Just as bad, CEOs paid based on
ROA might forgo projects with attractive IRRs if they drag down a firm’s overall returns.
- 44 -
Active Investments Become More Profitable
Controlling For Initial Levels
Return on Assets 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.
-2.0%
-1.0%
0.0%
1.0%
2.0%
3.0%
Active Passive Board No
Board
-3.0%
-2.0%
-1.0%
0.0%
1.0%
2.0%
Active Passive Board No
Board
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.
- 45 -
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-to-
assets 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.
- 46 -
Table X
Capital Structure Changes By Investment Type
Panel A: Active Investments Lever Capital Structure More Than Passive Investments
Active Passive Difference
1 yr 2 yr 1 yr 2 yr 1 yr 2 yr
Leverage 0.64 1.51 -0.45 1.56 1.09 ** -0.05
% Increase 43.6% 45.3% 35.4% 42.2% 8.2% 3.1%
Payout 5.37 0.53 0.91 10.00 4.46 ** -9.47 **
% Increase 32.6% 18.5% 12.7% 16.7% 19.9% *** 1.9%
Panel B: Board Investments Lever Capital Structure Less Than Non Board Investments
Board Seats No Board Seats Difference
1 yr 2 yr 1 yr 2 yr 1 yr 2 yr
Leverage -0.75 0.72 2.08 2.86 -2.83 ** -2.15
% Increase 33.8% 40.0% 58.1% 55.0% -24.2% ** -15.0%
Payout 7.36 0.76 3.58 -0.75 3.78 1.51
% Increase 26.7% 17.6% 40.0% 20.0% -13.3% -2.4%
The table shows the 1- and 2-year increases, from the date of the first Schedule 13D filing, in trailing twelve months debt-
to-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.
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.
- 47 -
Active Investments Increase Payout
Controlling For Initial Levels and Initial Cash
Payout Ratio Leverage
Figure 12. The chart shows the average 1-year change in payout ratio and in the debt-to-assets ratio for active,
passive, board, and no board investments, controlling for Tobin's q , initial levels of each variable, and initial
cash holdings.
0.0%
1.0%
2.0%
3.0%
Active Passive Board No
Board
-1.0%
0.0%
1.0%
2.0%
Active Passive Board No
Board
Figure 12 shows that, once again, these results are due to activism, not to stock picking. I
control for Tobin’s q, the initial value of each accounting metric, and the initial level of cash as a
percentage of total assets.43 While the average passive firm marginally increases its payout by
2.3% (t = 1.79) and decreases its leverage by 0.4% (t = -0.83), the average active firm increases
both payout and leverage. Payout increases by 2.5% more than passive firms (t = 1.26), and
leverage increases by 0.7% more than passive firms (t = 0.98). Thus, it appears activism earns
some of its gains by disgorging cash and leveraging the capital structure.
(C) Activism Reduces Firm Size
Activist shareholders also add value by counteracting management’s harmful tendency to
build empires. For CEOs, larger firms lead to greater compensation and greater job security. To
grow the firm, they might purchase assets which yield suboptimal returns. Thus, by divesting
unprofitable assets activists can improve future returns.
43 In Appendix D, Table D-8 shows more detail.
- 48 -
Table XI
Agency Cost Reductions By Investment Type
Panel A: Active Investments Shrink More Than Passive Investments
Active Passive Difference
1 yr 2 yr 1 yr 2 yr 1 yr 2 yr
Assets -4.8% -6.5% 0.0% 0.1% -4.8% *** -6.6% ***
Sales -2.4% 0.0% 1.7% 1.4% -4.2% *** -1.4%
Cash 22.1% 35.2% 30.5% 24.2% -8.3% 11.0%
Panel B: Board Investments Shrink More Than Non-Board Investments
Board Seats No Board Seats Difference
1 yr 2 yr 1 yr 2 yr 1 yr 2 yr
Assets -7.0% -9.6% -1.4% 0.5% -5.6% *** -10.1% ***
Sales -4.6% -3.1% 0.3% 5.3% -5.0% *** -8.5% **
Cash 26.0% 27.9% 20.6% 50.0% 5.4% -22.1%
The table shows the 1- and 2-year increases, from the date of the first Schedule 13D filing, in inflation-adjusted firm size,
measured by assets, sales, and cash. Each variable is expressed as a percentage of its value in the year prior to the
Schedule 13D filing. All data from Bloomberg. All data is winsorized at the 10% level. *, **, and *** indicate statistical
significance at the 10%, 5%, and 1% levels.
Table XI examines this hypothesis. It shows assets, sales, and cash, adjusted for core
inflation, expressed as a percent change from their values in the year prior to activism. Panel A
shows that while passive target firms grow, active target firms actually shrink. On average,
active targets reduce their assets by 4.8% more than passive targets over one year (t = -3.23) and
by 6.6% more over two years (t = -2.36). Similarly, activist targets reduce their sales by 4.2%
more than passive firms (t = -2.89) over one year and by 1.4% more (t = 0.58) over two years.
While active targets also appear to reduce their cash by 8.3% more than passive firms over one
year (t = 1.04), they appear to increase their cash by 11.0% more over two years (t = 1.11). The
inconsistent evidence likely reflects two conflicting forces: on one hand, activist target firms are
shrinking, but on the other hand, they are also increasing profitability.
Panel B shows that board seat investments shrink the most. Investments where activists
seek board seats shrink their assets by 5.6% more than other activist investments (t = -2.80) over
- 49 -
Active Investments Reduce Assets
Controlling For Initial Assets
Asset Growth Sales Growth
Figure 13. The chart shows the average 2-year change in assets and sales, adjusted for inflation, for active,
passive, board, and no board investments, controlling for Tobin's q and for the initial firm size.
-4.0%
-2.0%
0.0%
2.0%
4.0%
Active Passive Board No
Board
0.0%
1.5%
3.0%
4.5%
6.0%
Active Passive Board No
Board
one year and by 10.1% more (t = -2.55) over two years. They also shrink their sales by 5.0%
more (t = 1.04) over one year and by 8.5% more (t = 1.97) over two years.
To isolate activism from stock picking, Figure 13 controls for differences Tobin’s q and
in firm size.44 While passive targets actually increase assets by 2.8% (t = 1.81), active
investments decrease assets by 1.8%, shrinking 4.6% more (t = -2.12). Similarly, active
investments without board seats actually increase assets by 0.8% (t = 0.34), but board seat
investments decrease assets by 3.7%, shrinking 4.4% more (t = -1.49). The changes in sales show
similar patterns; while all investments grow their sales, the activist investments seeking board
seats grow the least. Taken together with the earlier evidence that activist firms increase
profitability, it appears that activist investments purposefully pare down, rather than
unintentionally shrink into irrelevance. Thus, activist investments, particularly those with board
seats, effectively reduce firm size. As Jensen (1986) argues, this decreases agency costs.
44 In Appendix D, Table D-9 shows more details.
- 50 -
Table XII
Target Firm Sales By Activism Type
Panel A: Active Investments Sold More Frequently
Active Passive Difference
6 mo 24 mo 6 mo 24 mo 6 mo 24 mo
% Sold 9.2% 17.9% 0.0% 4.2% 9.2% *** 13.7% ***
Avg. Premium 20.4% 39.8% - 21.5% - 18.3% *
�umber Sold 24 47 0 9
Panel B: Investments With Board Seats Sold Less Frequently, But For Higher Prices
Board Seats No Board Seats Difference
6 mo 24 mo 6 mo 24 mo 6 mo 24 mo
% Sold 2.2% 10.4% 19.3% 30.3% -17.0% *** -19.8% ***
Avg. Premium 44.0% 57.9% 16.8% 31.9% 27.1% *** 26.1% **
�umber Sold 3 14 21 33
The table shows the percent of all firms which sold themselves and the average premiums they received. Average premium is
calculated from one month before the hedge fund files a Schedule 13D until the date on which the sale closes. Panel A separates
active investments from passive investments. Panel B separates active investments seeking baord seats from those which do not. *,
**, and *** indicate statistical significance at the 10%, 5%, and 1% levels.
(D) Activism Sells Target Firms
Finally, activists add value by selling their targets to acquirers who values them more
highly. Executive managers might fail to sell the firm, even if it benefits shareholders, because
doing so imperils their job security.
Table XII measures how well activists sell their targets. First, it measures the percent of
targets which sell themselves, and second, it measures the average premium those firms fetched.
To capture the premium attributable to the activist, I measure the total buy-and-hold gross return
from one month before the Schedule 13D until the sale closes. Panel A shows that, relative to
their passive investments, hedge funds’ activist targets sell themselves more frequently and for
higher prices. Within the first 6 months, activists sell 9.2% of their targets, while passive
investors sell none. Over two years, activist investors sell 13.7% more targets (t = 4.93) and on
average sell them for 18.3% higher premiums (t = 1.64). Thus, activists both put their targets into
play and market them better than do passive investors. Panel B shows that when activists seek
- 51 -
board seats, they sell firms less frequently but for much higher prices. Again, this result is
consistent with the interpretation that activists use board seats to build long-term businesses.
When doing so, they’re less likely to sell—in particular, 19.8% less likely over two years (t =
3.77). But when they do sell, having built a strong business, they demand a healthy premium—
on average, 26.1% larger (t = 2.10). In total, relative to passive investments, activism unlocks
substantial value by selling the target firm.
(E) Improvements Differ by Strategy
If hedge fund activism, rather than sheer luck, catalyzes these observed changes, we
should expect activists’ professed strategies to predict target firms’ improvements with
reasonable accuracy. This section tests that hypothesis. In particular, I describe the differences
among the various activist strategies.
To describe these strategies, I run cross-sectional regressions like those in Section IV and
Section V. For all investments which use activist strategies, I regress the accounting statistics
described earlier on dummy variables for each strategy. When applicable, as described earlier in
this section, I control for the dependent variable’s initial value. To facilitate interpretation, all
non-dummy covariates are expressed as the deviation from the sample mean, and all constants
are suppressed. Thus, we interpret the coefficients on each dummy variable as that strategy’s
partial effect on a firm with average characteristics. In general, I find strong evidence that
different strategies work in different ways. Figure 14 shows the key results.45
Business strategy activism builds profitable businesses. Over two years, ROA increases
6.9% (t = 2.71) and margins increase 2.3% (t = 0.91). Perhaps to remove distractions from
building these businesses, activists also reduce target firms’ payout ratios by 10.0% (t = -1.78)
45 In Appendix D, Table D-10, Table D-11, Table D-12, and Table D-13 show more detail.
- 52 -
Panel A: Business Strategy Activism Improves ROA
Panel B: Capital Structure Activism Increases Payout
Panel C: Corp Gov & Sale Activism Decrease Assets
F = 2.74
-6%
0%
6%
12%
Business Strategy M&A Sale Capital Structure Corporate
Governance
F = 1.39
-20%
-10%
0%
10%
20%
Business Strategy M&A Sale Capital Structure Corporate
Governance
F = 1.11
-10%
-5%
0%
5%
10%
Business Strategy M&A Sale Capital Structure Corporate
Governance
- 53 -
Panel D: M&A Activism Sells Firms Most Frequently
Corporate Governance Activism Earns Highest Premiums
Figure 14. These charts plot the results of cross-sectional regressions on various accounting metrics: ROA (Panel A), payout
ratios (Panel B), and asset growth (Panel C). I also show cross-sectional variation of firm sale statistics (Panel D). For all
activist investments which use activist strategies, these dependent variables are regressed on dummy variables for each
strategy. Where applicable, regressions control for the dependent variable's initial level. All non-dummy covariates are
expressed as the deviation from the sample mean, and all constants are suppressed. Panel A, Panel B, and Panel C show 90%
confidence intervals and F statistics on the null hypothesis that all strategies generate equal improvements. In Panel D, 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. The red outlines (right axis) show a linear regression that regresses the average
premium on the strategy dummies.
Sale Odds
(left axis)
Sale Premium
(right axis)
0%
50%
100%
150%
200%
250%
Bus Strat M&A Sale CapStruc CorpGov
Sale Odds Ratio
0%
10%
20%
30%
40%
50%
Averarge Premium
and leverage by 1.2% (t = -0.73). Presumably due to increased profitability and lower payout
ratios, target firms’ assets increase 3.2% (t = 0.77) and sales increase 4.6% (t = 1.14).
M&A and sale activism effectively push target firms to sell themselves; M&A activism
sells firms more frequently, and sale activism sells them for higher prices. Within investments
that use activist strategies, those in which the hedge fund uses a M&A-related strategy are 2.3
times more likely to sell themselves as those which do not (t = 1.74). When a hedge fund
demands that a target firm sell itself, it earns on average a 24.6% premium (t = 2.04). When
hedge funds fail to sell these firms, they often seek divestitures, reducing total assets by 4.5%
over two years (t = -1.43).
- 54 -
Capital structure activism increases cash to investors. Target firms increase their payout
ratio by 10.4% (t = 2.21). Interestingly, leverage actually falls by 1.8% (t = -1.8), perhaps
reflecting heterogeneity even within the strategy—while some hedge funds want the firm to
leverage itself, others want the firm to de-lever a precarious balance sheet.
Corporate governance activism reduces agency costs. Target firms reduce their assets by
4.0% over two years (t = -1.10). As Jensen (1986) argues, management teams which ignore
shareholder interests may seek to grow assets at the expense of profitability. Activist hedge funds
seem to take particular interest in reducing firm size when they distrust management.
(F) Activist Ability Has �ot Declined
In unreported results, I find no evidence that activists’ abilities have declined over time. I
regress the changes in activist targets’ accounting measures on dummy variables for the
Schedule 13D filing year, controlling for Tobin’s q and for the accounting variables’ initial
levels. I conduct joint hypothesis tests on the null hypothesis that the yearly coefficients all equal
one another. While I find that activists’ ability to reduce assets has changed over time (F = 2.71),
it hasn’t decreased. Early in the decade, activists were very effective at reducing firm size,
towards the middle of the decade, activists struggled, and towards the end of the decade, activists
regained their effectiveness. No other accounting measures provide any significant results. Nor
do I find evidence that passive returns have changed over time. Thus, I find no support for the
hypothesis that activist skill has declined.
To summarize, I find manifold evidence that activism adds value. Target firms improve
their profitability, optimize their capital structure, reduce their potential agency costs, and sell
themselves more often and for richer valuations. In addition, activism adds value in the pattern
we would expect given hedge funds’ stated strategies. I find much more significant evidence that
- 55 -
activist hedge funds improve target firms than do previous studies.46 While this is in part due to
my unique sample, it is also because I dissect these improvements by strategy. In contrast, prior
studies looked at the results in aggregate—so different strategies canceled each other out.
Section VII: Conclusion
This paper argues that activist outperformance represents more than stock picking in
disguise. By comparing activist investments to passive investments while controlling for ex-ante
differences in target firms, I isolate hedge fund activism from hedge fund investing. Activism
earns substantial short-term and long-term returns by unlocking shareholder value; in particular,
activists revive struggling firms. They improve profitability, increase payout, spin off assets, and
sell target firms. Activists also tailor their approach to particular firms’ needs, focusing on
improving business strategies, advising on M&A transactions, selling target firms, optimizing
capital structures, or fixing corporate governance.
The new evidence presented in this paper expands the existing literature on activism. In
particular, while prior studies have shown that activist hedge funds earn excess returns, I present
strong evidence they earn these returns by doing more than picking stocks. I use several unique
tests to distinguish activism from passive investing—I compare activist investments to passive
investments, I control for ex-ante differences in target firms, I study activist instruments, and I
identify short-term events which contain only activist signals. Each test strongly supports the
hypothesis that activism adds value.
This paper has focused on the returns to corporate shareholders, sometimes at the expense
of measuring the returns to hedge fund investors. Most notably, I have made no adjustments for
46 See, e.g., Boyson and Mooradian (2008), Brav et al. (2008), and Klein and Zur (2008).
- 56 -
hedge fund fees. Thus, while I have shown that activism benefits shareholders, further research
might study whether it similarly benefits hedge funds’ limited partners.
In a broader context, this paper presents several lessons that might call corporate boards
to activism themselves. Given hedge funds’ ability to improve shareholders’ returns, perhaps
corporate boards should encourage hedge funds to invest. I also present evidence that expert
advice can turn around struggling firms. Perhaps undervalued companies should more frequently
hire outside consultants to revisit their policies on capital allocation, capital structure, M&A, and
corporate governance. Further research should investigate whether activist hedge funds follow
any predictable patterns that corporate executives might mimic.
Additionally, I show that agency costs detract from a firm’s value. In particular, why can
hedge funds unlock value beyond management’s ability even though hedge funds invest in many
companies, while managers devote their whole lives to just one firm? Most likely, this is due to
agency costs: whereas the hedge fund activists benefit from stock returns, the CEO benefits from
other factors—like compensation, pride, and job security. Perhaps corporate boards might
increase their firms’ values by compensating their CEOs less like CEOs and more like hedge
fund managers. Further research might investigate whether activists earn greater returns in firms
where managers’ interests are most poorly aligned with shareholders’ interests.
A1
Appendix A: Statistical Methods
(A) Event Studies
This essay uses the event study methodology. To determine how certain events affect
security returns, 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). Both models assume
that investments with greater risk offer greater reward. First, over a measurement window, I
calculate the firm’s risk. Then, over the event window, I assume the firm performed like other
firms with similar risk levels.
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).
itmtiiit RR εβα ++=
0)( =itE ε 2
)(iitVar εσε =
In this model, itR is the return to security i, and mtR is return to the valued-weighted market
portfolio in period t. iβ measures the security’s correlation with the overall market. iα represents
the idiosyncratic risk associated with a specific company. The error term itε has an expected
mean of zero, and the variance of itε is assumed to be homoskedastic with a normal distribution.
Fama and French (1992) recommend augmenting the market model to include more
possible risks—in particular, 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). Small firms tend
to be more risky and participate more frequently and for longer periods of time in earnings
downturns. Companies with high B/M ratios often suffer low earnings yields on their assets. To
A2
incorporate these factors into the market model, Fama and French suggest running time-series
regressions to determine the premiums and discounts associated with market capitalizations and
B/M ratios. They propose a three factor model:
ittitimtiiit HMLhSMBsRR υβα ++++=
0)( =itE υ 2
)(iitVar εσυ =
As before, itR is the return on security i and mtR is the return on the value-weighted index at time
t. iα is the intercept term. The two new terms represent the Fama French factors, where SMB
represents the difference in returns on portfolios of small and big market capitalization equities,
and HML is the difference between returns of portfolios separated by B/M ratios (Fama and
French 1992). Although these models are similar, most activism studies use the simpler market
model.1 Thus, I report results using the market model. To test for robustness, I check my results
against the Fama French three factor model.
To estimate the parameters, I follow two steps. First, I run times-series regressions over
an estimation window prior to the event. In this study, for daily returns I run regressions over the
200 day window ending 10 days before the event, and for monthly returns I run regressions over
the 36 month window ending one month before the event. In the diagram below, T-2 and T-1
represent the start and end of the estimation period, respectively. T0 represents the event day. T1
represents the end of the event window.
1 See Puustinen (2007), Boyson and Mooradian (2008), Brav et al. (2008), and Klein and Zur (2008).
T-2 T-1
Estimation
Window
Event
Window
T0 T1
A3
Second, I measure abnormal returns. I use the coefficient estimates from the time-series
regression to estimate counterfactual returns over the event window. Abnormal returns simply
equal actual returns minus expected returns
.
Under the null hypothesis, abnormal returns have conditional mean zero and conditional variance
,
where 200 represents estimation period’s length. To test returns over more than one day or one
month, I calculate abnormal returns (CARs) over the event window. For any two times t1 , t2 ,
CAR( t1 , t2) = .
This paper tests whether the mean CARs among a portfolio of activist investments differ
significantly from zero, where
and .
To test for statistical significance, the paper will calculate the variance of the mean CAR
,
where N is the number of events, and
.
This equation holds when the estimation period length T is large, such that
approaches zero (Dasgupta, Laplante, and Mamingi 1997).
A4
(B) Dummy Variable Regressions
This paper also makes extensive use of dummy variable regressions to analyze binary
variables like an activist’s strategy. First, I use dummy variable regressions to analyze whether
investments which I consider active enjoy greater returns or accounting outperformance than
other investments. To do so, I use a regression of the form
where Y i is the return to investment i, and the X ij ’s represent other covariates. I express all non-
dummy covariates as the deviation from the mean. For a passive target firm whose covariates
equal sample averages, returns equal . An activist target firm whose covariates equal sample
averages earns greater returns. If γ̂ is significantly greater than zero, I conclude that activism
generates significantly greater returns than passive investing.
Second, I use dummy variable regressions to measure the average partial returns on
various binary variables—most frequently, 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. I express all non-dummy covariates as the
deviation from the mean. Because I have suppressed the constant term, 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 example, if firm i’s covariates equal the sample averages,
A5
and if a hedge fund targets firm i with strategy 2 and strategy 3, then all the ’s equal zero, and
all the ’s equal zero, except and equal one, so we expect the return to be .
B1
Appendix B: Excerpted Hedge Fund Letters
To elaborate on the distinctions among activist strategies, this section excerpts various
hedge fund disclosures. The first section below presents a typical letter announcing that a hedge
fund will seek board seats. The next five sections describe business strategy activism, M&A
activism, sale activism, capital structure activism, and corporate governance activism,
respectively.
Investments Seeking Board Seats
In a letter to Motorola, 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.” Again this is only partially true. It is true that Sandy Warner, head of the Nominating Committee called me and offered seats to two of my Nominees if I would drop the proxy fight. However, you failed to mention in your conference call that I told Mr. Warner that I would gladly accept this offer if the Board would also accept Keith Meister. Mr. Warner replied summarily to this offer that Meister did not “qualify.” I asked Mr. Warner what does one have to do to qualify — lose $37 billion dollars? Mr. Warner then replied that the Board did not “know” Meister. 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). My offer to Motorola stills stands. 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. We disagree. In a political election when constituents believe their representatives’ performance was inadequate, they are certainly not denied information as to whether their representative acted in a grossly negligent fashion. 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. If as you have stated, we all want to benefit the stockholders of Motorola, then what possible reason is there for not putting Keith Meister on the Board. After all, how much can he eat at the Board meetings? On a positive side, having a highly intelligent, energetic individual like Keith, who has 145 million reasons to spend his time working toward the spin-off being accomplished, may well make this promise come true in a timely fashion. We ask
1 Motorola Inc., SEC Schedule 13D, filed 27 March 2008.
B2
the Board meet with Meister, put egos aside and let’s get on with the urgent business at hand.
Business Strategy Activism
In a letter to Pier 1 Imports, Elliot Associates wrote2:
Elliott is puzzled by the pace of cost-cutting and restructuring actions taken by the Company over the past two years. Whereas we believe a sense of urgency is appropriate, 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. It is imperative that the Company immediately focus on executing cost-cutting and restructuring initiatives in order to stem its current cash bleed. We suggest Pier 1 close down a significant number of stores, “right-size” the corporate and divisional SG&A, lease part or all of the corporate headquarters, and add new independent members to the Board of Directors. Elliott believes that these initiatives, if executed properly, will yield immense value to shareholders by creating a leaner, profitable company fit to navigate the intense competition of the future.
M&A Activism
In a letter to Brinks Co., MMI Investments wrote3:
We understand that BAX operations might be further improved, but conclude that this is insufficient grounds to retain them. The depressed market capitalization of BCO far outweighs any benefit of holding BAX. Furthermore, 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. 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. Any concerns about losing key people during a sale process can be addressed with meaningful divestiture incentive bonuses. This happens all the time and we would support such extra compensation for BAX executives in the event of a successful exit, whether that results in one or more transactions or a shut-down/liquidation. Most divestitures today, we believe, are achieved via an organized process, which includes a qualified investment banker with a board-approved mandate to approach qualified prospective buyers. Waiting for the buyers to come will likely result in no transaction.
2 Pier 1 Imports Inc., SEC Schedule 13D, filed 9 April 2007. 3 Brinks Co., SEC Schedule 13D, filed 20 April 2005.
B3
It is rare in life or business that the sun, moon and stars all align. 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, and a highly favorable environment in which to do so. We conclude that it is crucial for the future of BCO that a buyer is sought promptly. We would welcome the opportunity to discuss this or present further details to the Board.
Sale Activism:
In a letter to Gehl Co., 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, 2000 from Newcastle Partners, L.P. and CIC Equity Partners, Ltd. (collectively, the "Interested Parties") to acquire the Company, as described in such letter, for $18.00 per share in cash. Common sense aside, given that the offer represented a 67% premium to the Company's then current market price of $10.75 per share, we believe that the offer warranted, at minimum, a phone call to the Interested Parties to address any questions or concerns of the Company's Board of Directors. Instead, 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, or in any way seek any additional information about the offer. The Board's rejection of the offer, which rejection was announced in a press release only a few hours after it was delivered to you, clearly indicates to us that the offer was not given serious consideration by the Board. 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. Instead of giving any consideration to the offer, it appears that the Board spent their time preparing a press release attempting to belittle the ALL CASH offer. If there are concerns with any of the standard conditions contained in the offer, we invite the Board to contact either of the undersigned to directly address such concerns in a productive manner. 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.
Capital Structure Activism
In a letter to Ikon Office Solutions, Steel Partners wrote5:
We commend management for its efforts that have resulted in improved operating margins and a simplified capital structure. We also recognize and commend the
4 Gehl Co., SEC Schedule 13D, filed 20 March 2001. 5 Ikon Office Solutions, Inc., SEC Schedule 13D, filed 29 June 2007.
B4
Company for expanding the Company's existing share repurchase program. However, we believe that the Company's common stock continues to trade below its intrinsic value and your balance sheet remains highly inefficient. 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 $17.50 a share. This represents a 20% premium to the 20-day closing average of the shares. With the Company's excess cash and a conservative debt financing package, the Company should be in a position to self tender for at least $850 million of common stock, or approximately 39% of its outstanding shares at the aforementioned price per share. Based on our discussions with several leading investment banks, we believe this refinancing is highly achievable in a short period of time. The proposed recapitalization would provide immediate liquidity at a meaningful premium to the current share price for your shareholders who elect to tender. We believe it will be well received by the shareholders who desire liquidity. It would also be accretive immediately, and more so in the long term. As a long term shareholder, Steel would commit not to tender into this offer. Importantly, in addition to lowering the Company's cost of capital and enhancing its EPS growth, this recapitalization would result in a strong and flexible balance sheet that would support the Company's long term strategic vision.
Corporate Governance Activism
In a letter to Steel Gas Partners, Third Point Capital wrote6:
Sadly, your ineptitude is not limited to your failure to communicate with bond and unit holders. 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. (I was amused to learn, in the course of our investigation, that at Cornell University there is an "Irik Sevin Scholarship." One can only pity the poor student who suffers the indignity of attaching your name to his academic record.) Irik, at this point, the junior subordinated units that you hold are completely out of the money and hold little potential for receiving any future value. It seems that Star Gas can only serve as your personal "honey pot" from which to extract salary for yourself and family members, fees for your cronies and to insulate you from the numerous lawsuits that you personally face due to your prior alleged fabrications, misstatements and broken promises. I have known you personally for many years and thus what I am about to say may seem harsh, but is said with some authority. 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, SEC Schedule 13D, filed 14 February 2005.
B5
your fellow socialites. 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.
C1
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.
Especially because 13D Monitor selectively chose the hedge funds for my sample, it is worth
investigating whether certain outlier funds unduly influence the aggregate outcomes. This does
not appear to be the case. In this section, I analyze hedge fund heterogeneity in (1) short-term
returns, (2) long-term returns, and (3) target firm improvements. I find no evidence that outliers
have distorted my results.
Turning first to short-term returns, I look for outliers in any particular fund’s activist or
passive investments. For all hedge funds with at least three activist observations and three
passive observations, Figure C1 plots active returns against passive returns. For all funds above
In Short-Run, Most Funds' Active Investments Exceed Passive Investments
Results �ot Driven By Outliers
Figure C1. For each hedge fund with at least three activist investments and three passive investments, the chart
plots the fund's activist return aginst its passive return, from 30 days prior to each investment until 30 days after.
For funds above the 45° line, activist investments outperformed.
40%30%20%10%0%-10%-20%
40%
30%
20%
10%
0%
-10%
-20%
Passive Excess Return
Active Excess Return
Pershing
Greenlight
Nierenberg
DE Shaw
Cannell
Wynnefield
Newcastle
Stilwell
Southeastern
ElliotSandell
Harbinger
SAC
FEF
Discovery
Barington
Third Ave
ShamrockJANA
Third Point
Riley
Icahn
Ramius
C2
In Long-Run, Most Funds' Active Investments Exceed Passive Investments
Results �ot Driven By Outliers
Figure C2. For each hedge fund with at least three activist investments and three passive investments, the chart
plots the fund's activist return aginst its passive return, from one month before each investment until 24 months
after. For funds above the 45° line, activist investments outperformed.
120%100%80%60%40%20%0%-20%-40%
120%
100%
80%
60%
40%
20%
0%
-20%
-40%
Passive Excess Return
Active Excess Return
Nierenberg
SCSF
FEF
Cannell
Southeastern
Discovery
Wynnefield
Steel
SAC
Elliot
ValueAct
JANA
BaringtonRiley
Third Point
Icahn
Ramius
the 45° line, active investments outperform passive investments. The greater the distance from
the 45° line, the more the activist investments outperformed (or underperformed). For 73.9% of
the 23 funds, active investments outperformed passive investments. Further, I find few outliers.
Coupled with the strong median outperformance, this adds comfort that my prior results were not
skewed by outliers.
I use the same method to next analyze long-term returns. Figure C2 plots hedge funds’
active excess returns against passive excess return, from one month prior to the investment until
24 months after. If above the 45° line, a fund’s activist investments outperformed its passive
investments. Interestingly, the short-term outliers have disappeared; 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. In
C3
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
C4
Panel A: 82% of Hedge Funds Improve ROA
Panel B: 71% of Hedge Funds Increase Leverage
4%3%2%1%0%-1%-2%
4%
3%
2%
1%
0%
-1%
-2%
Passive Leverage Increase
Active Leverage Increase
Wynnefield
ValueAct
JANA
Steel
RamiusIcahn
Ramius
4%2%0%-2%-4%-6%-8%
4%
2%
0%
-2%
-4%
-6%
-8%
Passive ROA Increase
Active ROA Increase
Greenlight
Elliot
Southeastern
Farallon
Wynnefield
Steel
ValueActRamius
JANA Third Point
Icahn
C5
Panel C: 74% of Hedge Funds Reduce Assets
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.
20%10%0%-10%-20%
20%
10%
0%
-10%
-20%
Passive Asset Growth
Active Asset Growth
Highland
Greenlight
Breeden
Farallon
SCSF
Nierenberg
Cannell
SAC
Elliot
Southeastern
Discovery
StilwellFEFJANA Harbinger
ValueAct
Wynnefield
Steel
Barington
RileyIcahn
Third Point
Ramius
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.
Table C-1
Hedge Fund Sample
Activist Strategies Activist Instruments
Total Business Capital Corp Any Seek Win Proxy Any
Filings Strategy Sale Structure Governance Strategy Board Board Buyout Fight Lawsuit Instrument
Steel Partners 51 3.9% 19.6% 13.7% 11.8% 33.3% 37.3% 23.5% 9.8% 13.7% 3.9% 37.3%
ValueAct Capital 49 4.1% 10.2% 4.1% 4.1% 18.4% 16.3% 14.3% 10.2% 2.0% 2.0% 22.4%
Ramius Capital 39 17.9% 15.4% 20.5% 20.5% 59.0% 51.3% 38.5% 5.1% 17.9% 5.1% 56.4%
Carl Icahn 37 2.7% 37.8% 5.4% 2.7% 56.8% 29.7% 18.9% 24.3% 0.0% 5.4% 45.9%
Third Point 37 13.5% 24.3% 16.2% 27.0% 56.8% 29.7% 16.2% 2.7% 8.1% 8.1% 35.1%
Farallon 35 2.9% 5.7% 0.0% 0.0% 11.4% 2.9% 2.9% 5.7% 0.0% 0.0% 11.4%
Wynnefield Capital 29 10.3% 10.3% 0.0% 13.8% 37.9% 24.1% 20.7% 3.4% 3.4% 0.0% 24.1%
Harbinger Capital 24 12.5% 12.5% 12.5% 12.5% 29.2% 16.7% 8.3% 0.0% 4.2% 8.3% 54.2%
Riley Invetment Mgmt 24 16.7% 45.8% 20.8% 25.0% 70.8% 62.5% 50.0% 20.8% 25.0% 8.3% 66.7%
Financial Edge Fund 23 8.7% 21.7% 13.0% 17.4% 47.8% 34.8% 17.4% 0.0% 13.0% 13.0% 39.1%
Blum Capital 22 4.5% 9.1% 4.5% 4.5% 13.6% 4.5% 4.5% 0.0% 0.0% 0.0% 4.5%
Barington Capital 20 30.0% 30.0% 45.0% 35.0% 60.0% 50.0% 50.0% 5.0% 5.0% 10.0% 55.0%
Canell Capital 20 0.0% 15.0% 5.0% 0.0% 25.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
JANA Partners 18 11.1% 22.2% 22.2% 22.2% 55.6% 44.4% 33.3% 11.1% 16.7% 16.7% 50.0%
SCSF Equities 18 5.6% 22.2% 5.6% 0.0% 27.8% 22.2% 22.2% 16.7% 0.0% 0.0% 33.3%
Shamrock Activist Fund 17 29.4% 5.9% 52.9% 76.5% 88.2% 29.4% 17.6% 0.0% 17.6% 5.9% 47.1%
This table describes the hedge funds in my sample. For each fund, the table shows how many filings each fund made over the time period, what percentage of that fund's investments fall into each strategy category, and what percentage of that
fund's investments fall into each instrument category.
Shamrock Activist Fund 17 29.4% 5.9% 52.9% 76.5% 88.2% 29.4% 17.6% 0.0% 17.6% 5.9% 47.1%
Discovery Group 15 0.0% 26.7% 6.7% 20.0% 46.7% 20.0% 0.0% 0.0% 20.0% 0.0% 33.3%
Elliot Associates 15 13.3% 33.3% 0.0% 13.3% 66.7% 0.0% 0.0% 26.7% 6.7% 0.0% 33.3%
Joseph Stilwell 15 6.7% 33.3% 26.7% 6.7% 40.0% 46.7% 33.3% 0.0% 13.3% 20.0% 53.3%
Newcastle Partners 14 0.0% 21.4% 0.0% 7.1% 35.7% 71.4% 57.1% 21.4% 14.3% 7.1% 71.4%
Southeastern Asset Mgmt 14 21.4% 7.1% 7.1% 21.4% 42.9% 7.1% 7.1% 0.0% 14.3% 0.0% 21.4%
MMI Investments 13 0.0% 0.0% 7.7% 0.0% 38.5% 38.5% 23.1% 0.0% 15.4% 0.0% 38.5%
SAC Capital 13 0.0% 15.4% 0.0% 7.7% 53.8% 7.7% 0.0% 7.7% 15.4% 0.0% 23.1%
Highland Capital 12 0.0% 25.0% 8.3% 16.7% 33.3% 16.7% 0.0% 8.3% 8.3% 16.7% 33.3%
Nierenberg Investment 12 16.7% 0.0% 25.0% 33.3% 41.7% 8.3% 8.3% 0.0% 0.0% 0.0% 16.7%
Third Avenue Mgmt 12 8.3% 16.7% 16.7% 25.0% 66.7% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
DE Shaw 11 0.0% 9.1% 18.2% 9.1% 27.3% 18.2% 18.2% 0.0% 9.1% 9.1% 27.3%
Sandell Asset Mgmt 10 30.0% 40.0% 40.0% 20.0% 70.0% 70.0% 50.0% 0.0% 30.0% 10.0% 70.0%
Clinton Group 9 11.1% 33.3% 33.3% 22.2% 77.8% 33.3% 11.1% 0.0% 0.0% 0.0% 33.3%
Costa Brava 9 0.0% 11.1% 11.1% 0.0% 55.6% 55.6% 44.4% 11.1% 0.0% 44.4% 66.7%
David M Knott 9 0.0% 0.0% 0.0% 11.1% 33.3% 22.2% 11.1% 0.0% 0.0% 11.1% 22.2%
Greenlight Capital 9 22.2% 0.0% 0.0% 0.0% 33.3% 22.2% 22.2% 0.0% 11.1% 11.1% 33.3%
Cresecendo 8 25.0% 0.0% 12.5% 12.5% 37.5% 100.0% 100.0% 0.0% 25.0% 0.0% 100.0%
Breeden Capital 7 14.3% 0.0% 14.3% 14.3% 28.6% 42.9% 42.9% 0.0% 0.0% 0.0% 42.9%
Pershing Square 7 14.3% 14.3% 14.3% 0.0% 71.4% 14.3% 14.3% 14.3% 14.3% 0.0% 28.6%
Blue Harbor 6 16.7% 0.0% 16.7% 0.0% 16.7% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
New Mountain Vantage 5 20.0% 20.0% 20.0% 0.0% 20.0% 40.0% 40.0% 0.0% 0.0% 0.0% 40.0%
Relational Investors 5 80.0% 0.0% 40.0% 60.0% 80.0% 60.0% 40.0% 0.0% 0.0% 0.0% 60.0%
Total 718 10.4% 18.1% 13.6% 14.8% 44.0% 29.7% 21.7% 6.7% 8.8% 5.4% 37.0%
Appendix D:
Unreported Regressions and Additional Analysis
Table D-1
Characteristics of Target Companies
Panel A: Active vs. Passive Targets
(1) (2) (3)
Active Passive Difference
Mean Median StDev Mean Median StDev Mean t -stat
Market Cap 757.1 329.9 888.6 654.4 270.0 814.8 102.7 * 1.55
Tobin's q 1.6 1.3 0.8 1.7 1.4 0.9 -0.1 * 1.47
Abnormal q -16.6 -34.0 74.9 12.3 -4.2 81.4 -29.0 *** 3.68
P/E Ratio 15.8 16.2 18.1 13.3 15.3 18.5 2.5 * 1.36
Growth 3.0 2.3 16.4 4.1 3.9 16.4 -1.2 0.89
Margin 10.9 8.7 9.9 12.1 10.3 9.9 -1.1 * 1.34
Abnormal Margin 3.0 0.9 8.0 3.2 1.8 8.0 -0.2 0.23
Return on Assets 8.5 8.8 6.8 9.4 9.3 6.9 -1.0 ** 1.71
Return on Equity 20.0 16.8 18.5 22.3 19.1 19.7 -2.4 * 1.49
Abnormal ROE 0.7 -0.7 7.3 2.4 0.8 8.1 -1.7 ** 2.22
Leverage 22.2 20.5 19.2 22.8 19.8 20.6 -0.6 0.40
Abnormal Leverage 15.1 4.1 34.9 11.6 0.0 34.4 3.5 1.00
Payout Ratio 9.1 0.0 16.1 9.8 0.0 16.9 -0.7 0.42
Dividend Yield 0.5 0.0 1.0 0.6 0.0 1.1 -0.1 0.63
Cash/Assets 16.3 7.8 17.0 14.5 7.8 15.5 1.9 * 1.49
Capex/Assets -0.8 -0.5 0.8 -0.9 -0.7 0.8 0.1 0.84
R&D/Assets 2.2 1.9 1.6 1.9 1.5 1.7 0.3 * 1.30
The table reports the characteristics of hedge funds' targets. Column 1, Column 2, Column 4, and Column 5 report means, medians, and standard deviations. Column 3 and Column 6 report the
differences in means and associated t -statistics. Panel A separates activist investments from passive investments. Panel B separates activist investments which seek board seats from activist
investments which do not. All variables refer to the trailing twelve months prior to the Schedule 13D and are from Bloomberg or Compustat. Market Cap is market capitalization, Tobin's q is the
ratio of market value of equity plus long-term debt over book-value of equity plus long-term debt, Abnormal q is the target firm's Tobin's q as a percent of its industry's median Tobin's q , P/E Ratio
is the firm's trailing twelve months price-to-earnings ratio, Growth is the firm's one-year sales growth,Margin is the firm's EBITDA divided by sales, Abnormal Margin is the firm's margin less its
industry median margin, Return on Assets is the firm's EBITDA divided by its assets, Return on Equity is the firm's EBITDA divided by the book value of equity, Abnormal ROE is the firm's ROE
less its industry median ROE, Leverage is the firm's debt-to-assets ratio, Abnormal Leverage is the firm's leverage less its industry median, Payout Ratio is the firm's common dividends divided by
net income, Dividend Yield is the firm's total dividends divided by market capitalizatoin, Cash/Assets is the firm's cash and cash equivalents divded by total assets, Capex/Assets is the firm's capital
expenditures divided by total assets, and R&D/Assets is the firm's R&D expense divided by total assets. All data is winsorized at the 10% level. *, **, and *** refer to statistical significance at the
10%, 5%, and 1% levels.
Panel B: Active Investments Seeking Board Seats vs. Active Investments Which Do Not
(4) (5) (6)
Board Seats No Board Seats Difference
Mean Median StDev Mean Median StDev Mean t -stat
Market Cap 636.0 253.3 837.4 891.4 472.9 926.8 -255.4 *** 2.44
Tobin's q 1.5 1.2 0.7 1.7 1.5 0.8 -0.2 *** 2.47
Abnormal q -22.9 -42.6 72.6 -11.1 -29.8 76.7 -11.7 1.10
P/E Ratio 14.3 16.2 18.8 17.1 16.6 17.6 -2.9 1.07
Growth 0.5 1.0 16.6 5.5 5.8 15.9 -5.0 *** 2.54
Margin 8.4 6.9 8.8 13.5 11.5 10.4 -5.1 *** 4.13
Abnormal Margin 0.7 -0.8 7.2 5.0 3.7 8.1 -4.3 *** 3.54
Return on Assets 7.4 6.5 6.9 9.5 9.3 6.5 -2.1 *** 2.49
Return on Equity 17.0 13.9 17.9 23.0 20.3 18.6 -6.0 *** 2.59
Abnormal ROE 0.2 -1.6 7.2 1.2 0.0 7.4 -1.0 0.93
Leverage 19.7 16.4 19.4 24.7 25.1 18.8 -5.0 ** 2.21
Abnormal Leverage 12.8 4.0 33.5 17.1 4.7 36.3 -4.3 0.85
Payout Ratio 10.0 0.0 17.2 8.3 0.0 15.2 1.7 0.66
Dividend Yield 0.5 0.0 1.0 0.6 0.0 1.0 0.0 0.10
Cash/Assets 16.6 9.0 17.0 16.1 6.7 17.1 0.5 0.25
Capex/Assets -0.8 -0.5 0.8 -0.8 -0.5 0.8 0.0 0.35
R&D/Assets 2.0 1.8 1.6 2.3 2.0 1.6 -0.3 0.95
Table D-2
Short-Term Returns by Investment Type
Panel A: Active Investments Perform Better Than Passive Investments
Active Passive Difference
(-30,+30) (-2,+2) (+0,+2) (+0,+30) (-30,+30) (-2,+2) (+0,+2) (+0,+30) (-30,+30) (-2,+2) (+0,+2) (+0,+30)
5% -21.5% -7.5% -3.7% -12.2% -33.3% -7.0% -5.2% -24.2%
25% 1.6% 0.8% 0.1% -2.6% -10.7% -0.9% -0.7% -5.1%
Median 11.8% *** 4.4% *** 3.0% *** 5.6% *** 3.6% *** 1.9% *** 1.2% *** 2.0% *** 8.2% *** 2.5% *** 1.9% *** 3.6% ***
75% 25.4% 10.8% 9.7% 16.7% 19.2% 6.0% 4.4% 11.5%
95% 53.5% 25.9% 21.8% 38.8% 44.9% 16.7% 10.9% 30.3%
Average 12.9% *** 5.7% *** 4.6% *** 7.5% *** 4.6% *** 2.7% *** 1.8% *** 2.7% *** 8.3% *** 3.0% *** 2.8% *** 4.8% ***
% Positive 80.4% *** 79.9% *** 75.8% *** 69.1% *** 57.6% *** 67.4% *** 64.1% *** 57.4% ***
Patel z-statistic 10.77 16.50 17.16 8.89 4.04 9.27 7.82 3.00
Non par. z-statistic 8.91 8.76 7.61 5.75 3.37 6.95 5.76 3.26
N = 194 194 194 194 342 340 340 340
Panel B: Active Investments Seeking Board Seats Perform Better Than Active Investments That Do Not
Board Seats No Board Seats Difference
(-30,+30) (-2,+2) (+0,+2) (+0,+30) (-30,+30) (-2,+2) (+0,+2) (+0,+30) (-30,+30) (-2,+2) (+0,+2) (+0,+30)
5% -13.8% -7.8% -5.0% -8.7% -23.3% -6.1% -3.5% -13.6%
25% 6.1% 0.7% -0.8% -1.3% 0.9% 0.9% 0.1% -3.6%
Median 12.5% *** 7.8% *** 5.0% *** 9.3% *** 10.8% *** 4.1% *** 2.9% *** 5.0% *** 1.7% * 3.7% * 2.1% 4.3%
75% 29.8% 16.0% 14.1% 23.6% 24.4% 8.6% 7.6% 15.3%
95% 59.6% 29.2% 31.3% 40.6% 51.2% 21.6% 19.4% 35.1%
Average 17.7% *** 8.6% *** 7.6% *** 11.7% *** 12.6% *** 5.7% *** 4.6% *** 7.5% *** 5.1% * 2.9% * 3.1% ** 4.2% *
% Positive 88.9% *** 77.8% *** 71.1% *** 73.3% *** 77.9% *** 80.5% *** 77.2% *** 67.8% ***
Patel z-statistic 5.90 9.41 10.16 4.98 9.04 13.66 14.00 7.40
Non par. z-statistic 5.48 3.98 3.09 3.39 7.15 7.81 6.99 4.70
N = 45 45 45 45 149 149 149 149
The table presents average daily excess returns around the Schedule 13D filing for various event windows. I measure returns in excess of the market model using the value-weighted CRSP NYSE/Amex/Nasdaq index, estimating parameters over the 200 day
window prior to the Schedule 13D. Panel A compares active investments to passive investments. Panel B further splits active investments into those which seek board seats and those which do not. I report p -values for medians using the Wilcoxon Signed Rank
Test. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels.
Table D-3
Short-Term Returns By Type, Controlling For Undervaluatoin
Panel A: Industry-Adjusted Variables
(1) (2)
Dependent Variable: Activism Board Seats
Abnormal Returns Coeff. t -stat Coeff. t -stat
Constant 3.07 *** 5.09 5.65 *** 6.69
Abnormal q -1.14 ** -1.97 -2.07 ** -2.15
Abnormal Margin -0.09 * -1.56 -0.23 ** -2.23
Active 3.15 *** 3.37 - -
Board - - 2.45 1.24
No. and R2
290 0.07 122 0.11
Panel B: Non-Adjusted Variables
(3) (4)
Dependent Variable: Activism Board Seats
Abnormal Returns Coeff. t -stat Coeff. t -stat
Constant 2.45 *** 5.13 5.55 *** 6.45
Tobin's q -0.93 ** -1.96 -2.38 *** -2.46
Margin -0.02 -0.58 -0.03 -0.41
Active 3.75 *** 4.78 - -
Board - - 3.15 ** 1.71
No. and R2
441 0.06 162 0.07
The dependent variable is the abnormal return during the 5 days around the Schedule 13D filing date. Panel A controls
for abnormal q and abnormal margin and Panel B controls for Tobin's q and margin. Column 1 and Column 3 regress
all abnormal returns on Active , a dummy set to one for activist investments. Column 2 and Column 4 regress abnormal
returns from exclusively active investments on Board , a dummy set to one for board investments. All dummies refer
only to information contained in the initial schedule 13D filing. All nondummy variables are winsorized at the 10%
level and are expressed as the deviation from the sample average values. *, **, and *** indicate statistical significance
at the 10%, 5%, and 1% levels.
Table D-4
Abnormal Returns After 13D Filing Date
(1) (2) (3)
First Action First Board Seat First Board Win
(-30,+30) (-2,+2) (+0,+2) (+0,+30) (-30,+30) (-2,+2) (+0,+2) (+0,+30) (-30,+30) (-2,+2) (+0,+2) (+0,+30)
5% -30.2% -9.5% -3.1% -12.5% -23.4% -3.3% -2.8% -13.7% -29.1% -7.9% -5.4% -18.2%
25% -2.3% -2.2% -0.3% -1.9% -4.9% 0.2% 0.5% -3.5% -6.2% -2.2% -1.0% -5.6%
Median 7.5% *** 1.2% *** 2.8% *** 5.6% *** 6.3% *** 2.5% *** 2.6% *** 2.5% *** 5.2% *** 0.3% 0.7% *** 0.6% *
75% 20.0% 10.7% 9.3% 17.0% 18.4% 6.7% 4.6% 11.1% 20.7% 4.1% 3.3% 10.7%
95% 46.3% 21.7% 20.5% 26.9% 41.3% 13.6% 8.3% 34.1% 35.7% 10.2% 6.6% 25.8%
Average 8.3% *** 4.4% *** 5.3% *** 7.1% *** 6.9% *** 3.4% *** 2.5% *** 5.3% *** 6.2% *** 0.8% * 0.8% ** 2.5% ***
% Positive 73.6% *** 66.0% *** 71.7% *** 69.8% *** 60.4% ** 77.1% *** 77.1% *** 66.7% *** 66.4% *** 53.2% 61.5% *** 51.8%
Patel z-statistic 3.38 5.83 8.81 4.34 2.69 4.95 4.71 2.75 3.80 1.63 2.21 2.41
Non par. z-statistic 3.76 2.66 3.49 3.21 1.70 4.01 4.01 2.56 3.82 1.05 2.78 0.76
N = 53 53 53 53 48 48 48 48 110 109 109 110
The table presents average returns after the initial Schedule 13D filing, in excess of the value-weighted market, for various event windows. Column 1 analyzes schedule 13D/A filings in which, having already filed a passive 13D, a hedge fund first
announces activist intentions. Column 2 analyzes schedule 13D/A filings or news stories in which, having already announced activist intentions, a hedge fund first seeks board seats. Column 3 analyzes schedule 13D/A filings or news stories in which,
having already announced its intention to seek board seats, the hedge fund first wins board seats. I measure returns in excess of market model using the value-weighted CRSP NYSE/Amex/Nasdaq index, estimating parameters over the 200 day window
prior to the event. I report p -values for medians using the Wilcoxon Signed Rank Test. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels.
Table D-5
Long-Term Returns by Investment Type
Panel A: Active Investments Perform Better Than Paassive Investments
Active Passive Difference
(-1,+24) (+1,+6) (+1,+12) (+1,+24) (-1,+24) (+1,+6) (+1,+12) (+1,+24) (-1,+24) (+1,+6) (+1,+12) (+1,+24)
5% -132.5% -48.2% -77.7% -127.6% -153.3% -58.8% -89.4% -132.5%
25% -8.4% -12.6% -19.2% -18.0% -31.3% -20.1% -24.5% -30.3%
Median 20.1% *** 4.2% *** 7.4% *** 12.1% *** 10.4% * 0.1% -0.1% 5.3% 9.7% 4.1% 7.5% 6.8% *
75% 57.3% 19.9% 38.2% 47.5% 49.2% 20.3% 30.5% 45.9%
95% 172.3% 59.6% 111.3% 159.4% 125.2% 54.2% 75.2% 114.8%
Average 21.3% *** 4.9% *** 8.5% *** 13.5% *** 3.7% -0.6% -0.4% 1.5% 17.6% ** 5.4% ** 8.9% ** 11.9% *
% Positive 66.8% *** 56.3% *** 56.9% *** 61.1% *** 55.1% ** 50.0% 49.8% 53.7% ***
Patel z-statistic 8.43 2.88 3.87 5.55 1.08 -0.57 -0.29 0.16
Non par. z-statistic 5.70 2.39 2.58 3.87 2.08 0.54 0.48 16.29
N = 244 238 239 239 227 216 217 218
Panel B: Active Investments Seeking Board Seats Perform Better Than Active Investments That Do Not
Board Seats No Board Seats Difference
(-1,+24) (+1,+6) (+1,+12) (+1,+24) (-1,+24) (+1,+6) (+1,+12) (+1,+24) (-1,+24) (+1,+6) (+1,+12) (+1,+24)
5% -91.9% -37.0% -53.5% -75.9% -154.0% -55.8% -106.8% -144.7%
25% 0.2% -10.8% -11.6% -11.6% -8.1% -13.0% -24.1% -20.7%
Median 28.2% *** 6.1% *** 11.3% *** 18.3% *** 13.0% *** 2.0% 2.2% * 7.6% 15.2% ** 4.1% 9.2% * 10.8% **
75% 60.7% 22.4% 38.6% 53.1% 56.5% 19.1% 36.4% 40.6%
95% 150.7% 58.3% 95.7% 149.2% 177.8% 64.6% 116.2% 175.3%
Average 31.2% *** 8.3% *** 15.5% *** 21.6% *** 14.3% *** 2.8% ** 3.8% ** 7.8% *** 16.9% * 5.5% 11.7% * 13.8% *
% Positive 75.2% *** 58.5% ** 62.2% *** 66.4% *** 60.9% *** 55.4% * 54.5% 58.0% **
Patel z-statistic 7.48 2.67 4.19 5.15 4.97 1.89 1.84 3.19
Non par. z-statistic 6.03 2.32 3.14 4.06 2.45 1.30 1.11 1.87
N = 121 118 119 119 115 112 112 112
The table presents average monthly excess returns around the Schedule 13D filing for various event windows. I measure returns in excess of the market model using the value-weighted CRSP NYSE/Amex/Nasdaq index, estimating parameters over the 36 month
window prior to the Schedule 13D. Panel A compares active investments to passive investments. Panel B further splits active investments into those which seek board seats and those which do not. I report p -values for medians using the Wilcoxon Signed Rank
Test. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels.
Table D-6
Long Term Returns By Type, Controlling For Undervaluatoin
Panel A: Industry-Adjusted Variables
(1) (2)
Dependent Variable: Activism Board Seats
Abnormal Returns Coeff. t -stat Coeff. t -stat
Constant 6.67 0.90 7.81 1.00
Abnormal q -9.21 * -1.59 -14.40 ** -1.87
Abnormal Margin -1.01 ** -1.80 -0.01 -0.01
Active 4.32 0.45 - -
Board - - 2.45 1.24
No. and R2
251 0.03 150 0.03
Panel B: Non-Adjusted Variables
(3) (4)
Dependent Variable: Activism Board Seats
Abnormal Returns Coeff. t -stat Coeff. t -stat
Constant 4.60 0.75 15.33 ** 1.86
Tobin's q -7.07 * -1.36 -11.50 * -1.52
Margin -0.90 ** -2.11 -0.79 -1.25
Active 18.35 ** 2.17
Board - - 18.63 * 1.59
No. and R2
388 0.04 197 0.04
The dependent variable is the abnormal return during the 25 months around the Schedule 13D filing date. Panel A
controls for abnormal q and abnormal margin and Panel B controls for Tobin's q and margin. Column 1 and Column 3
regress all abnormal returns on Active , a dummy set to one for activist investments. Column 2 and Column 4 regress
abnormal returns from exclusively active investments on Board , a dummy set to one for board investments. All
nondummy variables are winsorized at the 10% level and are expressed as the deviation from the sample average values.
*, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels.
Table D-7
Profitability Improvements, Controlling For Stock Picking
Panel A: Return On Assets
(1) (2)
Dependent Variable: Activism Board Seats
Abnormal Returns Coeff. t -stat Coeff. t -stat
Constant -1.15 ** -1.70 1.20 0.89
Tobin's q 1.61 2.51 1.73 1.55
Initial Level -0.31 *** -4.16 -0.24 ** -1.76
Active 3.03 *** 3.05 - -
Board - - 0.82 0.46
No. and R2
108 0.26 40 0.10
Panel B: Margins
(3) (4)
Dependent Variable: Activism Board Seats
Abnormal Returns Coeff. t -stat Coeff. t -stat
Constant -2.53 *** -3.71 -0.06 -0.05
Tobin's q 0.16 0.24 0.23 0.24
Initial Level -0.12 ** -2.32 -0.20 *** -2.34
Active 2.87 *** 2.76 - -
Board - - 0.23 0.16
No. and R2
105 0.13 38 0.16
The dependent variable is the two-year increase in ROA (Panel A) and in EBITDA margins (Panel B). All regressions
control for Tobin's q and for the initial ROA (Panel A) or EBITDA margins (Panel B). Column 1 and Column 3 regress
improvements for all firms on Active , a dummy equal to one for activist investments. Column 2 and Column 4 regress
improvements for activist targets on Board , a dummy equal to one for board seat investments. All nondummy variables
are winsorized at the 10% level and are expressed as the deviation from the sample average values. *, **, and ***
indicate statistical significance at the 10%, 5%, and 1% levels.
Table D-8
Capital Structure Changes, Controlling For Stock Picking
Panel A: Payout
(1) (2)
Dependent Variable: Activism Board Seats
Abnormal Returns Coeff. t -stat Coeff. t -stat
Constant 2.25 ** 1.79 4.24 * 1.30
Tobin's q -1.36 -0.99 -0.56 -0.21
Initial Level 0.12 ** 2.02 0.16 * 1.39
Cash 10.45 * 1.50 10.83 0.70
Active 2.48 1.26 - -
Board - - 0.75 0.19
No. and R2
102 0.07 42 0.06
Panel B: Leverage
(3) (4)
Dependent Variable: Activism Board Seats
Abnormal Returns Coeff. t -stat Coeff. t -stat
Constant -0.37 -0.83 1.43 * 1.59
Tobin's q 0.05 0.11 -0.08 -0.11
Initial Level -0.05 *** -2.68 -0.05 ** -1.78
Cash -3.54 * -1.53 -4.07 -1.20
Active 0.65 0.98 - -
Board - - -2.08 ** -1.86
No. and R2
222 0.04 92 0.07
The dependent variable is the two-year increase in payout ratio (Panel A) and in the debt-to-assets ratio (Panel B). All
regressions control for Tobin's q, the initial payout ratio (Panel A) or leverage (Panel B), and the initial cash divded by
initial assets. Column 1 and Column 3 regress improvements for all firms on Active , a dummy equal to one for activist
investments. Column 2 and Column 4 regress improvements for activist targets on Board , a dummy equal to one for
board seat investments. All nondummy variables are winsorized at the 10% level and are expressed as the deviation
from the sample average values. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels.
Table D-9
Firm Size Reductions, Controlling For Undervaluatoin
Panel A: Assets
(1) (2)
Dependent Variable: Activism Board Seats
Abnormal Returns Coeff. t -stat Coeff. t -stat
Constant 2.77 ** 1.81 0.75 0.34
Tobin's q 8.75 *** 6.66 5.90 *** 3.15
Initial Level 0.00 -1.03 0.00 ** -1.89
Active -4.56 ** -2.12 - -
Board - - -4.42 * -1.49
No. and R2
522 0.09 256 0.06
Panel B: Sales
(3) (4)
Dependent Variable: Activism Board Seats
Abnormal Returns Coeff. t -stat Coeff. t -stat
Constant 4.95 *** 3.67 4.15 ** 2.03
Tobin's q 5.61 *** 4.85 6.29 *** 3.61
Initial Level 0.00 0.23 0.00 0.79
Active -1.19 -0.63 - -
Board - - -0.82 -0.30
No. and R2
494 0.05 244 0.06
The dependent variable is the two-year increase in assets (Panel A) and in sales (Panel B). All regressions control for
Tobin's q and for initial assets (Panel A) or initial sales (Panel B). Column 1 and Column 3 regress improvements for all
firms on Active , a dummy equal to one for activist investments. Column 2 and Column 4 regress improvements for
activist targets on Board , a dummy equal to one for board seat investments. All nondummy variables are winsorized at
the 10% level and are expressed as the deviation from the sample average values. *, **, and *** indicate statistical
significance at the 10%, 5%, and 1% levels.
Table D-10
Profitability Improvement By Strategy
Return on Assets EBITDA Margin
Coeff t -stat Coeff t -stat
Initial Value -0.4 *** -2.70 -0.2 ** -2.15
Business Strategy 6.9 *** 2.71 2.3 0.91
M&A Advice -0.7 -0.42 -0.6 -0.39
Sale -1.4 -0.86 -3.4 ** -2.11
Capital Structure 1.4 0.55 5.0 ** 1.81
Corporate Governance -0.3 -0.12 -0.7 -0.32
Other 0.5 0.22 2.0 0.87
No. and R2
29 0.50 26 0.46
Table D-11
Capital Structure Changes By Strategy
Payout Ratio Leverage
Coeff t -stat Coeff t -stat
Initial Value 0.0 0.11 -0.1 * -1.58
Cash 21.8 * 1.35 -6.1 * -1.47
Business Strategy -10.0 ** -1.78 -1.2 -0.73
M&A Advice 2.6 0.81 1.1 1.11
Sale -2.8 -0.81 1.0 0.90
Capital Structure 10.4 ** 2.21 -1.8 -1.17
Corporate Governance 3.4 0.62 1.3 1.11
Other 0.8 0.13 2.6 1.11
No. and R2
29 0.29 64 1.22
The table regresses the two-year change in (1) Return on Assets and (2) EBITDA margin on dummy variables for the various activist
strategies, considering only the sample of investments which announce activist strategies. The regressions control for the initial ROA or
the initial margin. The data is winsorized at the 10% level. Due to the full span of the dummy variables, the constants are supressed. *,
**, and *** indicate statistical significance at the 10%, 5%, and 1% levels.
The table regresses the two-year change in (1) payout ratio and (2) leverage on dummy variables for the various activist strategies,
considering only the sample of investments which announce activist strategies. The regressions control for the initial payout ratio or the
leverage as well as the initial cash divided by assets. The data is winsorized at the 10% level. Due to the full span of the dummy
variables, the constants are supressed. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels.
Table D-12
Firm Size Reductions By Strategy
Assets Sales
Coeff t -stat Coeff t -stat
Initial Value 0.0 ** -2.26 0.0 -0.06
Business Strategy 3.2 0.77 4.6 1.14
M&A Advice 0.7 0.23 3.6 1.26
Sale -4.5 * -1.43 -2.9 -0.93
Capital Structure 2.6 0.68 3.0 0.83
Corporate Governance -4.0 -1.10 2.3 0.65
Other 14.6 ** 1.65 12.0 * 1.32
No. and R2
185 0.05 178 0.06
Table D-13
Sale Efficiency By Activist Strategy
(1) (2)
Odds of Sale Avg Premium
Odds z -stat Coeff t -stat
Bus Strategy 0.23 ** -2.23 7.3 0.33
M&A 2.27 ** 1.74 13.7 ** 1.65
Sale 0.62 -0.98 24.6 ** 2.04
Cap Struc 0.95 -0.11 15.8 1.14
Corp Gov 0.47 -1.23 40.7 ** 2.26
Other 0.48 -0.62 74.8 ** 2.21
No. and R2
164 0.10 32 0.57
The table regresses the two-year percentage change in (1) assets and (2) sales on dummy variables for the various activist strategies,
considering only the sample of investments which announce activist strategies. The regressions control for the initial assets or sales. The
data is winsorized at the 10% level. Due to the full span of the dummy variables, the constants are supressed. *, **, and *** indicate
statistical significance at the 10%, 5%, and 1% levels.
Column 1 shows a logit model that predicts whether a target firm will sell itself, and Column 2 shows a regression that predicts the
average premium those firms receive. In Column 1, 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. I measure premium as the total return to shareholders from one month
before the activist first files a Schedule 13D until the sale closes. Both regressions include data only from investments which use activist
strategies. Column 1 presents R2, and Column 2 presents Pseudo-R
2. *, **, and *** indicate statistical significance at the 10%, 5%, and
1% levels.
References
Admati, Anat and Pfleiderer, Paul. 2009. The “Wall Street Walk’ and Shareholder Activism:
Exit as a Form of Voice. Stanford Graduate School of Business Research Paper No.
1918(R2).
Ang, James and Yingmei Cheng. 2006. Direct Evidence on Market-Driven Acquisitions Theory.
Journal of Financial Research 29: 199-216.
Aragon, George. 2007. Share Restrictions and Asset Pricing: Evidence From the Hedge Fund
Industry. Journal of Financial Economics 83: 33–58.
Bainbridge, Stephen. 2006. Directory Primacy and Shareholder Disempowerment. UCLA School
of Law, Law–Econ Research Paper No. 05–25.
Barber, Brad. 2007. Monitoring the Monitor: Evaluating CalPERS’ Shareholder Activism.
Journal of Investing 59: 66–80.
Berle, Adolf and Gardiner Means. 1932. The Modern Corporation and Private Property. New
York: Macmillan.
Becht, Marco, Julian Franks, Colin Mayer, and Stefano Rossi. 2008. Returns to Shareholder
Activism: Evidence From a Clinical Study of the Hermes UK Focus Fund. Working
Paper. Available at SSRN: http://ssrn.com/abstract=934712.
Bratton, William. 2006. Hedge Funds and Governance Targets. Working Paper. Available at
SSRN: http://ssrn.com/abstract=928689.
Bebchuck, Lucian. 2005. The Case for Increasing Shareholder Power. Harvard Law Review 118:
835–914.
Bebchuck, Lucian. 2007. The Myth of the Shareholder Franchise. Virginia Law Review 93: 675–
732.
Bebchuk, Lucian and Jesse Fried. 2003. Executive Compensation as an Agency Problem.
Journal of Economic Perspectives 17: 17–92.
Bethel, Jennifer, Julia Liebeskind, and Tim Opler. 1998. Block share Purchases and Corporate
Performance. The Journal of Finance 53: 605–34.
Bhide, Amar. 1993. The Hidden Costs of Stock Market Liquidity. Journal of Financial
Economics 34: 31–51.
Black, Bernard. 1990. Shareholder Passivity Reexamined. Michigan Law Review 89: 520–608.
Black, Bernard. 1992. Agents Watching Agents: The Promise of Institutional Investor Voice.
UCLA Law Review 39: 811–893.
Black, Bernard. 1998. Shareholder Activism and Corporate Governance in the United States. The
$ew Palgrave Dictionary of Economics and the Law 3: 459–465.
Boyson, Nicole and Robert Mooradian. 2008. Hedge Funds as Shareholder Activists from 1994–
2005. Working Paper. Available at SSRN: http://ssrn.com/abstract=992739.
Bradley, Michael, Alon Brav, Itay Goldstein, and Wei Jiang. 2007. Shareholder Activism and
Price Dynamics: Evidence From Closed–End Funds. Working Paper. Available:
http://finance.wharton.upenn.edu/~itayg/Files/Research%20Statement%20–
%20Website.pdf.
Brav, Alon, Wei Jiang, Frank Partnoy, and Randall Thomas. 2008. Hedge Fund Activism,
Corporate Governance, and Firm Performance. The Journal of Finance 63: 1729–75.
Carleton, Willard, James Nelson, and Michael Weisbach. 1998. The Influence of Institutions of
Corporate Governance through Private Negotiations: Evidence from TIAA–CREF. The
Journal of Finance 53: 1335–62.
Christoffersen, Susan, Chris Gêczy, David Musto, and Adam Reed. 2006. Vote Trading and
Information Aggregation. ECGI – Finance Working Paper No. 141/2007.
Clifford, Christopher. Value Creation or Destruction? Hedge Funds as Shareholder Activists.
Journal of Corporate Finance 14: 323–36.
Coffee, John. 193. Liquidity Versus Control: The Institutional Investor As Corporate Monitor.
Columbia Law Review 91: 1277–1368.
Dasgupta, Susmita, Benoit Laplante, and Nlandu Mamingi. 1997. Capital Market Responses to
Environment Performance in Developing Countries. The World Bank Development
Research Group Working Paper No. 1909.
Davis, Gerald and Han Kim. 2007. Business Ties and Proxy Voting by Mutual Funds. Journal of
Financial Economics 85: 552–70.
Del Guercio, Diane and Jennifer Hawkins. 1999. The Motivation and Impact of Pension Fund
Activism. Journal of Financial Economics 52: 293–340.
Del Guercio, Diane, Laura Wallis, and Tracie Woidtke. 2008. Do Board Members Pay Attention
When Institutional Investors “Just Vote No”? CEO and Director Turnover Associated
with Shareholder Activism. Working Paper. Available at SSRN:
http://ssrn.com/abstract=575242.
Deli, Daniel. 2002. Mutual Fund Advisory Contracts: An Empirical Investigation. Journal of
Finance 58: 109–33.
Fama, Eugene. 1970. Efficient Capital Markets: A Review of Theory and Empirical Work. The
Journal of Finance 25: 383–417.
Fama, Eugene and Kenneth French. 1992. The Cross-Section of Expected Stock Returns.
Journal of Financial Studies 47: 427-65.
Fama, Eugene and Michael Jensen. 1983. Separation of Ownership and Control. Journal of Law
and Economics 26: 301–25.
Fayele, Olubunmi. 2004. Cash and Corporate Control. Journal of Finance 59: 2041–60.
Gabaix, Xavier and Augustin Landier. 2008. Why Has CEO Pay Increased So Much? Quarterly
Journal of Economics 123: 49–100.
Gillan, Stuart and Laura Starks. 1998. A Survey of Shareholder Activism: Motivation and
Empirical Evidence. Working Paper. Available at SSRN:
http://ssrn.com/abstract=663523.
Gillan, Stuart and Laura Starks. 2000. Corporate Governance Proposals and Shareholder
Activism: The Role of Institutional Investors. Journal of Financial Economics 57: 275–
305.
Gillan, Stuart and Laura Starks. 2007. The Evolution of Shareholder Activism in the United
States. Journal of Applied Corporate Finance 19: 55–73.
Golec, Joseph. 1992. Empirical Tests of a Principal Agent Model of the Investor–Investment
Advisor Relationship. Journal of Financial and Quantitative Analysis 27: 81–95.
Gordon, Lilli and John Pound. 1993. Information, Ownership Structure, and Shareholder Voting:
Evidence From Shareholder–Sponsored Corporate Governance Proposals. Journal of
Finance 47: 697–718.
Graham, Benjamin. 1973 [2003]. The Intelligent Investor, Revised Edition. New York: Harper
Collins.
Grossman, Sanford and Oliver Hart. 1980. Takeover Bids, the Free Rider Problem, and the
Theory of the Corporation. Bell Journal of Economics 11: 42–64.
Greenwood, Robin and Michael Schor. 2008. Hedge Fund Investor Activism and Takeovers.
Harvard Business School Working Paper.
Hall, Brian and Kevin Murphy. The Trouble with Stock Options. Journal of Economic
Perspectives 17: 49–70.
Hu, Henry and Bernard Black. 2007. Hedge Funds, Insiders, and the Decoupling of Economic
and Voting Ownership: Empty and Hidden (Morphable) Ownership. Journal of
Corporate Finance 13: 343–67.
Jensen, Michael and William Meckling. 1976. Theory of the Firm: Managerial Behavior, Agency
Costs and Ownership Structure. Journal of Financial Economics 3: 305–60.
Jensen, Michael. 1986. The Agency Cost of Free Cash Flow: Corporate Finance and Takeovers.
American Economic Review 76: 323–29.
Johnson, Marilyn and Margaret Shackell. 1997. Shareholders’ Proposals on Executive
Compensation. Working Paper. Available at SSRN: http://ssrn.com/sbstract=121.
Kahan, Marcel and Edward Rock. 2006. Hedge Funds in Corporate Governance and Corporate
Control. University of Pennsylvania, Institute for Law & Economics Research Paper No.
06–16.
Karpoff, Jonathan. 2001. The Impact of Shareholder Activism on Target Companies: A Survey
of Empirical Findings. Working Paper. Available at SSRN:
http://ssrn.com/abstract=885365.
Karpoff, Jonathan, Paul Malatesta, and Ralph Walkling. 1996. Corporate Governance and
Shareholder Initiatives: Empirical Evidence. Journal of Financial Economics 42: 365–95.
Koller, Tim, Marc Goedhart, and David Wessels. 2005. Valuation: Measuring and Managing the
Value of Companies. Ne York: Wiley, John & Sons.
Klein, April and Emauel Zur. 2009. Entrepreneurial Shareholder Activism: Hedge Funds and
Other Private Investors. The Journal of Finance 64: 187–229.
Lipton, Martin and Steven Rosenblum. 1991. A New System of Corporate Governance: The
Quinquennial Election of Directors. University of Chicago Law Review 58: 187–253.
Lynn, Matthew. 2007. Wall street bosses spending too much time on the golf course. Wall Street
Journal. 17 November.
MacKinlay, Craig. 1997. Event Studies in Economics and Finance. The Journal of Economic
Literature 35: 13–39.
McConnell, John and Henri Servaes. 1990. Additional Evidence on Equity Ownership and
Corporate Value. Journal of Financial Economics 27: 595–612.
Parrino, Robert, Richard Sias, and Laura Starks. 2003. Voting With Their Feet: Institutional
Ownership Changes Around Forced CEO Turnover. Journal of Financial Economics 68:
3–46.
Pound, John. 1988. Proxy Contests and the Efficiency of Shareholder Oversight. Journal of
Financial Economics 20: 237-65.
Pound, John. 1992. Raiders, Targets, and Politics: The History and Future of American
Corporate Control. Journal of Applied Corporate Finance 5: 6–18.
Puustinen, Pekka. Determinants of Hedge Fund Activism and Shareholder Gains. Unpublished
MA, Helsinki School of Economics, Helsinki.
Rock, Edward. 1992. The Logic and (Uncertain) Significance of Institutional Shareholder
Activism. Georgetown Law Review 79: 445–406.
Romano, Roberta. 1993. Public Pension Fund Activism in Corporate Governance Reconsidered.
Columbia Law Review 79: 445–506.
Romano, Roberta. 2001. Less Is More: Making Shareholder Activism a Valued Mechanism of
Corporate Governance. Yale Journal on Regulation 18: 174–251.
Smith, Adam. 1776 [1977]. An Inquiry Into the $ature and Causes of the Wealth of $ations.
Chicago: University of Chicago Press.
Smith, Michael. 1996. Shareholder Activism by Institutional Investors: Evidence from
CalPERS. Journal of Finance 51: 227–52.
Sweren, Joshua. 2008. The Wandering Eye: Do Investors Steal Ideas from 13Fs? Unpublished
BA, Yale University, New Haven.
Strickland, Deon, Kenneth Wiles, and March Zenner. 1996. A Requiem for the USA: Is Small
Shareholder Monitoring Effective? Journal of Financial Economics 40: 319–38.
Wahal, Sunil. 1996. Pension Fund Activism and Firm Performance. Journal of Financial and
Quantitative Analysis 31: 1–23.
Yermack, David. 1995. Do Corporations Award CEO Stock Options Effectively? Journal of
Financial Economics 39: 237–69.