How Hedge Fund Activism changes Firms
Niko Markus Schindler
Abstract
This paper uses a sample of 946 hedge fund activist events between 2001 and 2006. Target firms prove to have abnormal buy-and-hold returns of 4.21 percent and 3.71 percent for value-weighted BHAR on the date of the 13D filing. Moreover, target firms are generally smaller and have stronger operational performance measures compared to their group of matched non-target companies. The operational performance measure of ROA increases after hedge funds announce their activist agendas, and target firms have higher CEO turnover. Furthermore, target firms prove to have higher production efficiency and a more efficient use of its assets.
1. Introduction
In corporate governance, the separation of ownership and operational control in public firms bears the risk of the agency problem. Thus, the cost of agency problem might cause severe damage to shareholder value. Shareholder monitoring might decrease the cost of the agency problem, yet it gives rise to another problem – the free rider problem, because all shareholders benefit from monitoring efforts. Bratton (2007) argues that hedge funds might be better capable of monitoring the executive body of public firms.
However, hedge funds with their ‘greed’ to “make returns that are not geared to the performance of the stock market” (Anderson, 2008, p. 140) generally suffer from bad reputation. In public, hedge funds are considered short-term focused with no real gain for shareholders. Even though previous literature has shown that firms targeted by hedge fund activists experience abnormal stock returns, a more in-depth picture of what parts of companies hedge funds target is still incomplete. Thus, fundamental knowledge of hedge fund activism is still missing, and Brav et al. phrases that lack of knowledge with unanswered question like: “Which firms do activists target and how do those targets respond? How does the market react to the announcement of activism? Do activists succeed in implementing their objectives? Are activists short term in focus? How does activism impact firm performance?” (2008, p. 1730.).
Therefore, this paper tries to address the question of how hedge fund activism changes firms. In section 1, previous literature on hedge fund activism is evaluated. Section 2 deals with the sample of hedge fund activist events used, and the methodology of how the target firms were matched with a control group. Section 3 highlights the characteristics of target firms, and section 4 shows how the markets react after hedge funds become activists. Section 5 analyses target firms performance and is subdivided into the impact of hedge fund activism on executives, operational performance, payout and capital structure, target firm’s earnings structure, investment structure and the working capital/current ratio of target firms. Finally, some conclusions are drawn on the findings of this paper.
! 2!
2. Literature Review
Hedge fund activism is generally classified as a strategy in which a hedge fund purchases a share of 5% or more in a publicly traded company with the intention to influence the target firm’s overall strategy (Klein et al, 2006, p. 1). Whereas hedge fund activism is a relatively new phenomenon, institutional investors such as pension funds and mutual funds have followed activist agendas for almost the past 30 years. In order to fully understand the impact of hedge fund activism, a recap of the different forms of shareholder activism and its results are important.
In corporate governance, the separation of ownership and control in publicly traded firms bears the risk of the agency problem between the firms’ managers and shareholders (Jensen and Meckling, 1976). According to Grossman and Hart (1980), monitoring benefits of all shareholders and causes the free-rider problem. Shleifer and Vishny (1986) argue that large minority shareholder might be the solution to the free-rider problem. These activist blockholders of mutual funds and pension funds are “the closest ancestors to hedge fund activists” (Brav et al., 2008, p. 1733). Pension funds and other activist investors have pursued their activist strategy on behalf of the SEC rule 14a-8 which came into effect in 1943 (Gillan and Starks, 2007, p. 34). Rule 14a-8 permits the issuance of shareholder proposals.
Generally, empirical research has shown that institutional investors like pension funds and mutual funds tend to target companies that perform relatively poorly in terms of market-to-book ratios and operating performance measures (Karpoff et al., 1996; Bethel et al., 1998; Smith, 1996). Therefore, institutional investors target firms that offer great upside potential for their activist actions.
However, research could not prove institutional investors to be successful activists. Karpoff et al. (1996) find little evidence that shareholder activism by institutional investors leads to an improvement of either performance or stock value. Similarly, Wahal (1996) and Romano (2001) argued that they found no evidence that activist events would have a significant impact on stock return or accounting measures of performance. Nevertheless, Smith (1996) concluded that shareholder activism was able to successfully change the governance structure of target companies and found statistically significant evidence that shareholder wealth increases when governance structure was changed successfully. Karpoff (1998)
evaluated all these empirical findings and concluded that, when averaged over all evaluated time periods, shareholder do not prove to have statistically significant effects on share values in the short-run.
The relative unsuccessfulness of institutional investors’ shareholder activism might have structural reasons. According to Romano (1993), “public pension funds face distinctive investment conflicts that limit the benefits of their activism” (p. 796). Thus, fund managers might be limited in their decision making by political pressure. Moreover, mutual funds gain from tax benefits only if they are diversified (Klein et al., 2006, p. 3). Therefore, mutual funds are neither allowed to own more than 10% of the outstanding securities of any company, nor to invest more than 5% of the fund’s total assets in one single security. Furthermore, investment funds that are broadly sold to the public are highly regulated and thus face restrictions on shorting, borrowing and investing in illiquid securities (Brav et al. 2008, p. 1734).
On the contrary, hedge funds are not subject to these regulations. Partnoy and Thomas (2006) note that “there is no generally agreed-upon definition of a hedge fund” (p. 22). Accordingly, the term “hedge fund” is not part of the federal securities law. Partnoy and Thomas (2006) define hedge funds upon “four characteristics: (1) they are pooled, privately organized investment vehicles; (2) they are administered by professional investment managers; (3) they are not widely available to the public; and (4) they operate outside of securities regulation and registration requirements” (p. 23). Furthermore, hedge fund managers are not restricted by pay-for-performance measures as mutual fund managers (Klein et al., 2006, p. 3). Hedge fund managers are generally compensated by a percentage of invested funds as well as a percentage of the hedge funds’ profits. Accordingly, hedge fund manages benefit financially from successful activism.
Therefore, because of these incentive structures, hedge fund activism compared to institutional investors’ shareholder activism tend to be directed at significant changes in individual companies, it has a strategic purpose as well as ex ante (Kahan and Rock, 2007, p. 1022). Moreover, parties that invest in hedge funds typically can not access or withdraw their investment for a fixed period of time, generally ranging from six months to several years (Partnoy and Thomas, 2006, p. 26). On the contrary, mutual funds need to deal with more investor redemptions. Thus, hedge fund managers tend to be more independent of their investors
! 4!
compared to mutual fund managers. In addition, Hu and Black (2006) find that hedge funds are able to buy shareholder voting rights undisclosed via the stock lending market, which enables hedge funds to hold large amounts of a company’s shares outstanding (compared to mutual funds) through both direct and indirect purchases.
Empirical research concerning hedge fund activism generally agrees on two conclusions. Firstly, hedge fund activists target firms with strong fundamentals and thus relatively high productivity, high ROA and liquidity (Brav et al., 2013). In addition, Boyson and Mooradian (2007), Brav et al. (2008) and (Greenwod and Schor (2009), find that the target firms are generally smaller compared to non-target companies. Secondly, empirical research has shown that shareholders of target firms benefit from abnormal returns around the date of the activist event (Schedule 13D filing date). Becht et al.’s (2009) field study of the Hermes U.K. Focus Fund results in evidence of excess returns to target firms. Brav et al. (2008) found abnormal returns of about 3.2 percent one day before the filing date (p. 1755). Klein et al. (2006) concluded “target firms earn, on average, 10.3 percent abnormal stock returns during the period surrounding the initial 13D filing. Furthermore, according to Clifford (2008), firms targeted by activists earn a 3.39% excess return surrounding the initial filing date.
Regarding improved operating efficiency, past empirical research differs in its conclusions. Klein et al. (2006) state they cannot find evidence that hedge fund activism improves firm performance but rather they show that several firm performance measures, such as ROA and ROE, decline one year after hedge funds started to become activist. On the contrary, Brav et al. (2008) found that both ROA and operating profit margins improved to a statistically significant extent two years after the 13D filing. Correspondingly, Clifford (2008) showed that ROA of targeted firms improved by 1.22 percent one year after hedge funds became activist. Additionally, Boyson et al. (2007) argued that hedge fund activism results in improvement of target firms’ long-term performance and improvement of cash positions.
Finally, Brav et al. (2014) concluded that target firms’ R&D expenses decline after hedge fund activist intervene. However, the target firms’ innovation output measured with the amount of filed patents do not shrink and partly even increase.
Thus, despite decreasing R&D spending, Brav et al. (2014) showed that “target firms’ efficiency improved after the intervention by hedge funds” (p. 27).
3. Data and Methodology A. Sample of Hedge Fund activist events
A centralized database for activist events of hedge funds do not exist. Thus, a sample for hedge fund activism is composed based on SEC 13D filings. Once a person or a group acquires more than 5% of shares in a company, it must be filed with the SEC under rule 13D within 10 days. This mandatory federal securities law is part of the Section 13(d) of the 1934 Exchange Act (Brav et al., 2008, p. 1736). Therefore, a list with ideally all hedge funds is matched with the SEC filings in the preset time frame. In the 13D filing itself, item 4 requires the filer to announce its intention behind the acquisition of the shares, so that passive hedge funds can be separated from activist hedge funds.
Amongst others, Professor Alon Brav from Duke University did major contributions to recent empirical research on hedge fund activism. After personal correspondence with Professor Brav, he was willing to share his data on hedge fund activist events with me. Professor Brav’s initial data set contains all hedge fund activist events from 1994 to 2011. For this paper, the time frame is set from 2001 to 2006. In the following sections on abnormal returns and target firms’ performance after a hedge fund became active, the paper deals with a sample of 554 and 809/810 hedge fund activist events respectively, as pointed out in Table 1.
Abnormal Returns Sample
- 19 HF events (companies do not have a Permno code) - HF events - several Firms are targeted multiple times per year - Firms delisted from CRSP
554 HF events
Target firms' performance Sample
- 22 HF events (companies do not have a Gvkey)
- 140 HF events - several firms are targeted multiple times per year - firms of which BE/ME data is unavailable
809*/810** HF events
* Peer Match ** Performance Match
HF events between 2001 and 2006
946 Hedge fund activist events in total
! 6!
B. Matching group
In order to examine the effect of hedge fund activism on target firms, several variables are tested against a group of non-target firms, which are structurally and regarding their performance similar to the targeted companies.
The first matching method is a mixture of the methods applied by Brav et al. (2008) and Klein and Zur (2009). Each targeted firm is compared to the average of the group of all matched companies along the same year, same two-digit SIC industry, and same Fama-French 5 x 5 size and book-to-market portfolio (Brav et al., 2008, p. 1767). The year before the hedge fund activist event happened serves as reference year, as proposed by Klein and Zur (2009, p. 200). Unlike Brav et al.’s approach that applies a Year-by-Year Peer Match, the matched group is kept constant over the compared years. This approach seems conceptually more reasonable, because a constantly matched reference group ensures a better illustration of the developments in a target firm and a non-target but similar company.
The second matching method is related to the beginning-of-period performance matching method as proposed by Barber and Lyon (1996, p. 369). Thus, each target firm is matched with one non-targeted company from the same two-digit SIC industry whose operating performance measure falls between 90 percent and 110 percent of that of the targeted company two years before the hedge fund activist event happened. The variable EBITDA/Revenue was used as operating performance measure, and the according data was pulled from Compustat (variable codes are EBITDA -- Earnings Before Interest and REVT -- Revenue – Total). The matching process was done in Stata with the psmatch2 command. In case Stata does not match the year, two-digit SIC industry code and the operating performance measure in the range of 90 percent to 110 percent, the target company was matched with a company from the same year, same two-digit SIC industry code and with the closest operating performance measure possible (from the Compustat output).
C. Data sources
For the calculation of abnormal returns, data was pulled from the CRSP database and Yahoo! Finance. For the calculation of target firms’ performance, Compustat and ExecuComp was used as source.
4. Characteristics of Target Companies
Table 2 evaluates the characteristics of target companies and gives an indication of the type of companies hedge fund activists target.
The statistically significant difference between target companies’ and matched firms’ total assets as well as market value ($-502.287 Mio and $-550.569 Mio respectively, both means winsorized) indicates that activist hedge funds are more likely to target rather small companies. Similarly, Greenwood and Schor (2009) found that target firms are generally smaller than non-target companies. From a hedge fund perspective, this outcome seems to make sense strategically, since smaller companies require fewer funds for a significant stake in the company, and thus fewer funds are ‘locked up’ in one investment. Furthermore, for a small company, the threat of a potential (hostile) takeover by the activist hedge fund is more severe, so that small firms are more likely to fulfill the hedge fund activist agenda.
Moreover, Brav et al. (2008) concluded that hedge fund activists are value investors, because the book-to-market ratio (amongst other valuation variables) has a statistically significant difference between target and non-target firms of 0.081 (2008, p. 1750). This paper, however, finds a winsorized mean of -0.013, so that Brav et al.’s conclusion cannot be confirmed. However, the differences are most likely due to differences in matching methods, which makes it difficult to come to a final conclusion regarding target firms’ book-to-market value.
Concerning operational performance, target firms seem to be more profitable in terms of both ROA (return on assets) as well as revenue growth (Growth Revenues), with a positive difference (winsorized) of 0.074 and 0.118 respectively; EBITDA is lower for target firms with a winsorized mean of $-75.470 Mio. Contrarily, Brav et al. (2008) found a negative growth, which might be again due to different matching methods and potentially due to different variables used from Compustat. However, both Brav et al. (2008) and Boysen and Mooradian (2007) found that hedge fund activists target firms with relatively low growth perspectives, but that are significantly more profitable in operational performance measures.
Furthermore, Table 2 shows that leverage is on average (winsorized) 3.6 percent higher for target firms, yet not statistically significant, which is confirmed by Brav et al. who found a difference in leverage of 2.8 percent (2008, p. 1753). In addition, target firms pay on average (winsorized) $-14.935 Mio lower dividends.
Table 2
Characteristics of target companies
Table 2 reports the characteristics of target companies and compares the target firms to a group of matched firms (with the matching process of target firms' performance as described in Section 2, one year before the hedge fund activist event (t-1). Columns (1) to (6) report the mean, winsorized mean, median, winsorized median, standard deviation and standard deviation winsorized of the target companies. The variables are winsozried at the 1st and 99th percentile. Column (7) to (9) reports the mean, winsorized mean and t-statistic of the difference between target firms and the average of each according group of matched companies. Data for Total Assets (in millions of dollars) is pulled from Compustat (variable AT); Market Value corresponds to MKVALT on Compustat. BE/ME is the market-to-book ratio defined as (book value of equity/market value of equity), and (SEQ/MKVALT) on Compustat. EBITDA corresponds to EBITDA on Compustat. ROA is return on assets, defined as EBITDA/lagged assets, and EBITDA/AT on Compustat. Leverage is defined as the book leverage ratio (debt/(debt + book value of equity)), which is (DT/(DT + SEQ)) on Compustat. Dividends is DVT on Compustat. Growth Sales is defined as the growth rate of revenues over the previous year, which is REVT on Compustat.
Firm Characterstic Mean(1) Mean (winsorized)(2) Median(3) Median (winsorized)(4) SD (5) SD (winsorized)(6) Mean(7) Mean (winsorized)(8) t-stat of Diff(9)
Total Assets 1005.152 847.441 234.061 231.919 3206.228 1717.918 -393.054 -502.287 -3.529 Market Value 573.192 532.433 125.035 125.031 1321.069 1035.096 -562.140 -550.569 -5.979 BE/ME 2.324 0.853 0.650 0.650 28.037 0.824 -5.280 -0.013 -0.757 EBITDA 93.707 81.841 14.508 14.230 348.577 197.831 -66.196 -75.470 -4.525 ROA 0.018 0.047 0.080 0.080 0.571 0.243 0.188 0.074 2.024 Leverage 0.221 0.317 0.208 0.207 3.057 0.409 -0.069 0.036 -0.551 Dividends 6.545 4.918 0.000 0.000 33.556 16.206 -14.101 -14.935 -5.947 Growth Revenues 0.243 0.118 0.049 0.049 2.123 0.449 -1.229 0.118 -1.208
Summary Stastics Difference with Matched Firms
8"
"
5. Stock Returns
The intrinsic question of hedge fund activism is the extent to which the activist attempts to create shareholder value. Thus, the following section examines how the market sees the effects of hedge fund activism on shareholder value.
Table 3 shows the average abnormal buy-and-hold returns of the targeted companies, which is calculated from the abnormal return in excess of the returns on the value weighted DJIA index. Data on returns of the individual firms is extracted from CRSP, and data on returns of the DJIA index is pulled from Yahoo! Finance. The abnormal buy-and-hold returns are calculated for 20 days prior to the Schedule 13D filing date to 20 days afterwards, a method also used by Brav et al. (2008 p. 1755). Additionally, the abnormal buy-and-hold returns are adjusted for each firms’ value-weight. Thus, total assets for the most recent fiscal year end before the 13D filing of all firms in question are added up, and each company’s abnormal return is adjusted for its share in the sum of all firms’ assets. Total assets as measure of size for the value-weighting seems conceptually most reasonable, because total assets are measured in book values, which is resistant to potential market fluctuations around the public announcement of the hedge funds to become activists.
Similar to Brav et al.’s (2008, p. 1755) findings, there is an average abnormal buy-and-hold return (BHAR thereafter) of 3.37 percent one day prior to the 13D filing. On the 13D filing date, BHAR is about 4.21 percent and 3.71 percent for value-weighted BHAR. BHAR increases gradually, except a small decrease between t +08 and t + 9, and t + 11 and t + 12. On day t + 10, BHAR amounts already to 6.86 percent, and on day t + 20, BHAR is 8.17 percent. The results on BHAR match quite well Brav et al.’s (2008) findings. However, Brav et al’s (2008) BHAR runs up more quickly in the last quarter. The slight difference might be due to different indices – Brav et al. (2008) calculated the excess return on various value-weighted indices (NYSE/Amex/NASDAQ index from CRSP), whereas this paper used the DJIA index from Yahoo! Finance.
Interestingly, the value-weighted BHAR follows a quite different path than the regular BHAR. From date t + 3 onwards, the value-weighted BHAR does not have the same tendency as BHAR anymore. Whereas BHAR continues rising after the 13D filing, the value-weighted BHAR does not rise anymore. From t + 3 to t + 20, the average deviation of value-weighted BHAR from BHAR is 3.03 percent with
Table 3
Buy-and-hold abnormal return around the filing of Schedule 13D filings
-0.60000%! -0.40000%! -0.20000%! 0.00000%! 0.20000%! 0.40000%! 0.60000%! 0.80000%! 1.00000%! 0.00000%! 1.00000%! 2.00000%! 3.00000%! 4.00000%! 5.00000%! 6.00000%! 7.00000%! 8.00000%! 9.00000%! t -2 0 ! t -1 9 ! t -1 8 ! t -1 7 ! t -1 6 ! t -1 5 ! t -1 4 ! t -1 3 ! t -1 2 ! t -1 1 ! t -1 0 ! t -0 9 ! t -0 8 ! t -0 7 ! t -0 6 ! t -0 5 ! t -0 4 ! t -0 3 ! t -0 2 ! t -0 1 ! 13D Filing ! t +0 1 ! t +0 2 ! t +0 3 ! t +0 4 ! t +0 5 ! t +0 6 ! t +0 7 ! t +0 8 ! t +0 9 ! t +1 0 ! t +1 1 ! t +1 2 ! t +1 3 ! t +1 4 ! t +1 5 ! t +1 6 ! t +1 7 ! t +1 8 ! t +1 9 ! t +2 0 ! Abnormal Return! WW Abnormal Return ! Buy-and-Hold Returns! WW Buy-and-Hold Returns! 10 # # #
maximum at 4.11 percent on date t + 18. Generally, value-weighted returns might be a less biased indicator, because small firms, for instance, seem to be more sensitive to macroeconomic conditions than large firms (Gertler and Gilchrist, 1991). Especially because hedge fund activist generally target relatively small firms, the sample of target firms is likely to be relatively volatile, in terms of macroeconomic conditions and anything related to it like access to capital markets and bank lending. Thus, value-weighted returns might decrease the volatility of the sample.
Furthermore, Brav et al. stresses that “market reactions are not an unbiased estimate of expected benefits from successful activism” (2008, p. 1757). If market prices were to adjust fully in anticipation of ex post effects of hedge fund activism, hedge funds would be incentivized to follow their activist agenda after the announcement. Thus, the abnormal market reaction is based on the expected benefit of the activist agenda, adjusted for the equilibrium probability that the hedge fund executes its announced activist plans successfully.
The abnormal BHAR around the 13D filing date leads to the assumption that the market anticipates that the announced hedge fund activist agenda actually improves the overall firm value. However, causality does not need to be given necessarily, and thus alternative causes of the abnormal BHAR need to be explored.
Brav et al. (2008, pp. 1760 – 1763) argues that the cause for abnormal BHAR might be buying pressure from the filing hedge fund (2008, pp. 1760 – 1763). Hence, Brav et al. finds that the 13D filing is accompanied by abnormally high trading volume (2008, p. 1760). The paper highlights that if the abnormal BHAR were truly related to the buying pressure only, negative abnormal BHAR returns should be observed after the 13D filing, which is not the case though. However, the value-weighted BHAR shows a slight tendency of reversal and stops rising after the 13D filing, which might be an indicator that the buying pressure is more relevant to the explanation of the abnormal BHAR than Brav et al. (2008) concluded. Thus, by incorporating the value-weighted BHAR into the analysis of abnormal BHAR, the market anticipation of value improvement of target firms after hedge fund activist agendas might be less important than earlier research has concluded.
Furthermore, abnormal BHAR might be due to hedge fund activists simply stock picking, rather than adding fundamental value to target firms (Brav et al., 2008, p. 1763). Brav et al. found this alternative cause reasonable, since he identified
! 12!
hedge fund activist as value investors (2008, p. 1764). However, hedge fund activists that are more likely to target firms with high book-to-market value could not be proved in this paper. Thus, the main argument for hedge fund activists being stock pickers rather than improving firms’ value fundamentally seems questionable. Similar to Brav et al.’s eventual conclusion, the abnormal BHAR is not likely to be caused by stock picking (2008, pp. 1763 – 1766).
Moreover, shareholders are not the only party who might benefit from hedge fund activism. Brav et al. identifies two key stakeholders: creditors and executives (2008, p. 1766). Thus, in the following section, the impact of hedge fund activism on stakeholders other than shareholders, amongst others, is analyzed.
6. Target Firms’ Performance Analysis A. Impact of Hedge Fund Activism on Executives
For the analysis of potential wealth redistribution amongst stakeholders, executives are examined. Brav et al. claimed that there is evidence of wealth redistribution from executives to shareholders (2008 pp. 1766-1767).
Panel A of Table 4 describes the average difference of CEO contracted pay between the listed variables and the matched group of firms along the two-digit SIC industry code, year, and the Fama-French 5 x 5 size and book-to-market dimensions. The t-statistic is calculated in Stata via a paired t test for CEO Contracted Pay and CEO-Pay-for-Performance; the t-statistic for the percentage of CEO turnover is calculated via a two-sample test of proportions. The variables are always winsorized at the 1st and 99th percentile, if not differently stated.
The first tested variable is CEO Contracted Pay (“TDC1” by ExecuComp), which is defined as total CEO compensation including option grants. In the year of the 13D filing, CEO Contracted Pay is on average $-250,222 lower in target companies compared to their matches. One year after the hedge fund activist event, the CEO compensation difference decreased to $-353,931. However, no average difference is statistically significant (neither at the 10 percent nor 5 percent level). Thus, from a statistical point of view, CEO Contracted Pay is not significantly different in target firms compared to the matched non-target companies. Contrarily, Brav et al. found that CEO compensation in the target companies is on average $914,000 higher (statistically significant at the 5 percent level) (2008, p. 1767)
Table 4
Target firm’s Performance Overview
Panels A to F outline various variables that map target firms’ performance at various levels compared to their matched non-target companies in years before and after hedge funds applied activist agendas. In the columns “Peer Match” (including the statistics concerning CEO Contracted Pay, CEO Pay-for-Performance and %CEO Turnover), target firms are matched to non-target firms along the year/industry/size/book-to-market dimensions; in the columns “(t – 2) Performance Match”, target firms are matched with non-targets based on operating performance (90 percent to 110 percent) two years before the hedge fund activist involvement. Then, difference is taken between target firms and the average of the group of matches, which is again averaged over all target companies and reported in the column “Diff w/ Match”. If not differently stated, the average is winsorized at the 1st and 99th percentile. Statistically significant average differences are marked at the 10 percent (“*”) and 5 percent (“**”) level.
Diff w/ Match Diff w/ Match winsorized t-statistic Diff w/ Match Diff w/ Match winsorized t-statistic Diff w/ Match z-statistic
t-2 -132.1575 -369.013 -0.262 0.298% 0.176% 0.111 2.65% 1.1406 t-1 -1099.905** -1012.583 -3.182 -0.108% -0.108% -0.042 4.032%* 1.7003 Event -117.5 -250.222 -0.303 -5.023%* -5.023% -1.879 2.982% 1.3074 t+1 -306.427 -353.931 -0.934 -2.495% -2.495%* -0.916 9.737%** 4.1651 (t+1) - (t-1) 713.1742 546.345 1.494 -1.875% -1.892% -0.478 5.705% -(t+1) - (t) -208.7117 -89.600 -0.512 4.841% 4.756% 1.413 6.755%
-CEO Contracted Pay ($1,000) CEO Pay-for-Performance % CEO Turnover
Panel A: Executive Compensation and Turnover (Year-by-Year Peer Match Only)
13
#
#
Diff w/ Match Diff w/ Match wins. t-statistic Diff w/ Match Diff w/ Match wins. t-statistic Diff w/ Match Diff w/ Match wins. t-statistic Diff w/ Match Diff w/ Match wins. t-statistic t-2 -34.393% -0.080% -0.982 -0.234% 0.684% -0.102 22.521%** 9.658% 2.127 1.729% 6.026% 0.049 t-1 -127.886% 0.125% -0.959 8.916% 2.173% 1.491 9.594% 5.187% 0.851 1.631% 2.586% 0.135 Event 2.033%* 0.094% 1.174 0.617% 0.706% 0.864 46.170% 3.089% 1.229 21.574% -7.377% 0.465 t+1 9.351% 0.383% 1.187 4.454%** 2.149% 2.158 4.922%* 5.226% 1.898 -9.294% 2.604% -0.761 t+2 0.074% -0.459% 0.101 1.317% 0.538% 1.389 4.434% 3.578% 1.493 13.370% 11.660% 1.989
(t+1) - (t-1) - 0.476% 1.045 (wins.) -1.586% 0.006% 0.004 (wins.) -9.555% 0.044% 0.0173 (wins.) -9.185% -1.456% -0.418 (wins.) (t+2) - (t-1) - -0.519% 0.986 (wins.) -6.501% -2.078% -1.458 (wins.) -11.910% -2.344% -0.885 (wins.) 13.501% 6.640% 1.348 (wins.)
Panel C: Payout and Capital Structure
Total Payout Yield Leverage
Peer Match (t-2) Performance Match Peer Match (t-2) Performance Match
14
#
#
Diff w/ Match Diff w/ Match wins. t-statistic Diff w/ Match Diff w/ Match wins. t-statistic Diff w/ Match Diff w/ Match wins. t-statistic Diff w/ Match Diff w/ Match wins. t-statistic t-2 -66.001** -66.968 -5.667 -115.001** -83.612 -3.679 -203.699** -304.295 -2.187 -966.177** -600.741 -2.640 t-1 -66.196** -75.470 -4.525 -153.001** -110.250 -3.629 -306.153** -342.915 -4.076 -1299.793** -760.846 -3.061 Event -75.839** -84.748 -5.112 -175.738** -131.374 -3.611 -360.121** -391.257 -4.795 -1529.855** -916.387 -2.909 t+1 -97.721** -106.823 -5.321 -166.285** -130.263 -3.332 -443.325** -467.674 -4.817 -1233.936** -974.888 -3.520 t+2 -120.453** -125.663 -5.689 -232.100** -178.735 -3.456 -488.136** -582.483 -3.680 -1750.747** -1378.680 -3.451 (t+1) - (t-1) -34.001 -32.880** -6.028 (wins.) -27.614 -28.151** -2.796 (wins.) -158.556 -137.328** -3.677 (wins.) -217.471 -289.901** -3.730 (wins.) (t+2) - (t-1) -54.075 -48.909** -6.584 (wins.) -68.682 -60.0212** -3.873 (wins.) -217.849 -251.535** -2.767 (wins.) -540.927 -564.534** -4.028 (wins.)
Panel D.1: Earnings Structure
EBITDA Net Sales
Peer Match (t-2) Performance Match Peer Match (t-2) Performance Match
Diff w/ Match Diff w/ Match wins. t-statistic Diff w/ Match Diff w/ Match wins. t-statistic Diff w/ Match Diff w/ Match wins. t-statistic Diff w/ Match Diff w/ Match wins. t-statistic
t-2 -276.323%** -37.283% -2.745 5.645% 0.708% 0.244 -5.245% -31.575% -0.155 4.448%* 1.673% 1.882
t-1 -305.086%** -69.709% -3.481 -235.263% -5.504% -1.344 -7.816% -30.549% -0.297 3.176% 2.392% 1.093
Event -233.335%** -49.519% -2.206 22.131%* 2.471% 1.654 -38.633%** -31.436% -3.053 0.917% 3.699% 0.098
t+1 -84.840%* -29.557% -1.679 61.714% 7.199% 1.573 6.061% -24.245% 0.118 55.693% 3.455% 0.721
t+2 -105.471% -47.176% -1.096 24.478% 6.104% 1.361 -30.704% -22.968% -1.518 -3.222% 1.713% -0.258
(t+1) - (t-1) 288.943% 48.170%** 5.622 (wins.) 49.938% 10.394%** 2.933 (wins.) 9.040% 8.390% 1.344 (wins.) 22.595% 1.154% 0.901 (wins.) (t+2) - (t-1) 331.410% 35.970%** 3.769 (wins.) 9.326% 10.055%** 2.615 (wins.) -35.286% 13.581%** 2.031 (wins.) -7.383% 0.032% 0.512 (wins.)
Panel D.2: Earnings Structure
COGS/SALES (wins. 90) SG&A/SALES
Peer Match (t-2) Performance Match Peer Match (t-2) Performance Match
15
#
#
Diff w/ Match Diff w/ Match wins. t-statistic Diff w/ Match Diff w/ Match wins. t-statistic Diff w/ Match Diff w/ Match wins. t-statistic Diff w/ Match Diff w/ Match wins. t-statistic t-2 -22.564** -23.270 -3.292 -42.505** -23.989 -2.367 -24.958** -23.919 -4.386 -14.344* -10.606 -1.818 t-1 -27.211** -27.096 -4.826 -52.152** -27.027 -2.557 -25.942** -24.712 -4.547 -15.076* -11.083 -1.651 Event -27.534** -30.368 -4.535 -70.010** -38.161 -2.552 -33.444** -30.790 -4.340 -16.099* -11.348 -1.531 t+1 -33.872** -37.521 -4.728 -41.901** -34.298 -2.134 -43.495** -41.957 -4.369 -17.020** -15.880 -2.029 t+2 -44.380** -48.880 -4.904 -86.675** -53.707 -2.435 -52.406** -51.935 -4.419 -14.541 -14.204 -1.482 (t+1) - (t-1) -6.889 -10.053** -3.612 (wins.) -10.864 -15.506** -1.997 (wins.) -15.483 -15.550** -3.881 (wins.) -6.178 -6.733* -1.729 (wins.) (t+2) - (t-1) -16.870 -20.833** -5.383 (wins.) -50.450 -31.444** -2.939 (wins.) -21.667 -22.827** -3.926 (wins.) -5.323 -7.481* -1.643 (wins.)
Panel E.1: Investment Structure
CapEx R&D
Peer Match (t-2) Performance Match Peer Match (t-2) Performance Match
Diff w/ Match Diff w/ Match wins. t-statistic Diff w/ Match Diff w/ Match wins. t-statistic Diff w/ Match Diff w/ Match wins. t-statistic Diff w/ Match Diff w/ Match wins. t-statistic
t-2 5.597% 8.932% 0.655 126.912%** 9.768% 2.062 46.971** 43.653 3.664 120.779** 46.053 2.151
t-1 -8.253% -7.280% -0.676 -11.453% -7.247% -0.494 29.621** 32.041 2.635 93.609* 60.335 1.908
Event 12.763% 3.034% 0.887 7.811% -8.081% 0.254 52.191** 54.998 3.860 115.958** 69.248 2.650
t+1 6.038% -1.345% 0.190 20.670% -0.152% 0.587 74.299** 96.056 3.302 84.517 56.084 1.681
t+2 -30.963%* -19.128% -2.079 -58.594%** -56.169% -2.853 64.122** 85.690 2.277 172.620** 135.866 2.650
(t+1) - (t-1) 24.937% 7.324% 0.945 (wins.) 56.338% 10.212% 1.457 (wins.) 43.819 63.213** 4.860 (wins.) -21.320 -12.750 -0.536 (wins.) (t+2) - (t-1) -13.806% -12.015% -1.292 (wins.) -20.386% -43.914%** -2.580 (wins.) 31.850 50.984** 3.846 (wins.) 49.838 58.121* 1.995 (wins.)
Panel E.2: Investment Structure
Return on Investment Total Cash From Investing
Peer Match (t-2) Performance Match Peer Match (t-2) Performance Match
16
#
#
Diff w/ Match z-statistic Diff w/ Match z-statistic Diff w/ Match z-statistic Diff w/ Match z-statistic Diff w/ Match z-statistic Diff w/ Match z-statistic t-2 1.451% 1.020 -5.653%** -2.647 -4.363%** -2.292 9.306%** 3.590 2.864% 1.625 -3.666% -1.451 t-1 -1.916% -1.323 -5.974%** -2.833 -4.470%** -2.361 6.492%** 2.439 6.362%** 2.203 -0.536% -0.202 Event -4.599%** -3.016 -6.026%** -2.835 -2.521% -1.296 7.756%** 2.811 7.120%** 3.893 -1.756% -0.633 t+1 -5.544%** -3.536 -5.550%** -2.551 -2.659% -1.281 2.314% 0.793 8.183%** 4.153 3.222% 1.113 t+2 -4.879%** -2.841 -5.400%** -2.331 -2.450% -1.124 4.464% 1.445 7.297%** 3.503 0.924% 0.331 (t+1) - (t-1) -3.628% - 0.425% - 1.811% . -4.178% . 1.821% - 3.758% . (t+2) - (t-1) -2.963% - 0.574% - 2.020% . -2.028% . 0.935% - 1.460% .
Panel F: Working Capital/Current ratio
(t-2) Performance Match Peer Match
% WC > 1 & < 2
% WC < 1 %WC > 2
Peer Match (t-2) Performance Match Year-by-Year Peer Match (t-2) Performance Match
17
#
#
! 18!
Similarly to Brav et al., the Pay-for-Performance variable is defined as the equity-based part of the CEO pay (including option exercise), which corresponds in ExecuComp to Restricted Stock Grant (RSTKGRNT) plus Options Granted (OPTION_AWARDS_BLK_VALUE) before 2006 and plus Grant Date Fair Value of Stock Awarded (STOCK_AWARDS_FV) and Grant Date Fair Value of Options Granted (OPTION_AWARDS_FV) after 2006, divided by Total CEO Pay (TDC1). In the event year, target firms’ executives earn on average -5.023 percent less equity based salary compared to their matches. Thereafter, the equity based compensation increases to an average difference of -1.875 percent, yet not at a statistically significant level. Brav et al. found a positive CEO Pay-for-Performance for the event year and the year after, both at a statistically significant level (2008, p. 1767). Despite different results, an approximate tendency of an increase in executives’ equity based salary might be derived from this paper as well. However, Brav et al. (2008) does not provide a sufficient description of the specific variables they used for their CEO Pay-for-Performance calculation, so that different calculations of the variables might be one of the reasons, why this papers’ results differ so much from Brav et al.’s (2008).
As Brav et al. suggests, a CEO turnover is classified as such if the CEO of a company is different from the CEO in the previous year (2008, p. 1770). The CEO turnover rate for the event year is 2.982 percent, yet not statistically significant (with two-sample test of proportions in Stata). One year after the hedge fund activists intervened, the CEO turn over rate increased up to 9.737 percent, which is significant at the 5 percent level. Brav et al. find a similar pattern (2008, p. 1770).
Overall, CEO Pay-for-Performance and CEO turnover seems to be enhanced by hedge fund activism. Thus, the change in CEO compensation and turnover seems to strengthen the hypothesis that hedge fund activists do not simply choose companies on a “stock picking basis”, but rather intervene seriously, because otherwise such severe changes are not very likely to happen (Brav et al., 2008, p. 1770). In addition, the change in CEO compensation and turnover supports the idea that hedge fund activism is responsible for the abnormal BHAR around the 13D filing date.
B. Impact of Hedge Fund Activism on Operational Performance
Following Brav et al. (2008), Panel B shows the operational performance based on ROA (return on assets, defined as EBITDA/Assets – “EBITDA/AT” in Compustat) and EBITDA/Revenues (“EBITDA/REVT” in Compustat). Similar to the following Panels, the target companies is matched with groups of non-target firms along the two-digit SIC industry code, year, and the Fama-French 5 x 5 size and book-to-market dimensions, as well as based on the industry and performance in year t - 2. The differences ((t + 1) - (t - 1) and ((t + 2) - (t - 1)) are calculated separately. Thus, occasionally, it seems as if the separate outcomes for (t + 2), (t + 1) and (t - 1) do not exactly match the separately calculated differences. This is the case, because when, for instance, the difference for a match is available in time (t + 1), but not in time (t - 1), then it does not enter the calculation for the overall average difference (t + 1) - (t - 1). However, as the difference for the specific match is present in (t + 1), it enters the calculation for the average difference in (t + 1), so that the difference (t + 1) - (t - 1) might seem to be distorted at first sight.
Consistent with previous research, target firms seem to have higher operating performance measures. ROA is constantly increasing after the 13D filing. In the even year, the target firms have a 5.546 percent higher ROA (significant at the 5 percent level) and 6.773 percent (significant at the 5 percent level) in year (t + 1). There is no statistically significant difference in year (t + 2). These findings are similar to Brav et al. (2008, p. 1771).
Furthermore, the EBITDA/Revenues variable seems to be higher for target firms as well. However, large outliers distort the outcome, so that the “Peer Match” is winsorized at the 10th and 90th percentile, and the (t - 2) Performance Match is winsorized at the 5th and 95th percentile. Similar to Brav et al., EBITDA/Revenues is higher for target firms, however, unlike Brav et al.’s findings, the positive excess operating performance declines after hedge fund interventions (yet still at a positive level).
C. Payout and Capital Structure
Panel C shows the Total Payout Yield and the Leverage. Brav et al. defines total payout yield as (dividend + share repurchase) / (lagged market value of equity) (2008, p. 1771). The ratio implies all the potential returns that an investor might
! 20!
obtain from owning equity in the company. This paper uses the Compustat variables (DVT + Purchase of Common and Preferred Stock) / (lagged MKVALT) in order to calculate Total Payout Yield. The data suggests a slight increase of Total Payout Yield in (t + 1) and a slight decrease thereafter in year (t + 2); yet not at a statistically significant level. The Performance match results in a statistically significant (at the 5 percent level) positive excess Payout Yield of 2.149 percent in year (t + 1) with a reversion in year (t + 2) (statistically insignificant). Brav et al. found a similar pattern at a statistically significant level. However, again Brav et al. might not have used the exact same variables from Compustat to calculate the Total Payout Yield, which may be one reason for the difference in results.
Brav et al defines Leverage as the ratio of debt to the sum of debt and book value of equity (2008, p. 1768). This paper uses the variables (DT / (DT + SEQ)) from Compustat for the calculation of Leverage. Panel C shows a weak relevering after the hedge fund activist event with 5.225 percent in year (t + 1) (significant at the 10 percent level) compared to 3.089 in the event year. However, compared to one year before the event, this relevering tendency is not significant. Brav et al. (2008) confirms the tendency of weak relevering at target firms after hedge fund activism.
D. Target firms’ Earnings Structure
Generally, Firms’ earnings are a function of revenues, sales production costs and investments. Thus, it might be interesting to test which variable hedge funds might target specifically. Therefore, in Panel D.1 and D.2 target firms’ EBITDA (EBITDA in Compustat), Net Sales (SALE in Compustat), COGS/Sales (Cost of Goods Sold, COGS/SALE in Compustat) and SG&A/Sales (Selling, General & Administrative Expenses, XSGA/SALE in Compustat) are tested against matched non-target companies.
As previous literature has shown, target firms tend to have higher operating performance measures. Thus, it is surprising that target firms have statistically significant lower EBITDA and Net Sales, throughout the whole time horizon. However, absolute variables do not incorporate the size of a company. Needles to say, larger firms are more likely to have higher EBITDA and thus can easily outperform small firms in such absolute measures. Especially because activist hedge funds tend to target smaller but profitable firms, one reason might be that the
relatively small EBITDA (or Net Sales) is compensated by the relatively small size. This idea is strengthened by the different results of Peer Match and Performance Match. The negative excess EBITDA and Net Sales are bigger for the Performance Match, which does not match along the size dimension. Overall, absolute variables can be less meaningful than ratios.
COGS/Sales is the percentage of sales revenue, which is used to pay for expenses that are directly dependent on sales. Thus, it is a measure of what percentage of sales revenues is used up by direct production costs, and, therefore, can be seen as a measurement of how productive a company is – the lower the ratio, the larger the share of sales revenue that does not pay off direct production costs. The same holds for SG&A/Sales, which is the share of sales that pays off costs that are not directly linked to the pure production of the good.
Moreover, target firms have proven to be stronger in operational performance measures compared to their matched sample. Thus, conceptually, it would make sense to expect target firms to be more efficient, and, therefore, to have lower COGS/Sales and SG&A/Sales ratios than their matched non-target firms. For the Peer Match, this assumption seems to be valid. Due to large outliers, average differences of COGS/Sales are winsorized at the 5th and 95th percentile. In the event year, target firms have a lower COGS/Sales ratio that is -49.519 percent (statistically significant) lower than the matched peer group. This ratio increases in year (t + 1) up to -29.557 percent with a reversion in year (t + 2) to -47.170 (neither are statistically significant though). A similar pattern is observed for the SG&A/Sales ratio. In the year of the 13D filing, target firms have a lower SG&A/Sales ratio by -38.633 percent (statistically significant), which increases slightly thereafter, yet not at a statistically significant level. Brav et al. (2013) show that hedge fund activism leads to productivity gains of the plants of the targeted companies. This conclusion cannot be completely confirmed. However, it seems that target firms are generally more productive than their matched non-target peer group.
E. Target Firms’ Investment Structure
Corporate innovation is a crucial component in analyzing hedge fund activism’s impact on target firms. Innovation is one of the most important factors for long-term growth of a company. Thus, CapEx (capital expenditures, CAPX in Compustat), R&D
! 22!
(research and development expenses, XRD in Compustat), ROI (return on investment, (net income before extraordinary items)/ (total invested capital), IBCOM/ICAPT in Compustat) and TCI (total cash from investing, investing activities - net cash flow, IVNCF in Compustat) are analyzed in Panel E.
CapEx seems to be significantly lower for target firms, and decreases further after hedge funds become activists. When comparing year (t + 1) to (t - 1), target firms show between -10,000,000 (Peer Match) and -15,500,000 (Performance Match) lower CapEx compared to the group of matched non-target companies. Year (t + 2) compared to year (t - 1) leads to an average difference between -20,800,000 (Peer Match) and -31,400,000 (Performance Match). However, lower CapEx does not necessarily serve as indicator for lower innovative power. CapEx is no flow of continuous investment, but rather punctual investments, which last until the investment is depreciated. Therefore, the target firms might have made the necessary investments in earlier years and thus are able to perform better in operational respects. Furthermore, Beneish et al. (2001) found that firms that experience an extreme up price movement tend to have lower capital expenditures than the control group. This might be an indicator for the abnormal BHAR target firms experience around the 13D filing date.
Similarly, R&D expenditure is significantly lower for target firms and decreases after hedge fund activists get involved as well. Year (t +1) compared to year (t - 1) reveals a decrease in R&D spending between 6,700,000 (Performance Match) and -15,500,00 (Peer Match); year (t + 2) compared to year (t - 1) shows a decrease in R&D expenses between -7,500,000 (Performance Match) and -22,800,000 (Peer Match). However, Brav et al. (2014) found that a decrease in R&D spending does not impact the quality of corporate innovation. Actually, the output of innovation (measured in patent counts) increases after firms are targeted by hedge fund activists. Furthermore, Brav et al. (2014) argued that hedge fund reallocate innovative assets, such as patents and human capital. Thus, despite a decline in R&D spending, hedge fund activism improves the firms’ base for a sustainable future growth.
Therefore, it is intuitive to assume the improved innovation quality has a positive impact on ROI, defined as (net income / total cash invested)1. However, the test results are not statistically significant, despite the negative return in (t + 2) of -19.128 percent (Peer Match) and -56.594 percent (Performance Match). The numbers show an improvement in year (t + 1) (not statistically significant though) and a sharp decline in the following year (significant at the 5 percent level).
Moreover, total cash from investing is analyzed. Compustat defines Investing Activities Net Cash Flow as an item that represents net cash received or paid for all transaction classified as investing activities on a Statement of Cash Flows. Total cash from investing is significantly positive for the entire period of time. However, positive total cash from investing is not per se a good or bad thing. Reason for a positive cash flow could be the sale of property or the sale of (financial) investments. Nevertheless, the positive cash flow account might be supporting Brav et al. (2014) findings of improved innovation quality, because it frees up funds for reallocation or restructuring of the more efficient investment structures after hedge fund activists get involved. On the contrary, positive total cash from investing together with the decrease in CapEx might generate some short-term cash, whose anticipation may explain partly the abnormal BHAR after hedge funds become activists.
F. Target Firm’s Working Capital/Current Ratio structure
The current ratio (current assets / current liabilities, ACT / LCT in Compustat) is an indicator for a firm’s short-term liquidity. The higher the current ratio, the more likely the firm is able to pay back its debt. Thus, working capital can be seen as one of the indicators for operational efficiency, because inventory, for instance, that is not sold yet, does not generate sales revenue, which could be used to invest or pay back debt. However, a relatively high current ratio might be a sign that a firm does not use its assets in the most efficient way possible. Thus, a current ratio between 1 and 2 is often quoted as target benchmark (Marshall et al., 2011, p. 83). Accordingly, the percentage of firms with a current ratio lower than 1, greater than 2, as well as lower than 2 but great than 1 (relatively efficient use of firm’s assets) is analyzed in Panel
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
1!Definition from “ISS Governance Services – Proxy Research; Company Financials – Compustat Data Definitions”, June, 2008.
! 24!
F. The average difference between target firms and its matches is calculated in Stata via a two-sample test of proportions.
Panel F shows that target firms tend to be more liquid and use their assets more efficiently compared to the matched non-target companies. In year (t + 1) target firms’ current ratios are on average -5.544 percent (z = -3.536) less below 1 compared to the peer group; in the year of the event -4.599 percent (z = -3.016). Contrarily, compared to year (t - 1) with -4.470 percent (z = 2.361) of target firms with a current ratio above 2 relatively to the matched group, the target firms increased in year (t + 1) to -2.659 percent, yet not statistically significant. Furthermore, target firms with relatively efficient use of its assets increased from 6.362 percent (z = 2.203) in year (t - 1) to 8.183 percent (z = 4.153).
To summarize, this paper finds in line with Brav et al. (2008) a slight improve in equity based CEO pay, an increase in CEO turnover, and a strong operational performance for target firms. Furthermore, target firms have negative CapEx, but positive total cash from investing. Finally, target firms improve their working capital after hedge funds become activists.
G. Differences with Brav et al.’s (2008) results
This paper uses the same data set than Brav et al. (2008), so that a more thoroughly overlapping of results might be expected. The differences most likely stem from differences in matching methods. Brav et al. (2008) applies a Year-by-Year Peer Match, whereas this paper keeps the matched group constant over the compared years. Additionally, Brav et al. (2008) does not mention how outliers are eliminated. However, as the results of this paper demonstrate, a winsorized result only at the 1st and 99th percentile can have a dramatic effect. Furthermore, it is not clear how Brav et al. (2008) select firms from all the firms that match the various dimensions. This paper averages all firms that possibly match the same dimensions of a target company, which Brav et al. (2008) rather did not do, as the punctually significantly different results show. Finally, Brav et al. (2008) might have used slightly different variables from Compustat, which results in partly different outcomes.
7. Conclusion
This paper uses the same sample of hedge fund activist events as Brav et al.’s (2008) research. Similar to previous research, this paper finds positive market reactions to hedge fund interventions in the form of abnormal buy-and-hold returns. However, the gradual rise of abnormal buy-and-hold returns after the 13D filing is not confirmed by the value-weighted BHAR. Thus, as value-weighted BHAR might not influenced by the relatively higher volatility of generally smaller target firms, hedge fund activists might not change equity prices as sustainably as Brav et al. (2008) concludes. Furthermore, target firms improve ROA after the hedge fund activist event and prove to have generally higher operational performance (measured with EBITDA/Revenues). There is weak evidence that target firms slightly relever after the event.
Moreover, in line with the stronger operational performance measures, COGS/Sales and SG&A/Sales ratios highlight that target firms are more generally more productive compared to their matched peer group. Target firms’ lower CapEx and R&D spending does not necessarily have an impact on target firms’ quality of innovation, as Brav et al. (2014) confirm. However, the significantly higher total cash from investing might be an indicator that the generation of short-term cash might be anticipated and leads partly to the abnormal buy-and-hold return. On the other hand, positive total cash from investing frees up funds for reallocating investment resources, which, in turn, improves the quality of innovation. Finally, the analysis of the working capital ratio shows that hedge funds improve the efficient usage of target firms’ assets.
In conclusion, hedge fund activism generates value for target firms on average, not because of simple “stock picking”, but rather because hedge fund activists transparently communicate their agendas and commit to it throughout their involvement in target firms. This paper shows that the change in target firms that hedge fund activism cause has a deeper impact than a superficial boost in equity prices.
! 26!
References
Anderson, G. (2008). Cityboy – Beer and Loathing in the Square Mile. London: Headline Publishing Group.
Becht, M., Franks, J., Mayer, C., Rossi, S. (2009). Returns to Shareholder Activism: Evidence from a Clinical Study of the Hermes UK Focus Fund. The Review of Financial Studies, 22(8), 3093-3129.
Beneish, M. D., Lee, C. M. C., Tarpley, R. L. (2001). Contextual Fundamental Analysis Through the Prediction of Extreme Returns. Review of Accounting Studies, 6, 165-189.
Bethel, J. E., Liebeskind, J. P., Opler, T. (1998). Block Share Purchases and Corporate Performance. The Journal of Finance, 53(2), 605-634.
Boyson, N. M., Mooradian, R. M. (2007). Hedge funds as shareholder activists from 1994-2005. Working Paper, Northeastern University, Boston.
Bratton, W. W. (2007). Hedge Funds and Governance Targets. The Georgetown Law Journal, 95, 1375-1433.
Brav, A., Jiang, W., Kim, H (2009). Hedge Fund Activism: A Review. Foundations and Trends in Finance, 4(3), 185-246.
Brav, A., Jiang, W., Kim, H. (2013). The Real Effect of Hedge Fund Activism: Productivity, Asset Allocation, and Industry Concentration. Working Paper.
Brav, A., Jiang, W., Tian, X (2014). Shareholder Power and Corporate Innovation: Evidence From Hedge Fund Activism. Working Paper.
Clifford, C. P. (2008). Value creation or destruction? Hedge funds as shareholder activists. Journal of Corporate Finance, 14, 323-336.
Del Guercio, D., Hawkins, J. (1999). The motivation and impact of pension fund activism. Journal of Financial Economics, 52, 293-340.
Del Guercio, D., Seery, L., Woidtke, T. (2008). Do boards pay attention when institutional investor activists “just vote no”?. Journal of Financial Economics, 90, 84-103.
Gertler, M, Gilchrist, S. (1991). Monetary Policy, Business Cycles, and the behavior of small manufacturing firms. NBER Working Paper.
Gillan, S. L., Starks, L. T. (2000). Corporate governance proposals and shareholder activism: the role of institutional investors. Journal of Financial Economics, 57, 275-305.
Gillan, S. L., Starks, L. T. (2007). The Evolution of Shareholder Activism in the United States. Working Paper.
Goetzmann, W. N. (2000). Hedge Funds: Theory and Performance. Working Paper.
Greenwood, R., Schor, M. (2009). Investor activism and takeovers. Journal of Financial Economics (92), 362-375.
Grossman, S. J., Hart, O. D. (1980). Takeover Bids, The Free-Rider Problem, and the Theory of Corporation. The Bell Journal of Economics, 11(1), 42-64.
Hu, H. T.C., Black, B. (2006). Hedge Funds, Insiders, and Empty Voting: Decoupling of Economic and Voting Ownership in Public Companies. Working Paper.
Kahan, M., Rock, E. B. (2007). Hedge funds in Corporate Governance and Corporate Control. University of Pennsylvania Law Review, 155(5), 1021-1093.
Kahn, C., Winton, A. (1998). Ownership Structure, Speculation, and Shareholder Intervention. The Journal of Finance, 53(1), 99-129.
Karpoff, J. M., Malatesta, P. H., Walkling, R. A. (1996). Corporate governance and shareholder initiatives: Empirical evidence. Journal of Financial Economics, 42, 365-395.
Karpoff, J. M. (1998). The Impact of Shareholder Activism on Target Companies: A Survey of Empirical Findings. Working paper.
Klein, A., Zur, E. (2006). Hedge Fund Activism. Working Paper.
Klein, A., Zur, E. (2009). Entrepreneurial Shareholder Activism: Hedge Funds and Other Private Investors. The Journal of Finance, 64(1), 187-229.
Marshall, D. H., McManus, W. W., Viele, D. F. (2011). Accounting – What the Numbers Mean. New York, NY: McGraw-Hill.
Partnoy, F. S., Thomas, R. S. (2006). Gap Filling, Hedge Funds, And Financial Innovation. Working Paper.
Romano, R. (1993). Public Pension Fund Activism in Corporate Governance Reconsidered. Columbia Law Review, 93(4), 795-853.
Romano, R. (2001). Less is More: Making Institutional Investor Activism A Valuable Mechanism of Corporate Governance. Working Paper.
Shleifer, A., Vishny, R. W. (1986). Large Shareholders and Corporate Control. Journal of Political Economy, 94(3), 461-488.
Smith, M. P. (1996). Shareholder Activism by Institutional Investors: Evidence From CalPERS. The Journal of Finance, 51(1), 227-252.
Wahal, S. (1996). Pension Fund Activism and Firm Performance. The Journal of Financial and Quantitative Analysis, 31(1), 1-23.
