The message of the Dow/DuPont merger and split up is simple: No firm is today “too big to target.” Activists can see the transaction as evidence that, even in the rare case where they lose a proxy fight (as they did at DuPont last year in a squeaker), the handwriting is still on the wall, and their game plan, if appealing, will ultimately prevail. Even though Trian could not win a majority vote to seat its candidates on the DuPont board, it held onto its stake, and the DuPont board quickly ditched their CEO in the wake of that fight and then approved the offer from Dow. Dow also was under pressure (from Third Point, an even more aggressive and short-tempered activist fund). The result was a marriage made somewhere other than in heaven.
Nor does this case stand alone. Lion Point Capital has now engaged Ally Financial (the former GMAC, which did fail in the wake of the 2008 crash), notwithstanding that Ally has been classified as a “systemically important financial institution” (or “SIFI”) by the FSOC. As soon as it became clear that even a SIFI could be stalked, AIG’s stock price began to soar, as market watchers predicted that it also would be targeted by activists seeking to downsize it. Lastly, MetLife downsized itself, beating activists to the punch in its effort to avoid being also classified as a SIFI
For most practitioners, the rise in activism is good news (whether they represent corporate targets or activist hedge funds) because it means more transactions and full employment. In this same vein, undertakers might welcome a return of the plague for the business it would generate. But for the more thoughtful, the question remains: where is this transition leading? Who wins and who loses? What happens in the long-run? This column will provide a brief tour of the empirical evidence and then turn to the policy issues.
1. The Evidence on Activism
The empirical evidence is in serious dispute. The one common finding, found by all studies, is that, on the filing of a Schedule 13D by an activist fund, there is a statistically significant price increase in the target’s stock, net of the market, which is normally in the range of 6-7%, but even higher if the filing is made by a multi-member “wolf pack.” Most studies do not find any long-term gains in operating performance or stock price—unless there is a follow-on takeover or restructuring. For example, a large-scale global study by Marco Brecht, Julian Franks and others found that meaningful long-term gains depend upon the realization of an outcome, typically a takeover or a restructuring. Absent a transaction, long-term gains erode. Even the impact of “liquidity events” (such as special dividends or stock buybacks) were insignificant (or even negative) in their study, and the gains associated with corporate governance changes were trivial (unless they implied an increased likelihood of a takeover).
The one outlier in these recent studies, which does find at least some modest long-term gains and improvements in operating performance at the target, was conducted by Professor Lucian Bebchuk and his colleagues. This study follows hedge fund targets for five years after their engagements and report that profits rise, but the data is thin and inconclusive. What it chiefly shows is that the short-term gains on the Schedule 13D filing do not erode. More importantly, there is a fundamental problem in this study with causality. The targets of hedge fund activism are not randomly distributed, but rather have generally underperformed the market in the years prior to the hedge fund’s engagement. That they thereafter improve back to the market’s norm does not prove that hedge funds caused this change. It could simply be a reversion to the mean, caused mainly by the efforts of the targets themselves. Others, including my co-author Darius Palia, have criticized the Bebchuk methodology for its failure to test for this possibility by using some form of a control group. Within recent months, however, one team of researchers has pursued this control group approach and reported surprising results. Cremers, Giambano, Sepe, and Wang constructed a control group that matched the firms in the Bebchuk study, except that they were not targeted by a hedge fund. Examining the performance of the control group over the same period, they find that “the value of the firms in our control group increases more than the value of firms in the target group.” Overall, they conclude that:
“[I]n the years following the interventions of activist hedge funds, the firm value of hedge fund targets deteriorates (sizably) compared to control firms.”
In short, based on this data, activist funds not only do not improve the recovery of the underperforming firms that they target, but may impede it.
Of course, this debate will continue, and Professor Bebchuk and his colleagues have indicated that they do not accept Professor Cremers’ findings and will publish a rebuttal. So it goes in academia. Nonetheless, the key point here is that simply showing that target firms improve post-intervention (a conclusion for which the Bebchuk data provided only equivocal evidence) does not demonstrate causation, which can only be shown by matching control group procedures.
One important finding does emerge, however, from the Bebchuk data. That study finds that a significant percentage of hedge fund engagements fall into a category they call “investment-limiting interventions.” These interventions typically involve large increases in leverage and shareholder payout and major decreases in investment, and they are motivated by a desire to halt or reduce capital expenditures by the target firm, particularly investments that fall into the category of “research and development” expenditures. Why? The view take by Bebchuk and his colleagues is that corporate managers overinvest in capital expenditures, because they are irrevocably committed to “empire building,” even when it is inefficient. Thus, in their view, hedge fund activism curbs this bias and leads the firm back to an “optimal” investment policy. This view that corporate managers persistently pursue “empire building” has a long history in corporate finance and was once dominant. It rested largely on the view that because larger corporate size implied higher executive compensation levels and because managers had firm-specific human capital invested in the firm, managers were risk-averse and favored conglomerate mergers because such a structure united countercyclical, co-variant subsidiaries that reduced the risk of corporate insolvency. Also, at least in modern times, larger firm was less susceptible to a hostile takeover, thus again protecting the manager’s human capital in the firm.
This theory once made sense and possibly explained why corporate managers made inefficient, “empire-building” conglomerate acquisitions. But it is now very dated. What it misses is that, beginning in the 1990s, the nature of executive compensation changed radically in U.S. firms, shifting from cash to equity compensation. Not only did the level of compensation soar, but the CEO of a large corporation today receives something like 63% of his or her compensation in equity. This gives the CEO very different incentives than in the past. Today, the CEO is incentivized to maximize the stock price of the firm, which will be retarded if the firm makes inefficient, empire-building acquisitions. As a result, the assumption that hedge fund activism is curbing inefficient empire-building seems dubious and unsupported.
Some data about the impact of hedge fund activism is clear: namely, its impact on research and development (“R&D”). One study by Allaire and Dauphin used the FactSet database to track the impact of a hedge fund “engagement” on R&D expenditures and found that over the four-year period following a hedge fund engagement, R&D expenditures at “surviving” target firms declined by more than 50% (expressed as a percentage of sales). This statistic likely understates the full impact, as not all target firms “survive” (i.e., they are acquired in a merger or they are broken-up in a restructuring), and in these cases the decline in R&D expenditures (although not measurable from financial reports) is almost certainly greater. This study did use a control, and in the control group R&D expenditures actually rose (modestly) over the same period, thus suggesting that causation is clear.
Even defenders of hedge fund activism find that R&D expenditures decline in the wake of a hedge fund engagement, but they have a justification. One study by Brav, Jiang, Ma and Tian acknowledges the decline in R&D expenditures after a hedge fund engagement, but argues that “innovation output” at targeted firms increased, at least as measured by patent applications and patent citations. In effect, they contend that targeted firms produce “more for less,” but the methodological problems in relying on such data to reach policy conclusions are enormous. Moreover, even if hedge fund activism made investments in R&D more profitable at target firms, this would not resolve the public policy inherent in the fact that hedge fund activism appears to be significantly diminishing total U.S. investment in research and development. The real problem from a public policy perspective is that R&D produces positive externalities. That is, society benefits more than the innovating firm does, because that firm cannot capture all the benefits from its innovation. Suppose for example that pharmaceutical Firm X discover a new drug. It will profit, but so will other firms that discover new applications, variations, or improvements on this new product. Thus, the social benefit is greater than the private benefit. But, as a result, if hedge fund activism reduces investment in R&D, it causes a social injury, even if it makes investment in R&D more profitable for the firms targeted by them.
In fairness, a revolution may be in progress within the public corporation, of which hedge fund activism is only the spearhead. This broader transition is redirecting corporate capital away from investment in capital expenditures and toward payout to shareholders. A recent study by the Roosevelt Institute phrased it this way:
“In the 1960s, an additional dollar of earnings or borrowing was associated with about a 40 cent increase in investment. In recent years, the same dollar is associated with less than 10 cents of additional investment.”
Under pressure, public corporations appear to be curbing investment in favor of shareholder payout. For example, the same Roosevelt Institute study reported that, from the second half of 2009 to 2013, corporations borrowed $900 billion, but paid out $790 billion to shareholders (while investing only $400 billion over this period). Thus, while corporate leverage is increasing (another consequence of hedge fund activism), it is funding dividends and buybacks, not long-term investment.
In 2015, the Wall Street Journal investigated the same shift and hired S&P Capital IQ, a research firm, to track firms in the S&P index. Their study found that S&P index firms “increased their spending on dividends and buybacks to a median 38% of operating cash flow in 2013, up from 18% in 2003.” In short, over ten years, shareholder payout more than doubled as a percentage of cash flow at the largest U.S. firms. At the same time, these S&P firms cut “spending on plants and equipment to 29% of operating cash flow from 33% in 2003.” Targets of activism cut the most, reducing capital expenditures in the five years after activists engaged them to 29% of operating cash flow, down from 42% the year before the intervention,  At the same time, these targeted firms boosted their payout through dividends and buybacks to 37% of operating cash flow in the year after the intervention(up from 22% in the year before the intervention).
Ultimately, the deeper question becomes whether the U.S. public corporation in the 21st century can fund R&D (or even retain its capital). The ability to fund R&D was arguably the great comparative advantage of the U.S. economy over the last half century: uniquely, the U.S. economy could create a Silicon Valley and invest in research that took years before it paid off. Today, at a time when activist hedge funds are multiplying like algae in a petri dish and when all are chasing the same limited number of targets, seeking to curb their investment in capital expenditures, it is debatable whether Silicon Valley could rise again under these circumstances.
2. Policy Options
If investors want corporations to prioritize payout over investment, they will eventually have their way, and neither Delaware nor the SEC are likely to stop them. But it is far from clear that investors have such a preference. Indeed, at various moments in 2015, prominent asset managers, such as BlackRock and Vanguard, issued public broadsides criticizing some investors for their preoccupation with the short-term. In the DuPont proxy fight last year, a coalition of BlackRock, Vanguard and State Street gave DuPont an initial (if short-lived) victory in a horserace that was decided by a nose.
So why then are activist funds such a formidable foe who usually win contested proxy fights? One answer is that they are generally not seeking control, but only a few seats on the board. Many will vote to stir up a sleepy management that has underperformed. But another answer may be even more important: an activist hedge fund can easily assemble a “wolf pack” of 20% to 30% (or possibly more) because a short-term profit is virtually riskless. That is, because an abnormal trading gain of 6 to 8% is a statistical near-certainty on the day the Schedule 13D is announced, others will join the wolf pack to get in on the profit. Some may leave soon thereafter, but most will stick around, at least long enough to see if a takeover bid is likely to be forthcoming.
This riskless profit may generate an excessive incentive for activism, but it is dependent on three factors: (1) that tipping and trading on such information does not amount to unlawful insider trading; (2) that the long 10 day window between when a shareholder crosses the five percent beneficial ownership level and when it must file its Schedule 13D gives those in the activist community sufficient time to learn and trade on this information of a forthcoming Schedule 13D filing; and (3) that the informed players in this recurring ritual do not constitute a “group” for purposes of the Williams Act (as poison pills would typically be triggered if the “wolf pack” were a “group”). If any one of these three could be chilled or modified, the “wolf pack” tactic would be less formidable.
The legal conclusion that a prospective Schedule 13D filer can “tip” others of its intent follows logically from the Dirks case, which makes a breach of some fiduciary-like duty a prerequisite to insider trading liability. For a time, decisions in the Second Circuit seemed to be softening the Dirk’s standard, but the Newman decision makes it clear beyond argument that the SEC must show that a tippee paid or promised some personal benefit to the tipper in order to establish liability. Under Dirks, as interpreted by Newman, this precondition for liability is largely missing from communications among activist funds. The activist organizing a “wolf pack” owes no fiduciary duty to the target’s shareholders, and its communications with other funds is arguably for a legitimate purpose that benefits its own shareholders (i.e., seeking to organize a successful proxy fight to increase the value of the target). No payment or promise of a “personal benefit” is being made by the tippee to the tipper, unless someone is so extraordinarily dumb as to promise a reciprocal tip.
To be sure, the SEC’s Enforcement Division has grumbled, suggesting that it feels organizers of a “wolf pack” can step over the line. But it has not announced any credible theory for its position, and the SEC has learned painfully not to undertake high-risk cases in this area. Much as the SEC may mutter, it is hesitant to act in a way that could embarrass itself. Thus, counsel to hedge funds advise caution and suggest that discussions among funds should be delayed until after a Schedule 13D is first filed. Still, the empirical evidence is that an extraordinary level of trading occurs during the ten day window period before the Schedule 13D is filed. Once, it was thought that this high level of trading during the window period was the work of a prospective bidder or proxy contestant, but the latest evidence is that those filing a Schedule 13D tend to trade only on the day that they cross 5% and the day after. Thus, the super-active trading during the remainder of the window period has to be the work of others, who are either informed traders or persons extremely skilled at decoding what is going on. The bottom line then is that, despite the SEC’s bluster, informed trading during the window period is normal. The SEC’s threats of litigation cause most to keep a low profile and not acknowledge their trading or tipping, but it occurs. Indeed, maybe all the SEC expects to achieve is to keep such trading out of sight, as they are legally powerless in most cases. This could change, as the Supreme Court has granted cert in the Salman case and will reconsider insider trading, but the facts of Salman involve a very simple fact pattern that is unlikely to extend the net of liability much.
If the SEC cannot stop informed trading surrounding the formation of a wolf pack, it could shorten the current 10-day window. Congress expressly gave them that power in Dodd-Frank, but the SEC seems unwilling to use its new authority. It knows that to do so would expose it to outraged criticism from institutional investors and knee-jerk academics, who both believe that activists are doing the Lord’s work. Both the U.K. and Australia have much shorter window periods, and the “wolf pack” is not readily observed in these jurisdictions. In the past, the SEC would have decided which side in this debate it favored and acted, but today avoiding criticism is an end in itself for the contemporary Commission.
This brings us to the final possibility: that the SEC could redefine the concept of “group” under Section 13(d)(3) of the Securities Exchange Act so as to reach the “wolf pack.” Here, the justification would be that the “wolf pack” is simply a device by which sophisticated parties can evade disclosure, and the SEC needs to respond to new developments. To be sure, some Second Circuit case law has defined the “group” concept narrowly. But the Second Circuit would probably give some measure of Chevron deference to revised SEC rules. What should a revised definition of “group” say? One possibility is that it could indicate that intentional tipping of a plan to organize a “wolf pack”-like coalition and file a Schedule 13D (and subsequent trading by the tippee) should be seen as strong circumstantial evidence that tipper and tippee were part of a “group.” Nothing else explains why the “tip” was given. This presumption could be rebutted, but its very existence would chill group formation. Absent such a rule, shortening the 10-day window might accomplish little, because sophisticated hedge funds could each buy up to 4.99%, file nothing, and deny that any “group” had ever been formed.
Of the various options, there has been much discussion of shortening the 10-day window, but none of changing the definition of “group.” This column has suggested that that may be the most important reform. If the SEC does not act (a near certainty), then, after the election, President Hillary Clinton, who has already criticized “hit and run activists,” might find this area one of the few where she could reach agreement with a Republican Congress.
 For a much cited study finding a 7-8% abnormal gain, see A. Brav, W. Jiang, F. Partnoy, and R.S. Thomas, Hedge Fund Activism, Corporate Governance and Firm Performance, 63 J. Fin. 1729 (2008); see also L. Bebchuk, A.Brav, and W. Jiang, The Long-Term Effects of Hedge Fund Activism, 115 Colum. L. Rev. 1085, 1122 (2015). For possibly the fullest study of Schedule 13D filings, see U.V. Lilienfeld-Toal and J. Schnitzler, What is Special About Hedge Fund Activism? Evidence from 13-D filings, at 2 and 25 (available at http://ssrn.com/abstract=2506704)(2014)(finding an 7% abnormal return on initial Schedule 13D filings and 4% on amendments).
 See Bebchuk, Brav, and Jiang, supra note 1, at 4-5.
 Professor Palia was the first academic to insist on the need for a matching protocol (such as propensity score matching or “neighborhood” matching). See J. Coffee and D. Palia, The Wolf at the Door: The Impact of Hedge Fund Activism on Corporate Governance (available at http://ssrn.com/abstract=2656325)(2015) at 86 and notes 173 to 174.
 Id at 7.
 Id. at 28.
 See Bebchuk, Brav and Jiang, supra note 1, at 1136-1139 (finding that 19% of activist interventions fall in this category). However, Professor Bebchuk and his colleagues use very narrow criteria to define “investment-limiting” engagements, defining them to include only engagements that cause the company to fall into the 5% most leveraged, or the 5% lowest investments, etc. If their criteria were expanded to look instead to the bottom or top quarter, respectively, a majority of engagements might well fall into this category.
 Id at 1137 n. 103 (noting that “managers have a tendency to invest excessively”). Even if true, this leaves open the question of whether activists seek to reduce R&D expenditures “excessively.”
 See generally Robin Marris, THE ECONOMIC THEORY OF MANAGERIAL CAPITALISM (1967); Oliver Williamson, Managerial Discretion and Business Behavior, 53 Am. Econ. Rev. 1032, 1055 (1963).
 For an overview of this transition, see Carole Frydman and Dirk Jenter, CEO Compensation, 2 Ann. Rev. Fin. Econ. 75-102 (201) at Figure 1.
 See Equilar, CEO Pay Strategies Report (2014) at p. 4.
 See Yvon Allaire and Francois Dauphin, “Hedge Fund Activism: Preliminary Results and Some New Empirical Evidence” (Institute for the Governance of Public and Private Organizations) at 24 (finding an average decrease from 17.34% to 8.12%, as a percentage of sales).
 Id. Over the same four year period, the control group’s investment in R&D went from 6.54% to 7.65% (of sales)—a modest improvement.
 See A. Brav, W. Jiang, S. Ma and X. Tian, Shareholder Power and Corporation Innovation: Evidence from Hedge Fund Activism (available at http://ssrn.com/abstract=2409404)(2014)(finding targeted firms filed 15.3% more patent applications compared to a matched sample).
 Other scholars have reported that attempts to use patent citations to measure innovation are confounded by serious methodological problems. See J. Lerner and A. Seru, “The Use and Misuse of Patent Data: Issues for Corporate Finance and Beyond” (Harvard Economics Department Working Paper, March 2015).
 See J.W. Mason, Disgorge the Cash: The Disconnect Between Corporate Borrowing and Investment at p. 19 (The Roosevelt Institute 2015)
 Id at 3. The sum of $740 billion (payout) and $400 billion (investment) is $1140 billion and the difference between that number and the $900 billion total borrowings is presumably accounted for by corporate earnings and stock issuances.
 See V. Monga, D. Benoit and T. Francis, “Firms Send Cash Back to Shareholders—Activists Push for Returns, Fueling Worries About Long-Term Investments,” The Wall Street Journal, May 27, 2015 at A-1.
 In April 2015, BlackRock’s CEO Lawrence Fink wrote a much publicized letter to the CEO of 500 of the nation’s largest corporations, criticizing the “short-term” orientation of activist investors. See A.R. Sorkin, “Dealbook: BlacRock’s Chief, Lawrence Fink, Urges Other CEOs to Stop Being So Nice to Investors,” N.Y. Times, April 12, 2015.
 Dirks v. Sec, 463 U.S. 646 (1983).
 United States v. Newman, 773 F. 3d 438 (2d Cir 2014).
 See Brav, Jiang, Partnoy and Thomas, supra note 1, at p. 1756 (showing that volume of trading during window period is double that both before and after window period)
 See Bebchuck, Brav, Jackson and Jiang, Pre-Disclosure Accumulation by Activist Investors: Evidence and Policy, 39 J. Corporation Law 1, 6 (2013).
 In United States v. Salman, 792 F. 3d 1087 (9th Cir 2015), the Ninth Circuit upheld the conviction where one brother tipped his sibling who then passed the information along to their brother-in-law. This is the simple “gift” case that Dirks expressly held criminal. Unlike Newman, the facts of Salman present the ideal fact pattern for the Department of Justice.
 See Section 929R of the Dodd-Frank Act of 2010, which amended Section 13(d)(1) of the Securities Exchange Act of 1934 to authorize the SEC to shorten the 10 day window, as it chose.
 See, for example, Hallwood Realty Partners, L.P. v. Gotham Partners L.P., 286 F.3d 613, 616-18 (2d Cir. 2002).
The preceding post comes to us from John C. Coffee, Jr., the Adolf A. Berle Professor of Law at Columbia University Law School and Director of its Center on Corporate Governance.