SEC Enforcement: Talking the Talk, but Walking the Walk?
Almost everyone has an opinion about securities enforcement. Many are disappointed (and even angry) that "few high level executives" have been prosecuted (criminally or even civilly) in connection with the 2008 financial crisis.1 Deep in their bunker, the SEC still has some diehards who maintain that fraud has been fully prosecuted, but, even there, attitudes are changing. The shift is much clearer at the Department of Justice (DOJ), which has just settled with JPMorgan for $13 billion and may be in hot pursuit of still unnamed defendants.2 Even if the SEC is presenting itself as a more aggressive enforcer under its new chair, questions remain about whether its behavior has truly changed.
Although there is a surplus today of opinions about how enforcement should change, there is a paucity of facts. Why is it that enforcers have underperformed? What practical steps are possible? Provocative new proposals are being made, but they too need to be informed by better factual evidence as to how regulators actually behave. Last week, speaking before the New York City Bar's Second Annual Institute on Securities Litigation and Enforcement as its keynote speaker, U.S. District Judge Jed Rakoff of the Southern District of New York offered a bold proposal and a critique. After rejecting some frequently invoked explanations for prosecutorial inaction (such as the "revolving door" theory), he presented an alternative theory: namely, that federal prosecutors have relied too much on deferred prosecution agreements under which they permit independent counsel to conduct internal corporate investigations to establish the basic facts.3 These investigations, often headed up by a former U.S. attorney or other eminent person, tend predictably not to find that senior executives behaved culpably. In the judge's view, this delegation of fact-finding to internal investigators has resulted in a loss of general deterrence that "far outweighs the prophylactic benefits of imposing internal compliance measures that are often little more than window dressing."4 Although the judge carefully took no position on whether fraud was at the heart of the financial crisis, he concluded that, to the extent it may have been, "the failure of the government to bring to justice those responsible for such colossal fraud bespeaks weaknesses in our prosecutorial system that need to be addressed."5
The judge's words will resonate with many, particularly those concerned that we have quietly accepted a double standard for criminal justice under which banks (and their executives) have been treated as "too big to jail." Nonetheless, there remains a missing factual foundation here. How is the government using its enforcement resources? Further, what is the alternative to internal investigations by counsel chosen by the corporation? How can the government investigate complex corporate misconduct when it largely lacks the manpower to conduct wide-ranging corporate investigations? How should the government prioritize among cases? To ask these questions is not to disagree with Rakoff that financial executives should be subject to the criminal law, but it is to highlight the key empirical questions that usually wind up submerged and ignored in normative debates. A necessary starting point is to ask: How in fact does the government allocate its scarce enforcement resources? Invariably, the government asserts that it is tough and getting tougher. But what the government says and what it actually does can diverge significantly.
This column will limit its focus to the SEC, but that may be where the problem is most serious. To any knowledgeable observer, it is obvious that the SEC is more seriously resource-constrained than the DOJ (because even a conservative Congress is sufficiently "law-and-order" oriented that it will not cut the DOJ's budget as severely as it may the SEC's).
The SEC's predicament is that, to justify a larger budget (which it clearly needs), the SEC must show a skeptical (and Republican) House of Representatives that it has done more. This need to improve on last year appears to have made the SEC very concerned with quantitative metrics (e.g., How many cases filed? How many settled in the fiscal year?). Much like a company approaching its IPO, the SEC has an incentive to inflate its numbers. At first glance, the SEC's recent numbers do look much improved. In fiscal year 2012, the SEC brought 734 enforcement actions—a near record.6 Also in fiscal year 2012, the SEC entered into 714 settlements—up 6.6 percent from 2011 and the highest number since 2007.7
But a closer look raises serious questions about whether these numbers have been padded. In an October 2013 story, investigative reporters at the Wall Street Journal found that on the last day of its fiscal 2012 year, the SEC filed some 22 enforcement actions.8 Moreover, for the last month of that fiscal year (i.e., September 2012), the SEC initiated some 128 administrative actions—up 86 percent from the same month the preceding year.9 Many of these actions were simply "follow on" actions in which the SEC simply bars a broker or investment adviser from the industry based on a prior conviction or enforcement action.
Alone, this padding might not mean much. But there are other, more significant problems with many of the cases that the SEC brings. According to NERA Economic Consulting, from 2003 to 2010, slightly over 40 percent of SEC settlements included no monetary payment.10 In fiscal 2012, this percentage of "zero dollar" settlements fell to approximately 34 percent of individual settlements and roughly 24 percent of company settlements.11 Improved as that is, such bloodless settlements (particularly in the case of corporate defendants that are seldom impecunious) raise a puzzling question: Why does the SEC regularly sue defendants when it is willing to settle for nothing? Arguably, this looks like illusory enforcement.
Moreover, in FY 2012, NERA found that the median company settlement (in those cases where cash was paid) fell some 28 percent to $1 million, down from $1.4 million in FY 2011.12 In addition, there were 20 fewer settlements with companies in FY 2012 in comparison with FY 2011.13 Even the way these median values are computed seems dubious. Computing the median value by considering only those settlements in which some cash was paid (and ignoring the numerous settlements in which none was paid) is much like my computing my batting average by counting only the plate appearances in which I did not strike out. (By the way, on that basis, I look much better).
Perhaps more important than the declining number and median size of corporate settlements in FY 2012 is the fact that the number of "high value" corporate settlements has fallen way below where it was in the period from 2003 to 2005. "High value" settlements are best measured by looking to the 90th percentile of all SEC company settlements. The 90th percentile figure was $80 million in FY 2003, $50 million in FY 2004, and $72.5 million in FY 2005.14 Yet, by FY 2011, it had fallen to $16.4 million, while in FY 2012, it rose modestly to $18.9 million—still less than one-quarter of the level in FY 2003.15 This fall in the size of the "high value" settlement means fewer "big" cases are being brought and likely measures the impact of resource constraints on the SEC. Unable credibly to threaten to go to trial, the SEC's staff may have to settle at reduced amounts, as defense counsel (who usually are alumni of the agency) are well aware of the SEC's logistical constraints.
The data that seems most inconsistent with the SEC's claim to be a tougher, more aggressive enforcer has just been released by the Morvillo, Abramowitz, Grand, Iason & Anello law firm. They have tabulated the number of enforcement actions filed by the SEC between Jan. 1, 2013 and Sept. 30, 2013 and find that some 526 enforcement actions were brought in this interval.16 But the majority of these actions do not involve current investigations of fraud or other securities law violations. Rather, some 30 percent of these cases (or 159 out of 526) were "follow-on" cases—that is, administrative proceedings to revoke licenses that follow after earlier SEC or DOJ cases in court.17 Another 21 percent (or 113 out of 526) were "delinquent filer" cases in which a public company was late in making a required periodic filing.18 Only 48 percent of the SEC actions in this period alleged a substantive violation of the federal securities laws. Even in these "core" cases, one-third of them (or 83 out of 259) did not allege scienter.19 This could mean either that these cases were relatively minor, or—more alarmingly—that the SEC did not allege scienter in order to facilitate settlement. Finally, even where scienter is alleged, the SEC may sue in an administrative proceeding (where the penalties tend to be lower and injunctive relief is not available). The SEC did so 27 times in the first nine months of 2013.20 Thus, out of the 526 cases brought by the SEC in 2013 (through Sept. 30, 2013), only 144 cases (or 27.3 percent) were both brought in federal court and alleged fraud.21 Again, it is uncertain whether this reflects the lower gravity of these offenses or the SEC's need to use administrative proceedings to facilitate settlement.