Peter Wallison, writing in today's WSJ, says Congress should abolish the private civil action for securities fraud. He notes, among other things, that the risk of liability is the real curse of SOX 404, that companies are compelled to settle meritless securities class actions, that the real wrongdoers get off because of indemnification and insurance, and that settlements are just transfers from innocent shareholders to lawyers and complainants rather than compensation for loss.
I sympathize. Indeed, in The Sarbanes-Oxley Debacle Henry Butler and I point out, consistent with Wallison, the “litigation time bomb" inherent in SOX.
So what, then, would be the solution to securities fraud? The SEC probably can't handle it alone, or at least the idea that it could would not be politically compelling.
Criminal liability? Not if you've been reading my blog. My colleague Christine Hurt, in The Undercivilization of Corporate Law agrees, but argues that
due to laws creating additional obstacles for private plaintiffs in both federal securities law litigation and state law fiduciary duty litigation, the civil system creates an unusually high number of Type II errors in trials for the same or similar misconduct.
But are the Type II errors of private securities litigation really greater than the Type I errors given Wallison's valid criticisms?
I don't have a global substantive answer to the obvious conundrum. But in Dabit, Preemption and Choice of Law I do have this to offer (footnotes omitted):
[P]rivate securities litigation, despite its defects, should not necessarily be abolished because it is a potentially effective way to deter securities fraud. Indeed, damages and settlements in securities cases dwarf the penalties imposed by other means. In other words, the efficiency of private litigation can be assessed only in the context of all of the tools for fighting fraud.
Congress, however, has never done such a global assessment. Congress and the courts applying implied remedies have backed into a series of sometimes contradictory moves. The Court first implied broad civil remedies under 10(b) and 10b-5 without any explicit Congressional approval or detailed assessment of the costs and benefits of these remedies. Congress' first move was to enact the PSLRA and SLUSA during a stock market boom when fraud was not a major concern and litigation abuses were perceived to be the biggest problem facing securities markets. The revelation of Enron and other frauds in 2001 and 2002 convinced Congress to take some kind of action. Congress considered reversing the PSLRA limits on securities fraud cases, on the supposition that the reduced deterrent effect of securities fraud suits might have been responsible for the frauds. But Congress chose instead to increase disclosure regulation and passed the Sarbanes-Oxley Act. Meanwhile, the concern with litigation abuse remains, figuring not only in Dabit, but also in Dura Pharmaceuticals, Inc. v. Broudo, which limited the fraud-on-the-market theory by tightening the loss causation requirement.
This chain of actions has created a system in which huge penalties are assessed against the victims, large bounties are paid to trial lawyers, innocent firms face burdensome new regulation under the Sarbanes-Oxley Act, and most fraudsters escape civil liability. The PSLRA not only did not fix the basic problem with burdensome securities class actions, but bolstered the argument for additional regulation when the market crashed. The only thing everybody seems to know about the scope and direction of securities regulation is that the federal government should remain in charge, as SLUSA made clear by blocking most state securities litigation. Yet federal law’s lurching policy shifts argue against a single set of answers provided by a single federal regulator, and for preserving a role for the laboratory of state laws.
The solution, I suggest, is jurisdictional choice for securities fraud actions. Specifically, I would limit federal preemption under SLUSA to state securities remedies that apply other than on a state-of-incorporation basis. This, of course, is basically consistent with Roberta Romano's suggestion a decade ago, except that it provides a legal mechanism for getting to this result that might have some political credibility against the background of proposals like Wallison's.
Of course there are problems with my suggestion. But are these problems worse than those of the other alternatives?
Comments