In a move that appears at once to be shrewd, savvy and largely symbolic, the SEC has modified its longstanding policy that it will not require a defendant to admit or deny liability, or facts that might establish its liability, in a settlement with the SEC. Now, such an admission may be required "when appropriate."1 Whatever the outcome in the SEC's mandamus appeal of Judge Jed S. Rakoff's Citigroup decision,2 Rakoff has effectively won the war, even if he loses the Citigroup battle. Although denying that Rakoff influenced them, the SEC conceded (effectively, if not formally) that its policy was simply too weak and equivocal.
Of course, personnel in the SEC Division of Enforcement may still think such an admission is seldom "appropriate." Indeed, overworked, understaffed, and experienced more at settling cases than taking them to trial, the "Division of Settlement" (as defense practitioners dub it) has every reason to go slow in implementing this change in policy. But, eventually, they will have to obtain some admissions in at least a few cases to make good on SEC Chairwoman Mary Jo White's promise.
This column will focus on what happens then. In the recent debate over the SEC's enforcement policy, many commentators have opined that requiring any admission by the defendant would expose it to astronomic liability in follow-on securities class actions. They have specifically claimed that an admission in an SEC settlement would result in either (1) offensive collateral estoppel, or (2) an admission that could be presented to the jury in a way that would make the case indefensible. As next discussed, the first conclusion is simply wrong, and the second is doubtful.
Those who have argued that offensive collateral estoppel could result from an admission are undoubtedly relying of Parklane Hosiery v. Shore.3 In that 1979 U.S. Supreme Court decision, the court did permit plaintiffs in a proxy fraud class action to assert offensive collateral estoppel based on an earlier SEC action (but gave the district court broad discretion to determine when collateral estoppel would be unfair). The key distinction, however, was that the earlier SEC action did not settle, but rather went to trial (and the SEC won a declaratory judgment). As the court said, estoppel is appropriate when "the party against whom estoppel is asserted has litigated questions of fact and has had the facts determined against him in an earlier proceeding."4 Settlements do not satisfy this requirement of a full adjudication and so do not trigger collateral estoppel.
But, even if offensive collateral estoppel does not result, can the plaintiffs introduce into evidence the express finding in the SEC's settlement agreement that the defendant violated the federal securities laws. Such an admission, if admissible in evidence, could resolve many cases as a practical matter. But here we encounter Federal Rule of Evidence 408 ("Compromise Offers and Negotiations"). That rule bars, with one notable exception, the admission of evidence relating to conduct or statements made during settlement negotiations that would otherwise be admissible "to prove or disprove the validity or amount of a disputed claim or to impeach by a prior inconsistent statement or an admission."5 Although this rule broadly precludes the admission of statements or conduct made "during compromise negotiations," the language of the rule arguably leaves open the status of a final settlement, including one approved by the court after a full hearing. Suppose then that defendants acknowledge in a frank and broad admission that they made material misstatements in a prospectus or offering document, and the settlement is approved, after a hearing, by no less a judge than Judge Rakoff, himself. Is this settlement now admissible?
The short answer is: Probably not under the existing case law, but the precise case has not yet been squarely faced. Both in the Second Circuit and elsewhere, courts have read Rule 408 as codifying a broader policy that seeks to encourage dispute resolution by barring the admissibility of even completed settlement agreements.6 Most of these cases involve a settlement agreement between the defendant and one plaintiff that another plaintiff whose alleged injury relates to the same transaction wishes to admit into evidence. Most of these attempts have failed with the court refusing to admit the evidence of the settlement agreement, but none have truly involved a settlement agreement in which the defendant admits its culpability in a settlement with a regulatory agency. Thus, some doubt still persists, particularly in a case in which the regulatory agency places the admission front and center in its negotiations with the defendant—and the settlement is approved by a federal court.7
How then is a court likely to evaluate the admissibility of this very sensitive and carefully negotiated admission? Here, we need to consider the one major exception in Rule 908. By its terms, Rule 908 does not apply to:
conduct or a statement made during compromise negotiations about the claim—except when offered in a criminal case and when the negotiations related to a claim by a public office in the exercise of its regulatory, investigative, or enforcement authority.8
This language fits like a glove the case of an admission in an SEC consent decree when federal prosecutors seek to use it in a later federal criminal prosecution. Such cases have arisen and have upheld the admissibility of the settlement agreement.9 But this exception cuts both ways. First, it certainly warns defense counsel that an admission in a SEC settlement and consent decree could haunt them in a later federal criminal prosecution. Conversely, this exception is also a negative pregnant. That is, if it expressly permits the admission of an SEC settlement in a criminal trial, it clearly does not cover the admissibility of that same settlement in a follow-up private class action. Defendants can reasonably argue that the drafters of the Federal Rules of Evidence knew when they wanted to authorize the admission into evidence of a civil settlement, and, outside the criminal law context, they chose to preclude the admissibility of even regulatory settlements.
Still, even if an admission in an SEC settlement is not admissible in a subsequent civil trial, this does not mean that it will have no impact. To the contrary, plaintiffs in the class action can be expected to point to the admission in briefs and pleadings. Courts have long shown a marked tendency not to dismiss class actions brought in the wake of SEC consent decrees, and the settlement amounts in those cases are typically higher. Particularly if only a handful of SEC settlements contain an admission by the defendant, one can reasonably expect that courts will view these cases as ones in which the SEC at least suspected a serious fraud. In the Southern District of New York, the majority of securities class actions are not dismissed on a motion to dismiss, and the chances of a dismissal on either a motion to dismiss or for summary judgment will predictably decline markedly when the defendant admits misconduct to the SEC in a prior proceeding.
Beyond even the impact on pre-trial proceedings, the possibility also arises that defendants can be embarrassed by their firm's admission, even if it is not directly admissible. Suppose the corporate defendant admitted that it failed to disclose material adverse information. On the deposition of the chief financial officer in a follow-on class action, plaintiffs might ask the CFO whether he believes the company omitted to disclose material adverse information. This question poses a delicate tightrope for the CFO, as he cannot deny the prior settlement or admission without violating the standard SEC consent decree. The CFO can possibly answer that he does not know, but even this answer can undercut his side's ability to contest the key issues at trial.
So what can defendants do when the SEC insists on an admission as the price of settlement? The options are essentially three: First, defendants can go to trial. But this carries the risk of offensive collateral estoppel if they lose under the Parklane Hosiery precedent. Second, they can offer a more attractive financial settlement if the SEC will drop its request for admissions (and the cost of an SEC settlement may rise for defendants when the SEC is prepared to seriously demand an admission). Third, they can seek to negotiate an admission that effectively admits nothing. To illustrate this last option, suppose the SEC sues under §17 of the Securities Act of 1933, which does not generally require proof of scienter. The defendant could admit liability under §17 and, more specifically, admit that it failed to disclose material information—and still escape automatic liability in follow-on private litigation. This is because plaintiffs must normally sue under Rule 10b-5 (not §17, which has no private cause of action). In a Rule 10b-5 case, plaintiffs must, first, plead with particularity a strong inference of fraud to avoid dismissal and obtain discovery and, second, prove scienter at trial. In this context, an admission as to materiality may have little real impact in a fraud-based action. Why would the SEC settle for such an empty admission? Possibly either because it was either risk averse about trials or because it did not feel able to commit the resources necessary for a trial in the particular case.
This scenario of "token" admissions that do not increase the prospect of private liability may seem remote, but it should remind us that there is no panacea. More must be done than revising the "neither admit nor deny" policy. Weak, slap-on-the-wrist settlements can continue under any system of legal rules. Thus, it is useful to ask what goals should be given a priority. In my judgment, Rakoff's decisions in Bank of America and Citigroup were not primarily motivated (and should not have been primarily motivated) by a desire to subject the defendants in SEC enforcement actions to higher private liability in follow-on class actions. Rather, the more important and appropriate goal was to increase the transparency of the process (without which there cannot be accountability). If so, the fact that admissions in SEC settlements may not be admissible in civil trials is not critical (particularly when, as this column has also stressed, such admissions are still likely to increase the size of private settlements in other ways).
But if greater transparency is the goal, ending the "neither admit nor deny" convention is only a partial and inadequate step in that direction. The even more serious obstacle to transparency is the way SEC settlements are negotiated. Typically, on the same day, a complaint, an answer and a stipulation of settlement are filed in a federal district court, and the court grants the requested injunction after only a cursory review. This settlement may have been the product of a year-long negotiation following the SEC's issuance of a Wells notice to the defendant. But this process is opaque and hidden from public view, much like the proverbial bottom of the iceberg. The problem here is that the complaint in a typical SEC enforcement action is often an artifact of the settlement negotiations. To settle, defendants insist on editing what the SEC alleges, and as a result the public never learns what the SEC believed actually happened.
To be sure, settlement negotiations must occur in private. But if we compare SEC settlements with the federal criminal justice system, a disparity is obvious between white-collar prosecutions and SEC enforcement actions. It is not the Department of Justice's standard operating procedure to negotiate the contents of the indictment with the defendant. Eventually, there may be a plea bargain to a lesser included count, but one learns at the outset what motivated the DOJ to indict. In contrast, many SEC complaints allege only scienter-less violations of §17 of the Securities Act or a failure to maintain adequate books or records or to supervise employees in a brokerage firm. If the SEC caught John Dillinger parked outside a Federal Reserve Bank with a machine gun in his car, their complaint might only allege double parking in front of a federally insured bank.
In this light, if the SEC begins, even sparingly, to require more admissions as the price of settlement and if the defense bar responds by refusing to settle (as some defendants will), the result may be to encourage the filing of non-consensual complaints that are more candid and not the product of a quiet plea bargain. That would be good if one wants to maximize transparency, but bad if one is an enforcement attorney who prefers to enjoy the quiet life.
The bottom line on the SEC's new policy is that if the SEC were to seriously pursue more admission in settlements (which is not yet likely), it would probably have to bring fewer actions or otherwise economize because it has limited enforcement resources. Last year (fiscal year 2012), the SEC brought over 700 enforcement actions. If it had sought admissions in even 10 percent of these cases, it might have had to bring significantly fewer cases. The tradeoff here is between quality and quantity. The SEC can bring many "slap-on-the-wrist" cases, obtaining what Judge Rakoff termed "pocket change" in his Citigroup decision. Or, it can pursue individual defendants seriously, often seeking to bar them from the industry. Such actions will predictably be fought fiercely (at least by the individual defendants, as the pending case against Fabrice Tourre illustrates). The SEC's traditional rationalization has been that it should settle quickly to speed compensation to victims. But the cost of maximizing compensation is that we minimize deterrence. Moreover, the real compensation comes typically not from the SEC, but from private plaintiffs in follow-on class actions (who will be greatly benefitted if the SEC obtains admissions). Deterrence is what the SEC should instead be seeking to maximize, and additional compensation may follow in its wake if the SEC does so.
On the practical policy level, to the extent that the SEC does need to husband its limited resources, the most sensible option may be for it to make greater use of the administrative proceedings. Its authority in this area was vastly expanded by the Dodd-Frank Act,10 but it so far has not made significantly greater use of this option. Administrative cases could be resolved more quickly and with less effort, and frankly many SEC cases do not require that the defendant be hauled into federal court.
Why then has the SEC not made greater use of this option? One possibility is that it is very slow to change and remains committed to doing things in the traditional way. Another is that it wants the favorable publicity that it generates from civil settlements in federal court to make its case for greater budgetary allocations from Congress. But Congress, much like the press, has also heard and understood Rakoff's message. Doing things in the traditional way is just not working for the SEC, and it cannot continue to pretend that hollow victories justify a victory parade.
John C. Coffee Jr. is the Adolf A. Berle Professor of Law at Columbia University Law School and Director of its Center on Corporate Governance.
1. On June 18, 2013, SEC Chairwoman Mary Jo White announced that the SEC policy would be changing at a conference sponsored by The Wall Street Journal and the CFQ Network. She indicated that the SEC would require an admission or acknowledgment of misconduct in cases involving harm to a large number of investors or in which a "heightened accountability or acceptance of responsibility through the defendant's admission of misconduct may be appropriate." See Joseph Edmondson Jr. and Barry J. Mandel, Monday Business Briefing, July 12, 2013.
2. See SEC v. Citigroup Global Markets, 827 F. Supp. 2d 328 (S.D.N.Y. 2011), stay granted, 673 F.3d 158 (2d Cir. 2012).
3. 439 U.S. 322 (1979)
4. Id. at 335.
5. See Federal Rule of Evidence 408(a).
6. See Weinstein's Federal Evidence, §408.04 (2d ed. 2013); (citing American Ins. v. North American, 697 F.2d 79, 82 (2d Cir. 1982); Alpex Computer v. Nintendo; 770 F. Supp. 161, 165-66 (S.D.N.Y. 1991); Playboy Enterprises v. Chuckleberry Publishing, 486 F. Supp. 414, 423 n.10 (S.D.N.Y. 1980); Buckman v. Bombardier, 893 F. Supp. 547 (E.D.N.C. 1995). But see Bank of America Nat'l Trust & Savings Ass'n v. Hotel Ruttenhouse Assocs., 800 F.2d 339, 345 (3d Cir. 1986) (once disclosed in court, judicially approved settlement is subject to discovery).
7. One bankruptcy court has ruled that the rules of evidence do not apply to final judgments entered in a prior judicial proceeding. See In re Chuck, 165 B.R. 1005 (W.D. Tex. 1993), aff'd 20 F.3d 1170 (5th Cir. 1994). Also, settlements can be introduced to show facts other than liability (such as that the defendant was on notice). See Brady v.Wal-Mart Stores, 455 F. Supp. 2d 157 (S.D.N.Y. 2006), aff'd 551 F.3d 127 (2d Cir. 2006).
8. Federal Rule of Evidence 408(a)(2) (emphasis added). There are also additional exceptions in Rule 408(b).
9. See, e.g., United States v. Prewitt, 34 F.3d 436, 439 (7th Cir. 1994) (securities enforcement settlement admissible in criminal proceeding).
10. The Dodd-Frank Act vastly expanded the SEC's authority to seek cease and desist orders against any person. See Securities Exchange Act of 1934 §21C.