FORTUNE — In a series of blistering opinions delivered over the past three years, culminating in his rejection on Monday of a $285 million proposed consent decree between the U.S. Securities and Exchange Commission and a Citigroup unit, Manhattan federal judge Jed Rakoff has been making both corporate defendants and their regulators profoundly uncomfortable. He previously held up approval an August 2009 SEC consent decree with Bank of America for about six months, and early this year he took a long, close look at a consent decree involving Vitesse Semiconductor before letting it squeak by.
There are several things bothering Judge Rakoff, and they are intertwined, which makes it difficult to tell how just how insoluble the situation might be.
The great fear among practitioners is that the judge, one of the brightest lights on the nation’s most important bench when it comes to securities litigation, is now saying that he will no longer sign SEC consent decrees unless a corporate defendant admits wrongdoing. If so, then the sky will fall, these observers say, because about 90% of SEC cases are currently concluded by consent decree, and the lynchpin for virtually every one of those is that the defendant doesn’t admit wrongdoing. If the company did admit wrongdoing, various legal doctrines would render that company a sitting duck in private class-action litigation, all but guaranteeing that the company’s liability would multiply ten-fold. It will usually make more sense for the company to fight the SEC case to the bitter end rather than capitulate to such a disastrous result. That prospect, in turn, scares the SEC, because it simply doesn’t have the manpower to litigate so many cases.
Though Rakoff’s Monday ruling can, indeed, be read that way, it can also be read — in light of his earlier rulings — less aggressively, and in a way that might still let everyone muddle through. Here I’ll try to lay out (1) what’s eating Rakoff and (2) what might conceivably placate him (short of requiring an admission of wrongdoing).
Opening the closed doors
What is unquestionable is that a procedure that has been fairly routinely followed since at least 1972 is unacceptable to Rakoff. That’s the “long-standing policy — hallowed by history, but not by reason,” as Rakoff put it Monday — in which SEC regulators and private parties work out their differences behind closed doors and then present a federal judge with a fait accompli settlement. Typically, the SEC simultaneously files a complaint against the defendant and then — that same day — files a proposed consent decree agreed upon between the parties in which the defendant neither admits nor denies the allegations of the complaint, but does acquiesce to paying a fine and promising to make certain internal reforms. The judge is then supposed to sign on the dotted line, forcing the company to make good on its promises on pain of being held in contempt of court.
So what’s wrong with that? At least three things, in Rakoff’s view.
1. He’s not a potted plant. This is not simply a statement about Rakoff’s character but, as Rakoff sees it, a constitutional imperative. There is a separation of powers issue if the SEC (which, for constitutional purposes, is part of the executive branch) is, in effect, ordering a federal judge to rubber-stamp its handiwork without engaging in meaningful, independent thought or review. And it’s impossible to engage in any such review, Rakoff is saying, when presented with a standard-issue fait accompli consent decree, because no facts have yet either been proven or admitted.
SEC-Citi settlement rejection: A public interest wake-up call
The other side of the coin here — as Rakoff himself acknowledged when he “reluctantly” agreed to sign a tweaked decree in the Bank of America case in February 2010 — is that there is also a separation of powers problem if a member of the judiciary branch, i.e., Rakoff, starts telling an executive branch agency, the SEC, how to deploy the agency’s limited resources — which is, in effect, what he’s coming close to doing. That would clearly invade the executive’s prerogatives. So there’s a delicate dance going on here, and Rakoff knows it.
2. These settlements might not serve the public interest. It’s usually assumed that if a settlement is reached through well-informed, arms-length bargaining between opposing parties, the settlement must be fair and reasonable. And Rakoff doesn’t really question that. He’s saying, if the SEC wants to work out a settlement with a defendant and then dismiss the case, that’s its business. But if it’s going to ask Rakoff to play a role — to issue an injunction and exercise the court’s contempt power — then the settlement also has to serve the public interest. And he’s not sure that the settlements he’s seeing do that.
He sees a temptation to toward a “cynical relationship between the parties,” as he put it in his first Bank of America (BAC) ruling. “The S.E.C. gets to claim that it is exposing wrongdoing on the part of the Bank of America in a high-profile merger; the Bank’s management gets to claim that they have been coerced into an onerous settlement by overzealous regulators. And all this is done at the expense, not only of the shareholders, but also of the truth.”
Rakoff also fears that high-level corporate executives who may have been bad actors themselves will be all too willing to have the company (i.e., shareholders) pay big fines so long as the SEC will agree to stop exploring just how high up the corporate ladder the wrongdoing extended. But such deals don’t serve the public’s interest.
Interestingly, while federal judges are required by law to consider the “public interest” when approving a government antitrust consent decree, there is no such express requirement when approving a securities consent decree, and the SEC contends that, as an independent government agency, it can be counted on to have already looked out for the public interest. But for Rakoff, the SEC’s view takes us back to the potted plant problem.
3. The routine language in consent decrees, whereby the corporate defendant agrees to “neither admit nor deny” the SEC’s allegations, may violate the First Amendment’s free-speech guarantees.
This is an interesting question with a comical history. If a corporation simply asserted in a consent decree that it “did not admit” wrongdoing, the SEC has long taken the view that such language would imply that the corporation still “denied” wrongdoing (which, in fact, it does) and would, as a consequence, let the company off too easy. So the SEC began requiring, decades ago, that the consent decree specifically state that the defendant “neither admits nor denies” wrongdoing.
What it will take to turn a new leaf at Citigroup
But in 1972 the SEC determined that even that clause was insufficient, because some corporations were signing the consent decrees and then turning around a few days later and issuing press releases denying wrongdoing, and claiming that they just signed the consent decree as a business decision given what it would have cost to litigate the suit. So, since 1972, the SEC has required corporations in the consent decree to promise not to deny the allegations of the complaint in public. The one glaring exception is that the company is permitted to deny the allegations where it matters most — in private civil suits brought by investors alleging the very same acts of wrongdoing that the SEC alleged.
This weird charade does not sit well with Rakoff, who thinks the practice may even violate First Amendment free speech protections. “Although we claim that these defendants have done terrible things, they refuse to admit it and we do not propose to prove it, but will simply resort to gagging their right to deny it,” he wrote last March in the Vitesse Semiconductor (VTSS) case.
How then might corporations be able to placate Rakoff without surrendering their precious desire to settle without admitting wrongdoing? Here are some possibilities:
1. Show him some hard evidence: It may be that Rakoff simply wants to see some hard evidence — depositions, emails, or mutually agreed upon facts — that give him a rational basis for exercising intelligent independent review. These would not necessarily have to be so damning as to leave the corporation legally defenseless against private suits.
After initially refusing to sign the Bank of America consent decree in September 2009, Rakoff okayed it the following February after the SEC presented him with about 50 pages of factual information whose accuracy the bank did not contest. At a minimum this enabled Rakoff to feel like he was exercising independent judgment about whether the settlement was in the public interest. In fact, Rakoff went further and wrote that it was now “clear to the court that . . . the [bank’s] proxy statement . . . failed . . . to disclose” material information. Though his findings were probably too tentative to bind the bank in later private civil litigation, they made it tougher for bank apologists to claim that it was pure as driven snow.
2. Historic cases require deeper inquiry. In the run-of-the-mill case, the public’s interest in knowing the precise truth about what happened will probably not be so high as it was in the Citigroup or Bank of America cases. The latter case involved one of the signature historical events in the financial crisis of 2008 — the shotgun sale of Merrill Lynch to Bank of America at a time when most bank shareholders had little idea how dire Merrill’s condition was or how much of the sales price would be going to pay bonuses to the very Merrill executives who had steered the company into the crisis BofA shareholders were being asked to bail it out of. Likewise, in the Citigroup (C) case the bank was accused of snookering its own customers by secretly designing and shorting a mortgage-backed CDO fund that was intended from the get-go to fail — a symbolic, hallmark transaction of an era.
And, in the Citigroup case, Rakoff said as much: “In any case like this [italics added] that touches on the transparency of financial markets whose gyrations have so depressed our economy and debilitated our lives, there is an overriding public interest in knowing the truth.” There wouldn’t necessarily be a comparable overriding interest in every case.
3. Drop the gag-order language. Go back to the pre-1972 situation in which corporations could deny wrongdoing in public if they wanted. So what? Once hard evidence has been made public showing why the SEC found wrongdoing, the corporate PR people will have a tougher row to hoe. But why not let the public hear both sides — the usual approach in a free country?
So those are my suggestions. If these measures don’t work, the SEC will probably have to eventually bring a mandamus action — a specialized form of appeal that is exceedingly hard to win — alleging that Rakoff is abusing his discretion. But if it does so, the SEC will be, in effect, asking a panel of appellate judges to tell Rakoff: Your role here is to forget about the independence of the judicial branch and just become a rubber-stamp.
Hard for me to picture that happening.