Robert Khuzami: nothing to see here

SEC enforcement chief Robert Khuzami testified before Congress to explain why the SEC allows lawbreakers to settle without admitting guilt. His threadbare rationalization has been given by regulators for decades, even as financial crimes and misdemeanors have become more common and more dangerous. It’s simply amazing that in the face of massive criticism Khuzami fails to offer any rationale applicable to the situation today. But he is on the short list to take over as Chair of the SEC, an obvious case of rewarding failure when failure was the goal of the Obama Administration.

Everyone in the financial sector knows they can get away with minor fines for the most egregious frauds, because the regulators no longer believe in beating the crap out of them. Here’s Khuzami proving that point:

Cases arising out of the financial crisis have been a particular priority of the Enforcement Division: to date, we have filed actions against 102 individuals and entities, naming 55 CEOs, CFOs, and other senior corporate officers, and obtaining orders for $2 billion.

That $2 billion Khuzami loves so much wouldn’t cover the investor losses from a randomly chosen group of ten real estate mortgage-backed securities from 2006, let alone the thousands sold by Wall Street. That group of executives doesn’t include any of the major Wall Street players. And Khuzami doesn’t even point out that individuals rarely pay significant sums; their employers pay them.

SEC Policy on Settlements

SEC policy is to accept a settlement

… only when our informed judgment tells us that the settlement agreement is within the range of outcomes we reasonably can expect if we litigate through trial. In making that determination, we take into account many factors, including: (i) the strength of the evidence and the potential defenses, including the possibility that the Commission might not prevail at trial, or prevail but be awarded less than the proposed settlement achieves; (ii) the delay in returning funds to harmed investors caused by litigation; and (iii) the resources required for a trial, including, most importantly, the opportunity costs of litigating rather than devoting those resources to investigating other cases.

That policy that might have worked in markets where reputation was crucial to survival of an investment bank, but if that were ever true in the past, it isn’t true today.

As a general rule, the parties settle if they agree on the range of outcomes and the chances for each outcome, and if they agree on the general range of costs associated with each outcome.
It’s pretty easy to evaluate a car crash case*. There are rules of thumb, and local lawyers have a pretty good idea of how things will go with local judges and juries. That’s why you see so many settlements.

Evaluating the Strength of a Securities Fraud Case

It’s hard to evaluate a securities fraud case. There is a lot of paper. The victims are always invisible, and many seem on first glance to be unappealing. The defense lawyers can throw out junk arguments, thinking that the jury won’t understand which arguments have merit and which are just smokescreens. The lawyers don’t have a clear idea of how a local judge or jury will react, and they don’t have a lot of experience with trials. The rules of thumb about cases were developed in wholly different times.

Here’s an example. The decisions to create RMBSs and how they are funded are made by one group of people. Another group decides on the specifics of the construction of the securities. Another group writes the offering materials and disclosure documents. Another group does due diligence. And yet another group does the selling. One argument that seems to work for these cases is the empty chair argument: why did you single out my guy when all those other guys were equally or more responsible?

Once it became clear that people took the empty chair argument seriously, and were furious with the SEC for singling out some schmuck, no matter how guilty, the solution was obvious. Name all of the schmucks and let them point fingers at each other. And for heaven’s sake, put blood on the floor: bring in victims to show how badly they were damaged.

Learning Nothing or the Wrong Lesson

That didn’t happen because of two specific decisions made by Khuzami. First, the SEC singled out one guy, and didn’t bother the higher-ups, especially the lawyers and accountants who made the decisions about disclosure and accounting. Second, the SEC decided only to prosecute a single example of RMBS fraud at each bank. Except, of course, Deutsche Bank, Khuzami’s former employer and issuer of RMBSs, where no case was filed.

Khuzami claims the SEC settles when it gets what it thinks is in the range of possible outcomes in a trial. Given the incompetence of their trial tactics, it may be so. But I don’t think so. The SEC should have tried a number of cases to learn what judges and juries think in these radically different circumstances. But because of Khuzami’s decision to sue each investment bank over a single RMBS, there was no opportunity to learn either new tactics or about the detailed responses of judges and juries to the horrible facts of RMBS fraud. The outcome is that cases were settled too cheaply.

The SEC knows this. It wins its silly million dollar insider trading cases because it has experience at trying lots of them. Not so with the hundreds of billions lost in RMBS cases. The SEC learned the lesson that it couldn’t win. It failed to learn the actual lesson: use better trial tactics.

Example

The SEC settled with Goldman Sachs in one of the ABACUS deals for $550 million. That’s a big number. But there were 25 ABACUS deals, with a value of $10.9 billion. The SEC only sued over one. That means that there is nothing useful for the losers in the other 24, and nothing for the investors in similar deals. If you calculate the losses for all of them, that $550 million looks like chump change. Furthermore, the ABACUS deals were masterminded by Jonathan Egol. He didn’t get sued, and was appointed as a Managing Director at Goldman Sachs. The only individual sued was Fabrice Tourre, a French guy who sold the deals. Empty chair, anyone?

The Resources Argument

Khuzami also makes an argument about resources: if the SEC settles, the resources can be used on other cases. If this had any merit, we’d certainly see more cases than we do. But it isn’t true. The lawyers go all the way through preparation for a trial, interviewing all witnesses, and considering all defenses offered in the defendant’s Wells Submission. Then they write a complaint. All that is left is to handle discovery and go to trial.

And yes, that is a load. But in February 2009, when Khuzami was appointed, it would have been easy for him to have demanded additional resources and done the necessary work. Instead, he and SEC Chair Mary Schapiro decided to use the prosecutors and investigators they had, all hired under Bush and Clinton, neither of whom wanted aggressive enforcement. He also forced a bunch of administrators back into field work instead of firing them and hiring people burning with desire to lock up cheats and frauds.

Conclusion

It is clear that Khuzami and Schapiro were doing exactly what Treasury Secretary Geithner and presumably President Obama wanted: let my bankers go.

Perhaps that is why Khuzami is on the short list for Chair of the SEC. Failing upwards is an easy way to reward the plutocrats while seeming to govern. This slovenly defense of an antiquated policy may just do the trick, but it reveals that the SEC is a thoroughly captured agency.
————-
* Suppose we have a car crash case. One outcome is total win for the plaintiff, another is a total win for the defendant, and then there are in between possibilities, partial liability for each party. Suppose the parties think that there is no possibility that the defendant gets off, a 10% chance that the plaintiff is 30% at fault, a 40% chance that the plaintiff is 10% at fault, and a 50% chance that the plaintiff is not at fault. Then suppose that the damages are set at $100K and costs of trial are $10K for each side. Then the expected outcome is easy to calculate:

.10 * .70 * 100K = $ 7,000
.40 * .90 * 100K = $36,000
.50 * 1.0 * !00K = $50,000

Total: $93,000
Less cost of trial: 10,000

The case should settle at $83,000, give or take a little. Plaintiff might take a bit less, for example, because she pays a smaller legal fee in a settlement than in a litigated case. Defendant might pay a little more for internal reasons, or because of somewhat different evaluations of the possibilities, or because the case is notorious and it might hurt future business not to pay.