Does the Securities and Exchange Commission suffer from trialphobia?

Ever since Judge Jed S. Rakoff rejected the S.E.C.'s settlement with Citigroup over a malignant mortgage securities deal, the agency has been defending its policy to settle securities fraud cases. But the public wants a "Law & Order" moment, and who can blame them?

Of course, there was one criminal trial. Federal prosecutors in Brooklyn brought a case against two Bear Stearns hedge fund managers who blew up the firm's internal fund, eventually leading to the demise of Bear. They were acquitted.

But so far, there's been no civil trial in a major case directly related to the biggest economic fiasco of our time: the financial crisis.

The S.E.C. contends that it has received more than $1.2 billion in penalties from financial crisis cases, having accused 81 people and entities, 39 of them chief executives and other senior officers. And it doesn't avoid trials altogether. The agency has averaged almost 14 trials a year from 2008 to 2010, compared with about eight from 2001 to 2003. Finally, in cases that haven't yet gone to trial, the S.E.C. has charged some low-level bankers from big Wall Street firms — but no masters of the universe.

As for the near future, the agency might actually have a financial crisis trial. Right now, it looks as if cases against the mortgage bank IndyMac, the brokerage firm Stifel Nicolaus and the executives who blew up the Reserve Primary money market fund could go to court. But do you see the pattern? None of those is a major investment bank. The S.E.C. is just not hauling in the big boys.

That could change if the S.E.C. sued Citigroup. As Judge Rakoff noted, Citigroup is a "recidivist," repeatedly flouting securities laws. In its settlement with the bank, the agency cited only one mortgage securities deal, but as my ProPublica colleague Jake Bernstein and I wrote, there are many more that look just as rotten.

Yet the reason for putting Citigroup in the dock goes beyond the bank itself. The S.E.C. is not getting big enough settlements out of the largest banks. It's not bringing enough financial cases. It isn't going after the big banks' top executives. It's being way too cautious in its interpretation of its role as defender of the fairness and sanctity of the markets. The frustration, shared by Judge Rakoff and the rest of humanity, is all the greater because the agency rarely, if ever, gets anyone to admit guilt when they settle.

This renders the settlements little more than turning on the light in a kitchen full of roaches. Instead of teaching the banks a lesson, the settlements merely show how the bad actors are scattered everywhere — and the public watches the banks scurry into the pantry to feast some more.

To the S.E.C., this view is profoundly unfair.

The agency's message is, "if you want to resolve a case short of a contested proceeding, come in and be prepared to provide the type of relief we would obtain at the end of a trial," said Lorin L. Reisner, the S.E.C.'s deputy director of enforcement.

"And where that's not available, we'll go to the mat."

On a case-by-case basis, the S.E.C.'s argument for settling is strong. While the public loves a court case, lawyers often believe that trials are failures. They are expensive, time-consuming and capricious, especially in financial cases that are often so complex they challenge even sophisticated juries.

Generally, securities regulators can rack up more enforcement actions by settling. And the agency would do only civil trials anyway; it's the Justice Department that undertakes criminal trials, which probably are a greater deterrent to white-collar crime.

Fair enough. But here's the rub: By taking this doctrine too far, the S.E.C. has undermined its negotiating position.

Agency officials continually advertise how few resources they have, how costly trials are and how irresponsible it is to shareholders to force a trial when a reasonable settlement can be won instead.

Last month, for example, Robert S. Khuzami, the agency's head of enforcement, trumpeted the S.E.C.'s "record-breaking performance during a period of resource constraints."

In doing so, the agency has Beltway blinkers on. Sure, it's speaking to Congress, but Wall Street is also listening. When it complains, even legitimately, about its budget or how costly and difficult trials are, the S.E.C. is inadvertently showing its belly to Wall Street in a sign of submission. It's whimpering that it will shy away from a trial, afraid of draining its coffers.

When it's not signaling its fear about spending money, S.E.C. officials are often talking about how complex financial crisis cases are. In a recent conversation with James B. Stewart of The New York Times, Mr. Khuzami almost sounded as if he were Citigroup's counsel, a role for which he is well suited since he held that role at Deutsche Bank before joining the S.E.C. In that interview, he made Citigroup's case for it. But the bank's lawyers get paid enough and don't need his help.

Above all, the S.E.C. worries about losing. That means it doesn't push cases that would broaden definitions of securities fraud, the ambitious cases that penetrate the gray areas and eliminate murk as a defense against wrongdoing.

In his interview with Mr. Stewart, Mr. Khuzami worried aloud that Citigroup might have made the proper disclosure in its mortgage deal when it mentioned that it was possible the deal might have an adverse impact on its customers. Might have? It absolutely did have a clear adverse impact. If you raise an issue as a mere hypothetical when you know for a fact that it's occurring, isn't that misleading? Shouldn't that be tested? And tested in court, so that a precedent is set?

To overcome its greatest fear, the S.E.C needs to realize that it can win even if it loses. A trial against a big bank could be helpful regardless of the outcome. It would generate public interest. It would put a face on complex transactions that often are known only by abbreviations or acronyms. Litigation would cost the bank money, too. And it could cast the way Wall Street does business in such an unflattering light that even if the bank won, it might bring about better behavior.

A trial would show boldness. And when the S.E.C. found itself at the negotiating table again, it would feel a new respect.