From time to time, I make a foray into expert-witness work. It's a good deal of fun to shift from the more abstract pursuit of strategic policy analysis to the hard-knuckled, win-or-lose atmosphere of the courtroom, especially when I think my advice helps to get a win.

Each case, of course, is different. But a number of them involve accusations against senior officers or directors for their alleged role in a bank's problems or even its demise. My involvement in these has led to this conclusion: Banks typically don't get into trouble because a meandering director or officer on his or her own does something totally terminal. It's more because supervisors were too slow to call a halt to problematic practices when violations pile up to the point at which supervisory duties demand intervention.

I worry that regulators' structural failure before the crisis to rein in problem banks before it was too late hasn't been corrected. Despite the very costly lessons of the financial crisis, the risk of supervisory error is even greater this time around. Bankers are basically the same. Supervisors may be a bit more chastened, but the fundamental paradigm of bank supervision is weaker because of the abundance of new regulations banks must follow. With so many rules to follow, a bank can't know which one regulators will care most about in an exam or which transgressions could turn terminal without rapid intervention.

How then do we ensure that supervisors take prompt corrective action on material lapses in time to prevent risk to the FDIC or, worse, to the financial system?

The federal bank regulators have taken steps since the crisis to improve their supervisory acumen, but more needs to be done. A good first step: Let's make public the numerical grades banks get on examinations, known as Camels ratings. This would make regulators subject to the market discipline they demand banks endure. (In the Camels system, banks receive a score in each category — as well as a combined composite score — for their capital, assets, management, earnings, liquidity and sensitivity to market risk.)

Disclosing Camels could have revealed the actual failing grades for troubled banks with time remaining to turn those institutions around. Focusing on Camels would also offset the current maniacal focus on capital as the trigger for an institution's troubles. Banks fail for lots of reasons other than capital, and Camels is supposed to catch those other triggers as well.

Before the financial system's latest near-death experience, there were three clear systemic crises. One of them — the 1998 Long-Term Capital Management fiasco — arose from what we now call a shadow bank and thus does not apply to lessons learned about bank supervision. The other two crises were the truly catastrophic savings and loan disaster of the 1980s and the follow-up crisis related to real estate finance in the early 1990s. By 2001, many had wiped their brows and thought the worst was behind them due to a new regulatory rulebook, much like many of us are now tempted to do.

In 2001, though, we received a vital reminder of what could still go very wrong. One of the nation's largest surviving savings associations — Superior F.S.B. — bit the dust. I testified then before Congress as to the causes of so large a miss after so many solutions had been enacted. I pointed then in part to the remaining ability of U.S. regulators to keep their own assessments — including that of Superior — secret. Had the Office of Thrift Supervision been required to disclose its Camels for Superior, the troubled bank's cover could have been blown long before it became one of the largest thrift failures preceding the 2008 crisis.

Had the OTS been forced to tell us what it should have known about Washington Mutual and Countrywide in 2007, it would have been subjected to widespread ridicule that would have forced it to intervene and, even at that late date, perhaps avert billions in losses across the financial system.

In 2001, when I called for Camels disclosures, bank regulators were joined by community bankers in close alliance against Camels disclosures. The regulators believed that public disclosure would compromise their ability to talk straight to troubled banks and community bankers. They worried that any disclosures about such straight talking would be painful, if not terminal, for the bank. I countered with reminders about ways some of the straight-talking could be muted to give banks time to clean up without giving regulators too much cover. But Camels remained secret, as it continues to be today.

In recent expert-witness work, I've read several inspector general reports on what the federal banking agencies, including the OTS, did and did not do from 2000 to 2009 for now-failed banks. The reason I've come back to Camels is because the IG reports are so startling in how many bright blinking lights each of these federal regulators either didn't see or didn't think much about. For good measure, state supervisors were no better. Reading the examination reports that underlie the post-mortems is even more depressing because the magnitude of the often simple control failures that went undisciplined is still more evident.

Camels disclosures would not mean that all of the assumptions and excuses that blinded examiners would come to light. Exam reports should stay secret. But releasing the aggregate Camels number will permit investors and analysts to spot disconnects between what they see as they survey a bank and what regulators should remedy or may be missing altogether.

Because Camels tracks critical issues like liquidity and interest-rate risk, disclosures will also provide a useful cross-check on undue reliance on capital adequacy. Strong capital can oversell a bank's health. All too many banks were well-capitalized just before their other weaknesses killed them. Prompt corrective action today is so hard-wired to capital that other causes of bank weakness — well known though they are — are often overlooked in disciplinary action until it's too late.

The hard-wired focus on capital is problematic not only for individual banks, but also for the financial system as a whole. We don't know Camels ratings because they're confidential, but I strongly suspect they are procyclical. They give ratings for past problems and so are lagging indicators. Currently, the scores reflect problems so late in the process it's as if they're merely meant to show regulators are aware of a bank's distress, but that doesn't help ready the bank for emerging risk. If regulators won't or can't make Camels public, they can and should study this question to be sure Camels work as they should and advance the Federal Reserve's countercyclical objectives. If the agencies don't do this, the Government Accountability Office should.

Bankers are of course responsible for the banks they run, but banks are unique institutions with public benefits over which supervisors must stand ever vigilant. Without reforms to current procedures that increase supervisory transparency, agency accountability and interagency coordination, we run the real risk of yet another systemic-risk catastrophe, not to mention unnecessary bank failures along the way that harm investors, depositors and their communities.

Karen Shaw Petrou is managing partner of Federal Financial Analytics.