Bank Exam Ratings May Not Be as Secret as You Think

WASHINGTON — For banks, the supervisory grade known as a "Camels" score is a critical and confidential yardstick of strength. But for many in the public, it is a riddle too tantalizing to ignore.

The ratings, in use since 1979, are typically known just to the bank's primary regulator that decides its score, as well as senior management.

While analysts, consultants and others have long argued for public disclosure of Camels ratings, the government's insistence on secrecy has not stopped interested parties from either mimicking the scoring process with reported information, or using clues to try and guess a bank's official rating.

"There is something of an attitude among bank regulators that the public isn't capable of handling this information for fear of sparking runs," said Jerome Fons, an executive vice president at Kroll Bond Ratings, which uses failed-bank data to score open institutions.

"It's hard to mimic" Camels "because it's not publicly available. But what I can observe is which banks have been closed and which haven't and I can correlate those closures to various financial ratios and metrics. We can reverse engineer Camels. That's the bottom line."

Fons' company is part of a cottage industry of private ratings organizations that use publicly available information to create an alternative to Camels for their clients. While Kroll's method works backwards from the point of failure, other firms try to create a more direct simulation of the Camels experience. Meanwhile, some experts believe certain public data points provide hints about a bank's actual Camels rating.

"If you gave 10 analysts the" public examination report "of a bank, they would all come up probably with the same estimate of a Camels rating," said Rebel Cole, a finance and real estate professor at DePaul University in Chicago and a former Federal Reserve Board economist. "It raises the issue of: Why do they make such a big deal about the secrecy of the Camels rating?"

Camels was first developed in the 1970s as part of the regulators' "Uniform Financial Institutions Rating System." Institutions were judged on five different components under the acronym C-A-M-E-L: capital adequacy, asset quality, management, earnings and liquidity. The system was revised in 1996, when agencies added the 'S' at the end of Camels for "sensitivity to market risk."

Banks receive a score of 1 to 5 for each component, and a final Camels score in that range representing the composite total of the components. As a bank's score gets higher, so too does the regulatory scrutiny. Banks with a 4 or 5 Camels rating are added to the Federal Deposit Insurance Corp.'s "Problem List" — also confidential — of institutions with a greater likelihood of failure.

While each bank's score is a tightly-kept secret, curious parties such as insurers and even government agencies not privy to what the bank regulators know have sought versions of the ratings for their own purposes. For example, the Securities and Exchange Commission, which lacks the official ratings, uses the model offered by Lord Whalen LLC's Institutional Risk Analytics, a private Southern California ratings firm that simulates Camels for their clients.

Insurers that write policies on banks are also interested in the information. In 2005, a joint regulatory guidance reminded the public about the confidentiality of Camels ratings after some insurers had asked banks to disclose them.

"The insurers want the Camels because it gives them a frame of reference in terms of enterprise risk," said Chris Whalen, senior vice president and managing director for IRA. "When you're writing directors and officers' insurance for a bank, op risk is huge. That is what you're insuring."

Yet most agree it is impossible to replicate the official ratings exactly, since the regulators likely include highly subjective information about individual institutions in determining a Camels score. One key element of the rating system is the quality of a bank's management, which is not necessarily quantifiable.

"Some will come up with 'Caels' — that's basically Camels without the 'M'," said Bert Ely, an independent consultant based in Alexandria, Va.

Nicholas Ketcha, a consultant and former FDIC director of supervision, said while the agencies are not opposed to firms doing shadow ratings, they view the official rating as purely a supervisory device, not a means for transparency about safety and soundness.

"The regulators' main objection to it always has been is, 'These are supervisory ratings. We use them for determining what kind of remedial actions we want to put the bank under. It's our way of monitoring, and it's not meant to be public,'" said Ketcha, who is executive managing director at FinPro Inc. in Liberty Corner, N.J.

Kroll's model uses data from the wave of failures over the past three years. The firm identifies nine financial ratios it says are most indicative of the chances a bank will be closed, such as brokered deposits to total deposits, nonperforming assets to total equity and the percent change in core deposits from the previous quarter. The ratios are then run through a statistical model that produces a numeric score. The score is then mapped to a letter grade of A through E (with A being best).

"What we try to do is predict bank failures, rather than necessarily mimic the Camels system," Fons said.

He added, however, that any unofficial rating lacks a key element: the privileged information regulators keep, which may be the most direct measure of a bank's likelihood of failure.

"Someone looking for a better system of predicting bank failures, it's going to be difficult to come up with something because of the role of the Camels rating in determining failure status," Fons said. "If the Camels rating says you are 5, they're highly likely to close you. … Whether or not the Camels rating is the right or best model, it will be by definition the only thing that matters. It is what supervisors are using to determine the viability of the bank."

Still, there are other firms that try to more closely parallel the Camels process.

Using call-report data for each depository institution, IRA produces its own component rating of 1 to 5 for each of the six Camels elements, and in turn a combined Camels score. Each component has its own methodology. For example, to gauge capital adequacy, a bank's various capital ratios — such as Tier 1 capital to risk-weighted assets — are judged against standard market benchmarks. Each time a bank's ratio is below the standard, that counts against its score for the component.

"We have found that using public information can give you a sufficient feel to understand what is going on in those areas of interest that are in the examination manual," said Dennis Santiago, chief executive officer and managing director for IRA.

While ratings firms cannot get verification from the regulators about whether their scores are accurate, Whalen said he has been told that another IRA tool — an independent estimate of banks' deposit insurance premiums, which take into account similar regulatory factors as in Camels — is "very close."

"That tells me that our Camels are in the right ballpark," he said.

Still, knowing a real Camels score is the Holy Grail. Rather than try to simulate Camels ratings, some believe clues to a bank's actual rating exist in the public sphere.

Drew Dahl, a professor at Utah State University, said an FDIC rule on public auditing for depository institutions provides a hint in limited cases. Banks with more than $5 billion of assets and Camels ratings of 1 or 2 can rely on its holding company reporting for complying with certain requirements of the rule. Dahl said an institution that uses that benefit — which would be reflected in public documents — is likely well-rated.

Yet there are other variables to consider. The relief is granted only in cases where three-quarters of the holding company's assets are in its bank subsidiaries. And the rule provides little insight about which institutions are poorly-rated, since a 1- or 2-rated institution can still choose to report at both the bank and holding company level.

"In other words, if a bank reports at the bank level, it may be highly rated, but if it reports at the holding company level, it must be highly rated," Dahl said.

Others point to the FDIC's "Problem List" as a potential resource. Publicly, the FDIC never cites which banks are on the list, but it does give their combined assets. Observers say if you know the size of a troubled institution you can try to guess whether it is on the list by how much the combined asset total changes each quarter.

"You can probably guess who most of the banks are on the problem list," said Ely.

Overall, Ely said, "I suspect you can get within one notch" of an institution's Camels rating. "You can probably say, 'This bank is probably a 1 or a 2, or probably a 2 or a 3.' You can certainly differentiate a 1 from a 5, or a 2 from a 4."

But, ultimately, he added, guesses about an institution's Camels rating are just that: guesses.

"As someone who has worked with call-report data for 25 years, I have a certain appreciation for what you can't do with call-report data. The Camels rating is based on more than just call-report data. You're limited in that regard."