Are State Regulators Too Lax?

A new study analyzing data from 15 years of state bank examinations delivers some explosive conclusions certain to add fuel to the already combustible relationship between federal and state regulators.

The top line: federal regulators downgraded banks' supervisory ratings twice as frequently as their state counterparts.

What's more, after federal exams, state banks reported higher nonperforming loans, higher levels of regulatory capital and lower returns on assets.

It gets worse.

States with more lenient supervision relative to federal counterparts had higher bank failure rates and more banks on the FDIC's problem list. Why? At least in part, the researchers blame "regulatory capture," saying state banking departments may cater to the banks in their state because they rely on the fees they pay.

But interviews with the lead author, California's banking commissioner and a Federal Reserve Bank president present a more nuanced view of what's going in state bank supervision.

First, the lead researcher says he's not out to torpedo the dual banking system. His goal is simply to highlight the inconsistency of oversight, explore its costs and quantify what benefits might outweigh those costs.

"This paper is not saying that dual banking is bad," says Amit Seru, an associate professor of finance at the University of Chicago's Booth School of Business. "What the paper clearly establishes is that you have two regulators who look at … the same bank virtually at the same time and are coming up with very different conclusions.

"That's worth having a public debate on."

Seru readily concedes he does not have conclusive evidence that state examiners go easier on their banks to protect their budgets. "I am not going to stand here and claim that we have conclusive evidence on this," Seru says.

California's commissioner of financial institutions, Bill Haraf, objects to pretty much everything about the research, from who did it to how it was conducted to the conclusions it draws. Kansas City Fed President Esther George says the researchers do not understand the cooperative nature of the examination process.

But before a detailing of their objections, a little background.

Oversight of the roughly 5,000 state-chartered banks in this country is shared by the state in which the bank is chartered and either the Federal Deposit Insurance Corp. or the Federal Reserve.

As of Sept. 30, the FDIC was examining 4,193 state-chartered banks while the Fed was examining 826 banks. (There are 1,333 nationally chartered banks, but they were not included because their exams are conducted by a single agency, the Office of the Comptroller of the Currency, so there is no inconsistency to study.) This paper, "Inconsistent Regulators: Evidence from Banking," covered exams of state banks with assets of less than $10 billion and Camels ratings of 3 or better over a 15-year period, 1996 through 2010. (Seru says the results hold even when data for banks with Camels ratings of 4 and 5 is included.) State banks are examined on a 12-month or 18-month cycle, depending on their size and condition. States alternate exams with the federal agencies. So a state bank examined by its state this month would see its federal examiner sometime between early and midyear 2013.

Conventional wisdom is that this rotation delivers the best of what both the states and the federal agencies have to offer. The states provide better knowledge of local conditions and a nearby door to knock on when a bank has a problem while the federal regulators bring a broader perspective and an ability to compare banks across a wider spectrum.

The study provides tons of statistics, but here are the key ones. For state banks overseen by the Fed, 76% of the downgrades were executed by the Fed and only 24% by the state regulator. For state banks under the FDIC, 65% of the downgrades were executed by the FDIC and only 35% by the state regulator.

While state regulators object to some of the data itself, one of their biggest concerns is who conducted it.

Of the four authors, two are academics. A former colleague of Seru's at the University of Chicago, Francesco Trebbi of the University of British Columbia, is the other academic.

But the two other co-authors both work in the Fed system. One is Sumit Agarwal, a senior financial economist at the Federal Reserve Bank of Chicago, and the other is David Lucca, an economist at the Federal Reserve Bank of New York.

"There is a bias there," Haraf says. "It would have been good for them to have consulted with us. I think we could have straightened out the records on some of their assertions and claims."

The Conference of State Bank Supervisors has invited the authors to present their findings to its members in May.

But because Camels ratings data is confidential, Seru could not have done this research without collaborators with access to it. "The data we got is proprietary," he says. "The analysis is done confidentially inside [the Fed] and then we get the regressions." Seru insists the Fed "didn't influence anything we wrote in this paper."

Haraf also notes that conditions can change during the time that elapses after a state completes an exam and a federal regulator comes into a bank.

"Conditions change in a 12- to 18-month period. Examination doesn't happen in a vacuum."

Seru doesn't buy that. "I wouldn't concede much to them, because when you are looking at a horizon of 15 years … this is not a one-off fact," he says. "It's not clear why it should systematically show in only one direction."

What worries Haraf most is this research turning into ammunition.

"This data can be used by people with a political agenda in favor of consolidated supervision," he says.

Early versions of what became the Dodd-Frank Act called for rolling all bank supervision into a single regulator. "You can imagine if that issue ever gets reopened, this type of research could be used in a political context," Haraf says.

George of the Kansas City Fed is more concerned about the study's impact on the collaborative relationship between federal and state supervisors. She says the alternating exam schedule isn't as black-and-white as the researchers assume.

"The distinction of the alternating exams really comes down to who is signing the exam report, whose letterhead is it on and who will have the discussion with the banker," she says.

"We don't supervise these institutions in isolation, where the state goes in and then goes back to their office and then we go in," she says. "There is an ongoing dialogue between the Fed and the states and the FDIC on how we calibrate our supervision."

In an interview, Seru admitted that it would be impossible to eliminate inconsistency — even examiners within a single agency will see things differently. "But I come back to how much of this inconsistency can we live with? All these entities are different, but I don't think we spend enough time thinking about the downsides of inconsistency."

The downsides he sees are delays in corrective action being taken, weaker banks and ultimately more failures. But he's willing to concede that the states may offer benefits that should be quantified and then measured against the costs. "Maybe the answer in the end is that we understand these costs of a dual system. There is inconsistency, but we are still happy to do this because we think the benefits outweigh the costs," he says. "That's fine, but I think we should have this debate."

"We need to sit down and figure out what those benefits are. What are the magnitudes of those benefits?"

As regulators struggle to implement the Dodd-Frank Act, we are at a "huge moment in banking history," Seru says. The agencies involved are expanding with the addition of the Consumer Financial Protection Bureau and the renewed interest in financial products at the Federal Trade Commission.

"How are we making sure that, as we expand the system, there won't be more and more of these coordination and inconsistency problems?" Seru asks. "We need to take this forward and try to understand what the benefits and costs of both sides are."