WASHINGTON — Top Republicans are moving beyond seeking to just revamp the Dodd-Frank Act and have set their sights on a different facet of the post-crisis regulatory environment: stress tests.
They are weighing potentially broad changes to the stress test program, particularly the Federal Reserve's Comprehensive Capital Analysis and Review, citing it as an example of regulatory overreach and floating a range of potential alternatives to the way the tests are performed and the consequences of failing them.
Their moves put them on a likely collision course with the central bank, which sees the stress tests as vital to the safety of the financial system.
"They are the single most important supervisory innovation, not just since the crisis, but really for the last 20, 25 years,” Fed Gov. Daniel Tarullo, the principal architect of the stress tests who is preparing to step down in April, said last week.
The Fed administers two different stress tests. The Dodd-Frank Act Stress Test, or DFAST, is required by the 2010 financial reform law contained in its name, while CCAR is an exercise created by the Fed without congressional input.
Together, the tests subject a large majority — some 80% — of U.S. assets under management to examination and provide a dynamic, rather than static, examination of not only the largest banks’ capital plans, but also the smaller regional banks whose continued function is critical to the economy in rebounding from a crisis. While there is no consequence of failure to DFAST, CCAR is another matter. Banks that don't pass CCAR for either quantitative or qualitative grounds can find their capital plans rejected or delayed.
But a draft version of a revamped version of House Financial Services Committee chairman Jeb Hensarling’s Choice Act would make extensive changes to the stress tests, including limiting CCAR to a biannual examination rather than an annual one, curtailing the applicability of the qualitative CCAR tests and limiting the Fed’s ability to stop banks from issuing dividends if they fail to pass CCAR.
Some Republicans want to go even further. Sen. Pat Toomey, R-Pa., said last week that he has "big problems" with CCAR. Those included "correlating" the effect of bank risk and capital allocation and the application of an implicit system of risk weighting and instituting models that the Government Accountability Office said in a report last fall were opaque for a program that has no statutory mandate.
“Isn't CCAR at least somewhat duplicative?” Toomey asked Fed Chair Janet Yellen during her semiannual monetary policy hearing last week. “And since it is very, very costly and not mandated by statute, would you consider bringing it to an end at some point in the foreseeable future?”
But the Fed appears wedded to the program. Yellen called it a "key part of our regulatory process” and one that is appropriately tailored to the largest banks and to prepare the financial system for a wide variety of potential systemic risks. The GAO report to which Toomey referred, while making suggestions for how to improve stress tests, also lauded them as a valuable innovation.
“It's been a cornerstone of our efforts to improve supervision, especially of the largest banking institutions that — whose stability is really critical to overall U.S. financial stability,” Yellen said. “The GAO in their assessment found that the stress tests have been useful and have played a useful role. They did not recommend that we end them.”
Wayne Abernathy, executive vice president for financial institutions policy and regulatory affairs at the American Bankers Association, said that, while banks don't necessarily oppose stress testing or CCAR per se, the public spectacle element of the exercise is probably a vestige of another era. The processes embodied in CCAR could easily be folded into the Fed's ongoing Camels supervisory process without compromising any of the effectiveness of CCAR, and Abernathy said he suspects that was what Toomey had in mind.
"It shouldn’t be this pass/fail sort of thing, because that doesn’t provide the kind of subtleties of supervision that you apply to … all the other elements of Camels," Abernathy said. "Toomey seems to be trying to get rid of that and making it more part of the regular supervisory process."
But there are already signs of opposition. Sen. Sherrod Brown of Ohio, the top Democrat on the Senate Banking Committee, told American Banker that he is pleased with the Fed's continuing efforts to tailor CCAR to focus on the riskiest banks but that the program's integrity should not be undermined by Congress.
“The Fed's recent update to its stress test requirements is another example of how the Fed is appropriately tailoring its rules for banks of different sizes and scales," Brown said. "But I am concerned by proposals that would eliminate an important tool that the Fed has used to ensure that big banks have enough capital to survive a crisis without needing a taxpayer bailout, simply to make it easier for banks to enrich their shareholders.”
Stress testing as a concept has been around for decades, generally as an internal process for financial institutions to wargame various economic contingencies on their portfolios. The idea is that one assumes certain conditions — a dropping GDP, illiquid markets for certain securities, etc. — and examines how those conditions will affect one’s balance sheet over a period of time. The current stress testing widow is nine consecutive quarters.
But the so-called top-down stress testing program currently in place was a product of the financial crisis. In early 2009 the Fed and other financial regulators implemented the Supervisory Capital Assessment Program, assessing the capital resiliency of 19 of the largest banks, finding that nine had no excess capital.
Dodd-Frank codified the need for an annual test, and 19 banks were subject to DFAST and CCAR when the program was run for the first time, in 2011. In the 2017 tests, 34 banks are slated to be tested.
DFAST and CCAR both use hypothetical scenarios to determine whether a bank’s capital would fall below minimum requirements in a stress environment. The tests differ in several ways, but the biggest difference is that DFAST runs each one fundamental way: DFAST tests a bank’s balance sheet against a standardized capital management plan, while CCAR tests the banks’ own capital management plan.
The Fed itself has long acknowledged that the stress testing program is imperfect and must be continually improved, and Tarullo in a speech last September unveiled his recommendations for how to make the program more dynamic and tailored to each bank’s capital reserves. Additionally, in January the central bank announced that it was eliminating the qualitative tests for all but the largest and most complex banks — effectively exempting 21 of the 34 banks from that more nebulous aspect of the examination.
Those changes have garnered fairly broad support from the industry and lawmakers in both parties. But banks’ apparent appetite for even broader changes appears not to have been sated.
The Clearing House Association in January issued a research note that sought to reverse-engineer the Fed’s stress tests — whose capital performance models have been described but not fully disclosed by the agency — to determine how the Fed expects certain asset classes to perform in a crisis. The note suggested that the Fed’s models appear to assume much poorer performance by small-business loans and residential mortgages in a crisis, resulting in a sort of de facto risk weighting of those asset classes.
When asked about the research note during her testimony before the House Financial Services Committee earlier this month, Yellen described it as a “highly flawed study” and said that she disagreed with its findings. She said it assumed an average rate of delinquency rather than a more granular examination of how loans actually perform.
But the program’s opponents will likely continue to cite the study as evidence that the process stifles lending and economic growth.
It is surprising that, after years of complaining about programs like living wills and the Consumer Financial Protection Bureau, the banking industry and Republicans have set their sights on CCAR.
Abernathy said in an interview last month that unlike some other ideas to come out of Dodd-Frank and the Basel accords, the stress tests actually do what they set out to do. But he added that the program could be improved further if it was simply folded into the existing supervisory program rather than being a separate annual "circus" where some banks pass and others fail.
“Of all the complex stuff that’s been thrown at the banking industry, the stress tests have been the most successful, and [are] the ones that are consistently getting better — far better than the living-wills process, and far better than a lot of the Basel III stuff,” Abernathy said. “I think where bankers get frustrated is where they have to go through a lot of complexity, where at the end … they wonder, ‘What good did that do?’ ”
Justin Bakst, director of Capital Markets at CoStar Group, a data firm that specializes in commercial real estate and has worked with many of the largest banks in their stress tests, said the costs associated with CCAR are high, but the banks get something for their money — namely, insight into their businesses’ operations that helps make better strategic decisions.
“If the primary focus of the CCAR scenario is to go by the letter of law, provide the regulators with what’s needed to perform these stress tests, and don’t leverage the data in any other components of the business, this is an expensive proposition,” Bakst said. “What many of the institutions have done — if not most of them — is they are taking this information and utilizing it to make better strategic decisions.”
Marcus Stanley, policy director for Americans for Financial Reform, said the fact that so much of the revamped Choice Act’s proposals related to CCAR deal with the Fed’s ability to limit dividend payments tips the industry’s hand on what is driving the push to reform stress tests.
Restricted dividend payments mean less money that flows to shareholders — typically large institutional investors.
“There’s people on one side of the wall who want a bundle of cash in their hands, and there’s regulators on the other side of the wall who are not handing them that cash,” Stanley said. “It’s not about process, it’s not about models … this is a straight-up money thing.”
Stanley said he doubts that Congress will curb CCAR legislatively, because, as Toomey pointed out, there is no statutory basis for CCAR. The Fed could make any or all of the changes that lawmakers seek, including discarding CCAR altogether, without any input from Congress. That theoretically gives the Trump administration a chance to take charge of the process.
“If you just appoint a set of regulators who want to get rid of it, you can get rid of it — you don’t need legislation,” Stanley said.
And Abernathy said the legislative proposals could be a feint designed more to influence the Fed than to handcuff it to any particular legislative vision. The Fed may be moving in that direction on its own.
"Often the regulators, in order to do the kinds of reforms they want to do, don’t mind, at least not privately, getting a nudge from Congress," Abernathy said. "I don’t know if that’s what is going on … but I wouldn’t be surprised if the regulators say, Where you’re trying to push us to go is kind of where we want to go anyway."
George Madison, a partner with Sidley Austin and former policy adviser to then-Treasury Secretary Tim Geithner, said that a wholesale dismemberment of the stress testing regime is a fairly remote possibility, even with Trump’s picks ascending to the Fed board. Stress testing is now part of the “regulatory DNA” at the Fed and elsewhere in the financial system, and killing CCAR isn’t going to change the need for bank supervision, of which CCAR is a critical part, he said.
“The president has a couple seats on the board of governors to put in place, and I do think over time there will be more, and who knows what kind of people he’s going to put in place,” Madison said. “But typically they are economists, and they are trained like every other economist, regardless of their political stripe. They know the importance of the Fed’s ability to manage the regulatory system with respect to banks … and to make sure they are not going to go under.”
And the Fed has broad authorities to sanction banks in various ways, according to Mike Alix, financial services consulting risk leader with PwC. These include curbing dividend payments, downgrading ratings and subjecting banks to heightened scrutiny. If anything, doing away with CCAR might create more trouble for banks because the application of penalties may be less predictable.
“They still have the power to use their examination resources to test whether firms are kind of qualitatively opaque,” Alix said. “And if not, they have other means by which they can create problems to the firms. The objection hasn’t gone away, the assessment hasn’t gone away, and the consequences for having weak processes will still be pretty significant for the institutions.”
Stanley said the coming fight over CCAR is likely to be among the most consequential aspects of Dodd-Frank reform, even if other programs get more headlines.
“This is what the Fed has chosen as its form of 21st-century consolidated bank supervision,” Stanley said. “In terms of just systemic stability and regulation of systemic risk, this is very, very important — pretty much as important as it gets, even though it’s flown under the radar.”