Why Big Banks Cut It Close in Fed's Stress Tests

WASHINGTON — On paper at least, it wasn't a stellar showing for the major banks in the results of the Federal Reserve Board's latest stress tests released Wednesday.

Two banks failed — foreign-based Santander and Deutsche Bank — while Bank of America was faulted for weaknesses in its capital planning process and JPMorgan Chase, Morgan Stanley and Goldman Sachs, meanwhile, were forced to revise their capital plans in order to pass the Comprehensive Capital Analysis and Review.

But analysts suggested that in the case of the U.S. banks that rode close to the line in passing the quantitative part of the test may have done so because of a willingness to push the limits of the stress test because they are comfortable that they can clear the central bank's qualitative aspect of the test.

"If the capital itself is not below the required norm but slightly above or equal, it doesn't necessarily mean it's not adequately capitalized," said Alex Tsigutkin, chief executive of regulatory reporting and risk management firm AxiomSL. "It could also mean that the banks are trying to be more capital-efficient."

Jaret Seiberg, an analyst with Guggenheim Securities, said "there is no competition to pass this by a huge margin."

"This is about having the qualitative systems in place so you can make payments to investors," he said. "Some banks on quantitative grounds had to resubmit because their initial distribution request was too high. But that's also a positive sign in that they are willing to seek distributions closer to the line because they are more confident on how they are going to fare on the qualitative part of the test."

Of the major U.S. banks, Bank of America was dinged the most. The Charlotte-based firm received a conditional "non-objection" to its capital plan this year after having to submit an "adjusted" capital management plan a year earlier when its initial plan fell short of regulatory capital minimums. The Fed asked the bank to "correct weaknesses in some elements of its capital planning process" by Sept. 30, and cited "weaknesses in [the bank's] loss and revenue modeling practices and some aspects of the [bank holding company's] internal controls."

But several other large banks had to submit adjusted capital plans after missing one or more of the regulatory capital minimums with their initial capital plans.

JPMorgan Chase failed the Tier 1 leverage ratio with its initial filing with a 3.8% capital level, below the 4% required minimum, and just squeaked by the required minimum Tier 1 common ratio with 5.0% under the severely adverse economic scenario posited by the Fed. Its revised capital plan showed a 4.1% Tier 1 leverage ratio and a 5.5% Tier 1 common ratio.

Morgan Stanley and Goldman Sachs, meanwhile, both failed the Tier 1 risk-based and total risk-based capital ratios with their initial capital plans. Morgan Stanley posted just under the required 6% for Tier 1 risk-based and scored 7.4% for total risk-based capital with its initial capital plan, below the 8% requirements. On its second attempt, it scored 6.2% and 8.2%, respectively.

Goldman Sachs posted 5.9% for Tier 1 risk-based capital, below the 6% requirement, and 7.6% for total-risk based capital with its original plan. It came back with 6.4% and 8.1% marks with its revised plan.

One senior Fed official said that this year is the first where any bank managed to pass the Tier 1 common ratios while failing some of the other regulatory capital minimums. The official said that the fact that three of the biggest banks met that fate is likely a reflection of the nature of this year's severely adverse scenario, which focused on vast corporate defaults. Those kinds of losses disproportionately affected the big banks with large exposures to trading and capital markets.

Mike Alix, a partner at PriceWaterhouseCoopers specializing in financial risk, said that in recent years the plodding advance of Basel III implementation has changed the way that the capital ratios are calculated. The distinction between regulatory and accounting capital are fading and the limitations on what can be considered "capital" for the purpose of calculating the regulatory minimum ratios are growing, Alix said. As a result, a 5% Tier 1 common ratio is increasingly easy to meet, while those other regulatory minimums are more elusive.

"The regulatory capital calculation itself is changing, and the relationship between common and non-common is change, and that is having effects on the results as you go through time," Alix said. "In recent years, the Tier 1 common ratio has been the binding constraint. Because of the changes I described, other ratios are coming in as the constraints."

Chris Maclin, head of finance risk and regulatory reporting at Wolters Kluwer, noted that in the years since the stress tests have been enacted, most of the banks have built up substantially more capital than they held before the crisis. But less progress has been made in developing the level of data collection and transparency that the Fed expects, especially from the largest banks. As a result, banks appear to have a handle on the quantitative aspect of the test, while the qualitative part has become the binding constraint.

"I think the bar will just continue to get higher," Maclin said. "Clearly the effort that the banks have put into passing the qualitative aspect has increased dramatically, but I think the Fed expects that to be an evolving process going forward."

Maclin added that the technological advances that the Fed is pushing banks to adopt are not going to be limited in their application to CCAR. The Fed's liquidity monitoring reports are one example of how the ethos of perfect transparency embodied in the qualitative portion of CCAR has metastasized to other areas of the central bank's regulatory and supervisory structure.

"I think the expectation now from the Fed is, 'What can we take from the CCAR stress test, and are there other measures we can monitor with similar processes?'" Maclin said.

The Dodd-Frank Act requires the Fed to conduct two separate rounds of stress testing on each bank with more than $50 billion in total assets. Both stress tests analyze banks' performance under three hypothetical economic scenarios of differing severity over the course of the next nine quarters. The Fed analyzes whether banks' capital levels dip below the regulatory minimum at any point during that stress event.

The first test, known as the Dodd-Frank Act Stress Test, runs banks' balance sheets through a standardized capital plan in order to make their capital reserves as comparable to one another as possible. Banks don't technically pass or fail the DFAST test, but if a bank falls below the minimum capital threshold in the first test, it is a sign that it may be undercapitalized for a stress event.

Banks can fail CCAR, however, and can fail in more than one way. CCAR examines banks' performance through the scenarios using their own capital plans, providing a more realistic assessment of how the bank would perform in the scenarios. If a bank falls below the minimum capital requirements in CCAR, the bank may resubmit its capital plan in order to stay above those minimums.

Two non-U.S. banks, Deutsche Bank Trust and Santander SA, failed CCAR on qualitative grounds despite having more than enough capital to pass the quantitative assessments. While those firms only represent small portions of the parent companies' global operations, they show that the Fed is concerned that the firm lacks some measure of internal controls to assess and reduce risk.

The CCAR results noted that the Fed had particular concerns with Santander's "governance, internal controls, risk identification and risk management, management information systems, and assumptions and analysis that support the BHC's capital planning processes." The Fed's objections to Deutsche Bank's capital planning processes, by contrast, are reflective of concerns about the bank's "risk-identification, measurement, and aggregation processes; approaches to loss and revenue projection; and internal controls."

Santander has the dubious distinction of having failed the test two years in a row, having received a qualitative objection from the Fed last year (Deutsche Bank was not included in last year's stress test).

There is no specific penalty for failing the test in multiple years — at least no regulatory penalty — except that Santander USA may not pay out a dividend even if the rest of the bank's global operation does pay a dividend. Banks that demonstrate a chronic inability or unwillingness to correct deficient behavior can also be subject to enhanced regulatory actions, including but not limited to cease and desist orders, the senior Fed official said.