First take: DFAST stress testing

March 21, 2014

Ten key points from the Federal Reserve's 2014 Dodd-Frank Act Stress Test (DFAST)

“The average person who relies on the banking system should be happy. Next week will tell us whether banks and bank shareholders will be happy.”

The DFAST results published yesterday (March 20th) are the Federal Reserve’s first stress test results released in 2014. Following this release, about half of banks disclosed their own company-run stress test results (required by March 31st). On March 26th, the Federal Reserve (Fed) will release its 2014 Comprehensive Capital Analysis and Review (CCAR) results.[1]

  1. Improved industry capitalization: The DFAST results show that the 30 largest US BHCs have an average Tier 1 Common ratio of 11.5% which is an increase from 11.1% in 2013.  Moreover, despite the more stringent stress test in 2014, the minimum Tier 1 Common ratio for the industry remained resilient at 7.6%, up from 7.4% in last year’s stress test under the severely adverse scenario. These results suggest improved asset quality.

  2. All but one of the 12 new BHCs fared well: Only one of the 12 new[2] BHCs breached the minimum Tier 1 Common ratio of 5%. In comparison, none of the original 18 BHCs did, and the original BHCs fared better with lower loss rates and a higher ratio of PPNR to average assets (3.0% vs. 2.1%).

  3. Stressed Tier 1 Common ratios are lower at about half of the very largest banks – likely the impact of new Fed projections of assets, loans, and RWAs: The Fed utilized its own projections of BHC balance sheets and RWAs for the first time in 2014, in order to enhance the comparability of results across BHCs and to ensure that industry-wide loan and asset growth projections are consistent with prior recessions (i.e., assets and loans grow rather than decline). This change in modeling approach, however, contributed to an increase in RWAs and to greater erosion in stressed capital ratios that we estimate to be 50-100 bps. This new element of the Fed’s “black box” proved to be the most significant and is a tool the Fed may leverage to keep capital in the system.

  4. Higher analyst payout expectations have likely caused a few banks to overshoot in their capital plans: Because the results display somewhat greater Tier 1 Common dilution from last year (390bps versus 370bps), banks have roughly the same capital flexibility. Against a backdrop of significantly greater payout expectations by equity analysts, we believe that some bank capital plans have tested the Fed’s boundaries resulting in a few banks having to quickly adjust their proposed capital actions (these banks know who they are and are currently making adjustments that will be disclosed as part of next week’s CCAR results). Even those few banks with significant capital flexibility may be restrained from greatly increasing dividends due to relatively high payouts in prior years – regulators still look unfavorably on dividends versus buybacks given the difficulty in ratcheting down dividends in future years.

  5. Uncertainty remains around ultimate capital plan approval on March 26th: The Fed’s expectations for this year are heightened with respect to the sophistication and comprehensiveness of the capital planning process. While the 12 new BHCs will likely not be graded on the same curve as the original BHCs, there may be a regulatory desire to make a statement and ensure that the newcomers understand the significance of the stress test mandate even if they do not view themselves as truly systemically important. Therefore, the Fed will likely qualitatively object to, or require resubmission of, the capital plans for one or two of these new entrants in addition to one or two of the original banks. See PwC’s Regulatory Brief, Stress testing: Failures on the horizon? (November 2013).

  6. Counterparty credit shock has muted impact:  The Fed included a counterparty shock component for the first time this year that applies the Fed’s global market shock parameters to counterparty positions and assumes that a BHC’s largest single counterparty defaults (90% loss given default).[3] However, the counterparty and trading losses interestingly remain flat at $98 billion. This may be attributable to the exclusion of designated central clearing firms as a possible defaulting counterparty, even as greater volumes are being directed toward central clearing, as well as the exclusion of G-7 sovereigns.

  7. Adverse scenario’s rising interest rate impact also minimal: The adverse scenario (disclosed for the first time this year) features rising interest rates and a steepening yield curve, highlighting BHCs’ interest rate sensitivity. This analysis is timely given current Fed tapering and recent speculation regarding the end of the zero interest rate policy. This scenario shows an increase in unrealized losses on Available-for-Sale (AFS) securities of $79 billion for the 14 Advanced Approaches banks (as compared to the severely adverse scenario) that must be recognized in the AOCI component of capital.  However, the adverse scenario also demonstrates an expansion of net interest margins in a rising and steepening rate environment resulting in an increase in PPNR for these banks of $106 billion, fully offsetting the unrealized AFS losses over time.

  8. Basel III standardized projected capital ratios show surprising strength:  For the first time, BHCs incorporated the Basel III standardized approach into their stress test projections (for 2015 quarters), which is a more stringent framework than Basel I. At first blush, it is surprising that for many of the banks the trough values for the more conservative Common Equity Tier 1 (Basel III) ratio are higher than the corresponding Tier 1 Common (Basel I) ratio. This seemingly counterintuitive result is likely explained by the difference in timing of the troughs of the two ratios, with the Tier 1 Common low point occurring early in the planning horizon before banks can rebuild capital through earnings retention.

  9. Loan loss rates improve from last year: Portfolio loan loss rates improved 60 bps overall, from 7.5% to 6.9%.  In a portfolio-by-portfolio comparison, credit card portfolios improved the most, by 1.5% (from 16.7% to 15.2%), while commercial and industrial portfolios improved by 1.4% (from 6.8% to 5.4%).  The commercial real estate portfolio was the only asset class with increased loss rates, which increased by 40bps to 8.4%.

  10. Leverage Ratio is less of a binding constraint:  Despite speculation that the Leverage Ratio could be the binding constraint for the industry, improvement in the overall capitalization levels of the 30 banks has resulted in the ratio being less of a binding constraint than in previous years.  The Leverage Ratio (stressed minimum) improved slightly from 5.9% to 6.1% and is well above the 4% minimum level for BHCs. 

[1] The primary differences between DFAST and CCAR is that CCAR includes the BHCs’ submission of proposed capital actions and the Fed’s object or non-object determination. Capital actions under DFAST are standardized based on the BHC’s existing dividend payout rate.

[2] The 12 new DFAST and CCAR participants this year are BMO Financial, BBVA Compass Bancshares, Comerica, Discover Financial Services, HSBC North America Holdings, Huntington Bancshares, M&T Bank, Northern Trust, RBS Citizens Financial Group, Santander Holdings, USA, UnionBanCal, and Zions Bancorp.

[3] The eight BHCs required to perform the counterparty default scenario include the six BHCs that have been subject to the global market shock (Bank of America, Citigroup, Goldman Sachs, JP Morgan, Morgan Stanley, and Wells Fargo), and the two largest custodial banks (Bank of New York Mellon and State Street).

Related reading