Stress Tests: Less Stressful Than Ever

By Marcus Stanley

(Note: see also AFR slide presentation laying out these findings)

The chart below shows that this year the Federal Reserve permitted banks to pay out 190 basis points (1.9 percentage points) of their risk-based capital to shareholders, an unprecedented level compared to previous years.

SOURCE: Federal Reserve and AFR Calculations

At the same time, the chart also shows that the stress tests appear to have been less stressful than ever this year, with the total estimated impact of a severe economic downturn reducing bank risk-based capital levels by just 330 basis points (3.3 percentage points). That’s a level significantly lower than previous years and far lower than estimates of the capital losses suffered by banks during the financial crisis. For example, a 2013 study by the Boston Federal Reserve found that the ten banks with the largest capital losses during the financial crisis lost an average of 6.6 percentage points in capital over that period, with many becoming insolvent. That’s likely a significant underestimate since it doesn’t take into account all the extraordinary assistance received during the financial crisis; a more recent estimate by the Federal Reserve finds even higher financial crisis losses.

There are a couple of different interpretations of these outcomes. The regulator-friendly interpretation is that increased oversight has succeeded in reducing bank risks sufficiently that current capital levels are more than adequate to permit record post-crisis payouts to investors.

That might be true but it’s hard to verify from the outside, particularly since bank capital levels actually don’t seem to have increased very much in recent years. Tier 1 risk-based capital levels were at 13.9 percent going into the 2017 stress test process, as opposed to 13.5 percent going into the 2015 stress tests.

Assumptions about how the largest Wall Street banks will weather a crisis are heavily dependent on one’s assessment of the complex hedging strategies banks use to try to mimimize trading and counterparty losses and how such strategies will perform in a crisis. That’s a level of detail that only bank supervisors have access to, and whether it’s reliable depends on the banks’ own internal risk modeling — perhaps a reason that banks are now lobbying hard to remove the supervisory assessment of bank internal risk management from the approval process for capital distributions to shareholders.

Another interpretation is that the Federal Reserve is already making concessions to big banks in order to head off pressure for policy changes that would weaken or eliminate regulatory authority. The CHOICE Act that recently passed the House would severely impact Federal Reserve regulatory in areas ranging from agency funding to the vulnerability of their key policy choices, including stress testing practices and parameters, to being overturned in industry lawsuits. The Treasury Department’s recent report on bank regulation adopted some 75 percent of recommendations by big bank lobbyists, including a number that would sharply limit the autonomy of the Federal Reserve in areas like stress testing.

Perhaps regulators are already deciding that discretion is the better part of valor in the current environment. If so, that’s not a good signal for their ability to protect the public from financial sector risks.