Unpacking the Federal Reserve’s COVID-19 Sensitivity Analysis

The analysis was severe; the Fed should avoid punitive assumptions in the next round of submissions to promote economic recovery

Key Takeaways:

  • The 2020 DFAST contained severe macroeconomic scenario assumptions, even for current conditions.
  • The Fed’s COVID-19 sensitivity analysis (“COVID-19 analysis”) added even more stringent assumptions. Notably, and contrary to much commentary, it was also broader than just adjustments to GDP, unemployment and U.S. Treasury rates.
  • The sensitivity analysis failed to consider any of the benefits of the government’s $6 trillion in fiscal and monetary actions to bolster the resilience of small businesses and consumers.
  • Despite its severity, the sensitivity analysis results demonstrated the fortitude of the banking sector in the face of unprecedented stresses.
  • The Fed should avoid the imposition of additional procyclical shocks on banks as they design the next round of COVID-19 sensitivity analyses, due to be released in September.

Federal Reserve


In mid-May, the Federal Reserve (“the Fed”) announced that it would re-evaluate “the severely adverse scenario” that it had included in the its 2020 supervisory stress tests to estimate the potential economic impact of government imposed lockdowns arising from the COVID-19 event. On June 25th, the Fed published their “Assessment of Bank Capital during the Recent Coronavirus Event” at the same time as the results of its 2020 Dodd-Frank Act Stress Tests (“DFAST”) and the related Comprehensive Capital Analysis and Review (“CCAR”). The Fed used the COVID-19 analysis to support its decision to suspend share repurchases, place caps on dividends, and require banks to resubmit capital plans and conduct additional stress analyses later in the year. This is the first time the re-submission mechanism has been triggered since the introduction of the stress testing process.

The Sensitivity Analysis Made Major Changes to the 9-Quarter Scenarios and Assumptions

Despite a prevailing market narrative that the Fed’s COVID-19 analysis made only changes to three macro scenario inputs in the 2020 stress test, the changes were broader and more material. As background, the Fed’s COVID-19 analysis incorporated three plausible alternative downside scenarios including:

  • A “V-shaped” rapid recovery that regains much of the output and employment lost by the end of this year;
  • A “U-shaped” recovery that is more prolonged, with only a small share of lost output and employment is regained in 2020;
  • And a “W-shaped” double dip recession with a short-lived recovery followed by a severe drop in activity later this year due to a second wave of COVID-19 outbreaks.

As mentioned, all three of these scenarios incorporated sizeable changes to the GDP, unemployment and Treasury rates assumptions. Importantly however, the Fed also made significant changes to:

  • The assumptions regarding the U.S. Volatility Index (“the Vix”);
  • The level of market risk and risk weighted assets;
  • Some of the Global Market Shock (“GMS”) factors;
  • And credit-based inputs and rating assumptions for specific portfolios.

These additional alterations served to increase the balance sheet and risk weighted assets and inflate potential loss estimates for certain exposures under both the 9-quarter planning horizon and the instantaneous GMS component. Moreover, these changes may have more materially amplified losses and post-stress capital requirements than what was implied by the changes to the unemployment, GDP and U.S. Treasury rates inputs. Importantly, the Fed’s approach to the COVID-19 analysis did not factor in the role the government’s $6 trillion fiscal and monetary spending have played in mitigating the worst impacts of the COVID-19 event on consumer and small business credit exposures.

The assumptions the Fed included in their COVID-19 analysis regarding the macro scenarios, capital markets, balance sheet and credit exposures are charted below. Where possible, we have also provided the assumption included in the DFAST 2020 scenario and a description of how those assumptions were changed in the COVID-19 analysis.

The COVID 19 Macro Scenario Assumptions

Clearly the changes to the unemployment and GDP scenarios are a dramatic departure from the assumptions included in the 2020 DFAST, though the impact of a government imposed shutdown of the economy did have unprecedented (if albeit somewhat predictable) impacts on those assumptions. That said, the assumptions previously included in the 2020 DFAST and remained unchanged in the “V”, “U” and “W” shaped recovery scenarios were already draconian and included a 50% decline in equity prices, a 660 bp widening in BBB/Baa2 rated  credit and a 28% decline in the HPI. The Fed’s approach also assumed  a peak VIX of  82.7%, however it is unclear from the analysis the duration or  shape of the VIX adjustment.

Capital Markets Assumptions

The COVID-19 analysis applied a 30% increase in market risk weighted assets (“mRWA”) which is a significant adjustment not only because of its sheer size but because the stress test assumes a static balance sheet over the course of the projection path. While it is true that mRWA assets jumped in the later part of the first quarter, this was largely reported to be a temporary consequence of banking organizations with large capital markets presences stepping in to support market liquidity. A review of some banking organizations initial earnings reports for the 2Q20 suggests the increase in mRWA has abated and trended downwards towards pre-COVID mRWA levels.

The Fed indicated they adjusted select GMS components in their COVID-19 analysis, however the magnitude of those alternations was not disclosed. Those adjustments below are summarized below.

Balance Sheet and Credit Based Inputs and Rating Assumptions

Similar to the change in mRWAs  assumptions, the Fed’s COVID-19 analysis assumed larger balance sheets and risk weighted assets (“RWA”) than firms had assumed in their 2020 DFAST submissions. The increase in balance sheet and RWA has an amplified impact given DFAST only assumes a static balance sheet throughout the projection period. There was certainly a jump in RWAs in late March,  driven by both the increased short-term investments caused by a significant spike in operational deposits and the considerable drawdown on wholesale committed lines of credit. However, an initial review of second quarter earnings reports suggest that end of period balance sheets and RWAs  have return to more normalized levels and below what was reported for many firms at the end of the first quarter.

The Fed’s  COVID-19 analysis also incorporated significant adjustments to corporate credit, both in terms of direct adjustments to credit ratings and to inputs to loss models for specific industries. Such blanket moves, without a full consideration of all the factors that drive credit ratings, is extremely stringent.


The COVID-19 sensitivity analysis results demonstrated the banking sector was extraordinarily resilient even when faced with unprecedented economic and market disruptions. Given the rigor of the COVID-19 assumptions in the most recent sensitivity analysis, it hard to imagine how the Fed could rationalize the use of even more draconian assumptions in the upcoming capital re-submissions this fall.  In particular, an approach which employs implausible levels of severity would appear to run counter to the Fed’s commitment to non-procyclical capital requirements. It would also inhibit the economic recovery by hampering banking organizations’ ability to provide much needed capital and liquidity to US businesses and consumers.

Coryann Stefansson is Managing Director and Head of Prudential Capital & Liquidity Policy.