The biggest news from the US Federal Reserve this week won’t be the results of Wednesday’s FOMC meeting. Chair Powell and other Fed officials have already conditioned markets to expect a dovish tilt to the FOMC statement, Summary of Economic Projections, and press conference.
Rather, we’re looking forward to the results of the annual stress tests on large US banks slated for release on Friday June 21st. Financials are 13% of the S&P 500, after all, and most of that weighting rests with the largest institutions at the center of this yearly exam.
Beyond their impact to bank stocks, however, we’re always curious to see exactly how the Fed constructs its “baseline, adverse, and severely adverse” scenarios at the heart of this process. These change every year and amount to the central bank’s answer to the question “What’s the worst that could happen?”
Here’s the answer from the Fed’s February 2019 document outlining the stress tests (link at the end of this section).
- US real GDP: 2.25% in 2019, 1.5% in 2020, 2.0% in 2021
- Unemployment: 3.5% in 2019, increasing to 4.0% by the first half of 2021
- Inflation: 2.25% in 2019, 2020, and 2021
- Short term Treasury yields: 2.5% in 1H 2019, 2.75% in 2H 2019, flat thereafter
- 10-year Treasury yields: 3% in 2019, 3.25% by the start of 2020, 3.5% in 2021
- Corporate/mortgage spreads stay flat
- US equity prices rise 5% annually in 2019, 2020 and 2021 and volatility declines
- International GDP growth: 5.75% – 6.0% in “developed Asia”, 0.5% – 0.75% in Japan, 1.50% to 1.75% in Europe
Severely Adverse Scenario
- US real GDP: an 8% decline from pre-recession peaks, bottoming in Q3 2020
- Unemployment: rises to a peak of 10% by Q3 2020
- Inflation: falls to 1.25% in Q1 2019 but rises to 2.0% by 2H 2020
- Short term Treasury yields: zero percent through the end of 2021
- 10-year Treasury yields: 0.75% in Q1 2019, 1.5% in Q1 2021, 1.75% in Q1 2022
- Investment grade corporate spreads widen to 550 basis points in Q3 2019 from 200 bp in Q4 2018. Mortgage spreads widen by 350 bp.
- US equity prices fall by 50% through the end of 2019 and the CBOE VIX Index hits 70. House prices fall by 25% and commercial real estate prices by 35%.
- Severe recessions around the world, save developing Asia, and deflation. The dollar rallies against all major currencies except for the Japanese yen.
Adverse Scenario (essentially a watered-down version of the previous case)
- US real GDP: a 2.75% decline from peak, 0.75% growth in 2020, 3.25% growth in 2021
- Unemployment: a 7% peak instead of 10%
- Inflation: remains near 2%
- A decline in short and long term Treasury yields, but with the 10-year back to just below 2.0% by 2021
- A 20% decline in US equity prices, troughing in Q4 2019
Three important points about all this:
#1: Inflation. We were struck by the stress test’s assumption that inflation will remain at 2% in a garden-variety recession (the “adverse scenario”) and still show a 1% increase in a crisis (“severely adverse scenario”). Here’s what the TIPS market is saying just now:
- 5-year TIPS breakeven/expected inflation: 1.51%
- 10-year TIPS breakeven/expected inflation: 1.68%
- For reference: May headline CPI was 1.8% higher year, core was +2.0%
Takeaway: the market is implicitly questioning the Fed’s stress test assumption that inflation can/will remain near 2% over its scenario horizon, and we’re nowhere near even an “adverse” economic environment at present.
#2: Interest rates: while we appreciate the Fed’s need for internal consistency, its baseline scenario (published in February) was clearly – and dramatically – wrong on Treasury yields. Two-years pay 1.85% at present, not 2.5%. Ten-years yield 2.08%, not 3.0%. These are large misses across just a 4 month window (February’s release of the stress test criteria to now).
Takeaway: if the Fed’s baseline scenario is already out of date, what does that say about the one labeled “severely adverse”? In the Fed’s assumptions regarding that environment, Treasuries still have a positive yield. Given that 10-year German bunds now have a negative 0.25% yield (and nothing has “hit the fan” yet), this seems optimistic.
#3: The speed/degree of the recovery. We’ve done this analysis of the Fed’s stress test scenarios for several years now, and the same question always nags at us: “what will spark a recovery in economic conditions, and how effective will it be?” The traditional policy tools to counter a recession – fiscal and monetary stimulus – may be weaker now than in 2008/2009, for example. US debt to GDP was 63% in 2007; it is 105% today. Fed Funds were 5.3% in 2007; they are 2.37% today.
Takeaway: to our thinking the greatest weakness of the Fed’s stress test scenarios is that they focus on “winning the war” (maintaining financial system liquidity in a deep recession) but pay less attention to whatever comes after. What if US rates go negative in a crisis and don’t flip back to positive afterward? Stress of a different kind, that, but one scenario worth considering.
Summing up: the Fed’s bank stress tests highlight the central bank’s own institutional blind spots. The purpose of these annual exams is to reassure markets and society as a whole that the financial system can withstand a serious shock. But because they also reflect how the Fed sees its own efficacy in managing its dual mandate of employment/inflation, the scenarios may well fall short of true worst-case outcomes.
As with all things, markets therefore will be the final judge of how robust the financial system is to a shock, and how well the Fed has done its job in preparing for that possibility.
The Fed’s February 2019 outline of scenarios: https://www.federalreserve.gov/newsevents/pressreleases/files/bcreg20190213a1.pdf