“Everyone has a plan until they get punched in the mouth.” That’s Mike Tyson’s most famous quote, and as we reviewed the Federal Reserve’s 2020 Stress Test “severely adverse scenario” this weekend we couldn’t help but think about Iron Mike. Instead of a regulatory exercise, the US central bank finds itself essentially living in its own simulation. Or, in truth, a darker version of it… Punched harder than it might have thought possible, sending their plans to the mat for an 8 count.
We’ve been looking at these Stress Test scenarios for years, and always wondered as much about the assumptions for the recovery from recession as much as how bad a downturn might get. Both these matter because the whole point of the Stress Test process is to make sure the US banking system could act as a counter-cyclical balance during recession. Yes, credit standards might increase but banks should still be lending in a downturn to cushion the economic blow.
Here is how the Fed modelled a severe recession in the 2020 Stress Test:
#1: Real GDP growth:
- Q1 2020: -5.3%
- Q2 2020: -9.9%
- Q3 2020: -7.6%
- Q4 2020: -5.3%
Takeaway: so far so good – this really is a severe recession – and the 2021 numbers don’t have GDP growth positive until Q4 (+2.9%).
#2: US Unemployment rate:
- Q1 2020: 4.5%
- Q2 2020: 6.1%
- Q3 2020: 7.4%
- Q4 2020: 8.4%
Takeaway: instead of a gradual rise in unemployment, the COVID-19 Crisis is causing mass joblessness all at once. The Fed’s 2020 adverse scenario had unemployment peaking at 10.0% seven quarters after the start of a recession (Q3 2021). We’ll get there much sooner, one assumes.
#3: Three month and 10-year Treasury yields:
- 3-month Treasury yields of 0.1% for the entire 3-year forecast window (2020 – 2022).
- 10-year Treasury yields bottoming at 0.7% at the start of the recession (Q1 2020) and rising by 10-20 basis points every quarter throughout the recession forecast.
Takeaway: these are broadly in line with what we’ve seen thus far (if a tad high), but the jury is out about the assumption that long rates will rise monotonically over the next 2-3 years.
#4: BBB Bond Spreads:
- Q1 2020: 450 basis points
- Q2 2020: 520 basis points
- Q3 2020: 550 basis points
- Q4 2020: 550 basis points
Takeaway: in these numbers we have a partial explanation for why the Fed launched its investment grade corporate bond buying program last week. On March 23rd (last Monday), spreads were 488 basis points, double their March 10th levels. Without Fed intervention, we would have blown through the central bank’s adverse scenario. While Chair Powell may not be using the Stress Test levels to set crisis policy, the speed of the move to clearly dangerous levels must have been part of the calculus.
#5: Dow Industrials and CBOE VIX Levels:
- Q1 2020: 22,262, 69.1
- Q2 2020: 18,623, 70.0
- Q3 2020: 16,910, 66.0
- Q4 2020: 16,581, 60.3
Takeaway: while the Stress Test scenario is explicitly not a Fed forecast, it is still worth noting that their “downside target” for the Dow in a severe recession is that year-end 2020 number of 16,581. At the worst close of 2020 on March 23rd, we were just 11% away. Not a bad call, in other words. As for the VIX levels in the Stress Test, we simply cannot imagine a year when the VIX averages 59. Even in the 12 months from October 2008 to September 2009, the VIX only averaged 41.
Summary: the Fed’s Stress Tests were borne out of the 2008 experience, a long and painful recession. What we have now is something completely different, the proverbial punch in the face. That tells us that the Federal Reserve will do whatever it can to limit the reaction of capital markets to its worst-case scenario. We just hope they’re wrong about their VIX levels…