There is an old chart from Bank of America that shows how every Fed tightening cycle ends in a crisis.
… the only question is when. It is perhaps worth noting that the last time we showed this chart was in late 2018, just before Powell capitulated on the Fed’s last tightening cycle, and not long before the covid pandemic sparked the biggest monetary and financial stimulus in world history. In other words, it is virtually assured that every time the Fed pulls away the punch bowl the market crashes.
Another thing that is virtually assured, is that by the time the tightening Fed realizes it has made a policy mistake, the economy is in a recession.
It’s why as DB’s Jim Reid writes in his Thursday Chart of the Day, the Fed’s decision will have profound implications for the start date of the next recession.
Indeed, having accelerated their taper, it now opens the path for the first hike in the cycle to occur in H1, with March now widely cited as the month of “liftoff” but for the purpose of Reid’s “thinking note,” let’s say the first rate hike is in June.
So given that the Fed tightening eventually contributes to recessions, when should the next one occur?
A look at the table in a recent DB note “When the Fed hike: what happens next?” suggests the median and average time to the next recession is 37 and 42 months after the first hike. So that takes us to July 2025 and December 2025 respectively. That said, the earliest gap over 13 cycles is 11 months and that would take us to May 2023.
Alternatively, if we go a different route and look at the yield curve, today’s CoTD shows that the 2s10s pretty much always flattens after the first hike. By an average of 80bps over the next 12 months. So a June 2022 first hike may bring an inversion by June 2023 given the 2s10s currently sits at just under 80bps. Obviously that assumes a flat curve from now until June next year which is a big assumption.
Anyway, a recession occurs around 18 months after the 2s10s inverts on average. So that would signal the average recession starting around December 2024. Prior to the pandemic the earliest a recession occurred after a 2s10s inversion was 9 months so that would be around March 2023.
While Reid concedes that there are some “sweeping assumptions used here” this gives you a framework for thinking where we are in the cycle. Clearly each cycle is different and many argue that the Fed is already behind the curve and as such they should have been tightening policy already. This should mean a more compressed cycle relative to history.
Alternatively as we saw in the mid-1960s, with the Fed making an error by keeping policy too loose, they delayed the eventual recession to late-1969 but left us with an inflation problem that created big economic problems in the 1970s that the energy shock then compounded.
So that is the trade off, one which even former Fed uber-dove Kocherlakota is railing against (yes, once the Fed’s biggest dove, the former Minneapolis Fed president is now an ardent hawk).
All else equal, at this stage history would suggest a US recession in 2024 or 2025 is a realistic assumption, and while it could certainly come earlier, that would assume the earlier end of the historical template.
Incidentally, 2023 is when most respondents (31%) to the latest Deutsche Bank market survey said the next recession would hit, with 2024 just a few percentage points behind at 29%.
Incidentally, the market seems to agree because as we showed recently, traders are now pricing in an inverted 4Y1Y – 2Y1Y forward swap curve, meaning that fed funds in 2025 are seen below 2023, suggesting at some point in the interim, the Fed’s next easing cycle will begin.
