Reversion to ‘un-inversion’

The slope of the yield curve has a long and storied reputation as a recession predictor. Researchers such as Frederic Mishkin, who literally wrote the textbook on monetary policy and financial markets, have been writing about the power of this indicator since the 1990s.

Even further back, you can find references in the 1960s from Reuben Kessel who examined the “the cyclical behaviour of the term structure of interest rates” for the National Bureau of Economic Research.

On the New York’s Fed dashboard, the yield curve indicator has been flashing red for the last two years, with the implied recession probability consistently above 50%. There’s also a helpful Q&A which emphasises why we should care:

“The yield curve has predicted essentially every US recession since 1950 with only one “false” signal which preceded the credit crunch and slowdown in production in 1967. There is also evidence that the predictive relationship exists in other countries notably Germany, Canada, and the United Kingdom”

You can argue about whether an inverted yield curve causes a recession (typical story: banks that are forced to pay more for their deposits restrict the supply of credit) or merely pre-empts a recession (typical story: investors see a downturn coming and therefore expect short-term rates to fall in the future). However, the empirics are strong in our view.

Feline uncomfortable?

When I was growing up, my father had an odd favorite expression: “there are more ways of killing a cat than drowning it in milk”. Rather than a lesson in feline lacto-asphyxiation, this was a 19^th century saying which meant that there are many ways to achieve the same objective. There are lots of different cuts of the yield curve (5-year vs. 30-year, 3-month vs. 10-year, 3-month vs. the 3-month rate in 18 months’ time etc. etc.), and academics argue endlessly over which is more prescient.

However, I prefer to just keep it simple. The difference between 2-year yields and 10-year yields is the typically used market default. On that metric, the yield curve has been continuously inverted from July 2022 until the week before last. Should we celebrate the fact that over two years of inversion have now come to an end? Unfortunately, not.

The chart and table below show the alignment between yield curve ‘un-inversion’ and US recessions. It makes for uncomfortable reading. In each of the last four economics cycles, the curve was flat or inverted until just a few months before the recession begun and started to steepen 3-6 months beforehand. In all those cases, Federal Reserve rate cuts were not enough to save the day and avert an economic and market downturn.

So is a recession inevitable today?

No, we are dealing with small samples here. Other ‘cast iron’ indicators have proven distinctly brittle. I’m looking at you, the Senior Loan Officers’ Survey! This yield curve signal could end up looking just as suspect.

Nonetheless, it does suggest that we should be on the lookout for signs of deteriorating data. The recent employment report offered something for everyone, but few hints of a sudden stop in economic activity. The economic consensus attributes a 30% chance to a recession in the US over the next 12 months. We worry that’s on the low side given the kind of signal thrown up here, but don’t have much conviction in such statements given so many unusual developments in the post-pandemic economy.

What does this mean for our asset allocation?

First, we’ve stayed on the cautious side on risk assets by maintaining an underweight in credit despite a largely benign default environment. Credit investors don’t yet seem to be demanding much compensation for the risks signaled by the yield curve.

Second, we’ve had a long-standing enthusiasm for US treasuries and only recently reappraised that view considering increasingly extreme pricing at the front-end. The yield curve has only ‘un-inverted’ because the market now thinks policy rates will be below 3% by the end of next year (see chart above). Getting more than the implied 250bp of rate cuts without a recession would be unprecedented. It is possible, but is tough to envision.

It looks like the government market is finally getting the message that recession risks are real.