Monetary Policy & Inflation | US
Which Markets Perform During US Recessions?
- By Bilal Hafeez
- 3 min read
Summary
- US equities tend to underperform from around six months before a US recession onwards.
- Bonds tend to outperform from around three months before a recession onwards, but the pattern was more mixed in the 1970s/early 1980s.
- We maintain our underweight bond and equity view.
Our call that a US recession is likely by next year has focused our readers’ minds. Indeed, I am now commonly asked how to position for one. Our asset allocation view has been to underweight bonds and equities and overweight cash and commodities.
We like cash because your downside is limited, and it gives you optionality to buy into large drops in markets. We like commodities because there are significant supply-demand imbalances in commodity markets, which policymakers are struggling to manage. But is it right to be underweight bonds and equities? We look at how these assets have performed around US recessions since the 1970s.
The punchline is that equities underperform around US recessions – especially from six months before a recession onwards (Chart 1). They have fallen on average by 4% versus cash in the six months leading up to the seven US recessions we have had since 1970 and by 6% in the three months leading to them. The falls are even larger after a recession, with stocks dropping an average of 11% in the six months after.
Meanwhile, US bonds tend to outperform both in the immediate run-up to recessions and after. The clearest performance is in the three months leading up to a recession when bonds outperform cash by 3%. They also rally after the start of a recession.
A breakdown by recession reveals the 1980 and 1990 recessions did not see equity weakness into the recession. Meanwhile, the COVID recession was the only one to see strong equity performance after the start of the recession (Table 1). For bonds, the 1973, 1980 and 1981 recessions gave more mixed performances (Table 2).
We see parallels for bonds with the 1970s and 1980s, which suggests our underweight bond positions may not be out of line with the historic data. Meanwhile, our underweight equity view is consistent with most of the historic data.
