By Michael McCracken, Assistant Vice President and Economist
Forecasting recessions is notoriously difficult. Even so, knowing that a recession may be imminent enables consumers and businesses to make better strategic decisions.
For example, households may want to increase their savings in case they lose jobs, while companies may reduce investments to adjust for lower expected future demand.
It is therefore not surprising that a lot of research has gone into identifying consistent, simple-to-use signals of an oncoming recession.
In a previous post, Michael Owyang and Hannah Shell discussed how the “slope” of the yield curve for U.S. Treasuries is often considered one of these useful signals.1 Here, the slope is defined as the difference between the yield on a long-maturity rate (such as the 10-year) and the yield on a short-maturity rate (such as the one-year).
This slope is used as a signal because it has switched from positive to negative prior to the last seven recessions. This slope has gained momentum in the news because, as shown in figure below, it is near its lowest level since the last recession.
However, as the saying goes, don’t put all of your eggs in one basket. For example, the slope of the U.S. yield curve turned negative in 1966, yet a recession did not occur for many years after.
In the context of other countries, Australia has had its yield curve invert three times in the last 25 years and at no time experienced a recession! Clearly, there are other factors at work beyond just the slope of the yield curve.
As a general matter, data that have a forward-looking component should also be considered as a predictor of future recessions. For example, individuals buy stocks based on a firm’s expected future, rather than past, performance. Hence, a rapid decline in stock returns is often considered a predictor of future recessions.
As another example, a rapid decline in the University of Michigan’s Index of Consumer Expectations (a product of their Surveys of Consumers) is also considered a predictor of future recessions, because the survey explicitly asks individuals about their outlook on economic conditions over the next few years.
Another intriguing signal has been recently studied by researchers at the central bank of Norway.2 They showed that residential investment (measured by its contribution to GDP growth and plotted in the figure below) is a useful predictor of recessions and has more predictive content than stock returns and surveys of consumer expectations.
Importantly, they also showed that residential investment and the slope of the yield curve are better at predicting recessions jointly than they are individually. These results suggest keeping an eye on both before predicting a recession.
What do these indicators say about a recession occurring in the near future? Most suggest that one is not imminent. For instance:
There are some concerns that the yield curve will decline further and eventually go negative as the Federal Reserve continues to hike interest rates.
Even if that were to happen, though, one would be wise to consider movements in other leading indicators, such as those mentioned in this post, to get a better sense of whether a recession is truly on the horizon.
1 Owyang, Michael; and Shell, Hannah. “Is the Yield Curve Signaling a Recession?” Federal Reserve Bank of St. Louis On the Economy, March 24, 2016.
2 Aastveit, Knut A.; Anundsen, Andre K.; and Herstad, Eyo I. “Residential Investment and Recession Predictability,” Norges Bank Working Paper 24/2017, November 2017.