The US bond market is the best predictor of recessions
In this paper, Edward Leamer proposes a new way of displaying and analyzing macroeconomic time series to form recession forecasts. His new methodology focuses on the spread between the yield of the 10-year Treasury and the 3 month Treasury. The interest rate spreads in the figure below reveal that a flat or inverted yield curve is a symptom of the last year of the expansion, but the current slope of the yield curve of about 140 basis point fits best the third year before recession not the first. In addition, it is clear that Volcker and Burns chose record negative values of the interest rate spread not long before the recessions came.
Leamer then looks at the 3-Month Treasury yield, the inflation rate, the unemployment rate and housing starts and what these four variables are signaling about recession probabilities:
- The inflation data raises great concern because it reveals that inflation rose to very high levels in the year before recessions in the Volcker and Burns eras, and the current dramatic increase in inflation matches those episodes;
- The unemployment data reveals that the current low unemployment rate is also a symptom of the last year of recession, if the covid-19 data are excluded. With the covid-19 data our recent unemployment levels have taken a path never seen before;
- The housing starts data reveal that starts decline very near the end of the expansion, and the housing starts through May 2022 do not support the conclusion that we are in the last year of this expansion.
To reconcile the conflicting signals from all the above variables, he uses a probit analysis to create a “horse race” among them that might predict recessions. He concludes:
[I]t’s the bond market yields , not inflation, unemployment or housing starts that are the critical forecasting variables. It’s an inverted yield curve which is correcting itself with an increase in the spread, and also a high 3-month Treasury yield that best predict the outcome.
A New Way of Forecasting Recessions
Author: Edward E. Leamer
From: UCLA
Is the bond market in Europe the best predictor of recessions?
Based on monthly data from 1970 to 2022 and the AUROC performance metric, we show that yield curve inversions generally predict recessions in the euro area. However, there are two important limitations. First, the forecasting capability of the yield curve has tended to weaken since the global financial crisis. Second, the overall performance is driven by core eurozone countries, especially Germany.
Risk premia, and more particularly the credit risk component, blur the relationship between the yield curve slope and the probability of a future recession within the so-called peripheral countries.
Do Yield Curve Inversions Predict Recessions in the Euro Area?
Authors: David Sabes, Jean-Guillaume Sahuc
From: Banque de France
