Introduction
Some news outlets in the US have reported that the yield curve for US-government bonds is inverted, or, in other words, the interest rate for bonds with a shorter horizon are now higher than interest rates for US-government bonds with longer maturities. This “inverted yield curve”" is seen as a good predictor for an upcoming recession. For better illustration the interest rates for 2 year, 3 year and 5 year bonds:
Figure 1: The interest rate for 2 year, 3 year and 5 year bonds from November the 20th until December the 7th
On December the 3rd the 5 year interest rate was below the 3 year interest rate and on par with the 2 year interest rate. Since then the 5 year interest rate has fallen below the 2 year interest rate. Most, but not all, experts interpret this as a sign for an upcoming recession. But why?
Yield Curve and Inverted Yield Curve
To understand the link between recessions and the yield curve, let’s look at the following graph, which portrays the yield curve from 1960 until end of octobeer. The data is taken from Gurkaynak, Sack and Wright.
Figure 2: Yield curve from June 1961 until November 2018. Yields are end-of-month yields.
Usually long-term interest rates reflect that buying and holding long-term bonds carries more risk, in particular inflation risk. Thus, long term nominal interest rates should be higher than short-term interest rates. As one can see from the graph most of the times the yield curve is normal, i.e. short-term interest rate is lower than long-term interest rates. Or, in other words, the yield curve is upward sloping. However, several months stand out because short-term nominal interest rate is higher than the long-term nominal interest rate. Some examples for an inverted yield curve followed by a recession are:
- December 1978 - followed by the 1980-82 recession
- May 2000 - followed by the bursting of the Dot-Com bubble
- July 2006 - followed by the Great Financial Crisis
Because an inversion of the yield curve was usually (with one exception) shortly followed by a recession, an inverted yield curve is seen as a good predictor for recession.
As a word of caution, one should not confuse ‘prediction’ with ‘causation’, i.e. it is not the inverted yield curve that causes a recession but rather the underlying economic conditions that cause a recession are also reflected in the yield curve. For instance, the Recession in 1980-82 was not caused by the inverted yield in January 1980 but rather by a reversal in monetary policy. The 1970s were characterized by a high inflation attributed to oil price shocks. To combat the rising inflation the Federal Reserve adopted a strongly contrationary monetary policy that subsequently caused the recession. Similarly for the other two recessions in our examples.
However, comparing the yield curve from the week of December the 3rd to previous yield curves there is one striking difference: Only a small section of the yield curve is inverted, i.e. only the 3yr yields and the 2yr yields were higher than the 5yr yield, while for other inversions much larger parts of the yield curve were inverted. Hence, the inverted yield curve only weakly predicts a recession.
Conclusion
Compared to previous periods where an inverted yield curve was followed by a recession, the spread is very small compared to previous recessions and it is only on a small section of the yield curve. This seems to be a case of confirmation bias as the last recession was 10 yrs ago and thus pundits expect a recession and are looking for signs of that recession.