Running from the shadow of the yield curve

In recent days the equity market has appeared to take fright at a change in the difference between the yield US long term interest rates and the yield on short term us interest rates.

This difference between US long term interest rates and US short term interest is called "the yield curve".

Figure 1: US Ten Year Bond Yields minus US 90 Day Treasury Yields

There are a couple of popular ways of measuring the US yield curve. The first one is the difference between US 2 year bond yields minus US 10 year bond yields. The second is the difference between US 90 day Treasury bill yields minus US 10 year bond yields. We show this measure of the yield curve in Figure 1 above. It is widely said that when either of these differences in interest rates or measures of the yield curve falls below zero, then that movement forecasts a coming US recession.

Our attention was drawn to this issue in October 2017. At that time, the President of the Philadelphia Federal Reserve Patrick Harker, did an interview with Bloomberg in which he drew attention to this issue. He said the Fed was closely watching the shape of the yield curve as the Fed tightened monetary policy by putting US short term interest rates.

He said that the Fed was wary of allowing the yield curve to "invert". This suggested that the Fed was very aware of the reputation of the yield curve as a forecaster of US recessions.

We then decided to examine the effectiveness of the yield curve as a forecaster of US economic activity. As our measure of the yield curve, we chose the difference between US 90 day treasury yields minus US 10 year bond yields. As our measure of US economic activity, we chose the Chicago Fed national activity indicator. The Chicago Fed provides us with a history of the official timing of US recessions. Our research found that the yield curve was in fact quite a good forecaster of US recessions.

Since 1982 we found that when the yield curve "inverted" a US recession had indeed followed by around five quarters in the period up until 2000. Since 2000, we found that when the yield curve "'inverted" that a US recession had indeed followed by around nine quarters. However, we found that the most reliable measure seemed to be not when this measure of the yield curve fell below zero, but rather when this measure of the yield fell below 50 basis points. This means it fell below a level when US 10 year bond yields were half of one percent higher than 90 day Treasury bill yields.

Let us take as an example the most recent US recession beginning in 2008. On this occasion, the US yield curve inverted in the final quarter of 2005. It remained inverted until the end of 2006. The reason that it remained inverted, may have been that US inflation remained stable through 2006. It was not until 2007, that long lags between falling US unemployment and US inflation led rising inflation to lead to rising US ten year bond yields.

It was in fact this rise in bond yields, and the rise in US mortgage rates that followed that actually caused the slump in US activity, which showed finally itself as the US Great Recession of 2008 and 2009. It was the lesson of this period that it is NOT the yield curve itself that caused the recession. It is the sell-off in bonds and the rise in bonds yield AFTER the yield curve is no longer inverted that ACTUALLY causes the recession. The fact that it is not the yield curve inversion that leads directly to the recession which may explain why the lead time between the inversion of the yield curve and the beginning of the recession is so long.

In the past week, the level of the US ten year bond yield has fallen to a level where it is 0.45% above US 90 day treasury yields. This means that the yield curve is only just beginning to invert. It needs to remain inverted for more than one quarter to even begin to signal a possible later recession. Should this recession occur, it will not arrive before five quarters later. Based on the data since 2000, the recession will not arrive until nine quarters later. This means at worst no US recession before 2020 and probably no US recession before 2021. On either scenario, the US economy will see plenty of economic growth and plenty of US corporate earnings growth before then.

Conclusion

We can all sleep easy in our beds, the US recession bogeyman is not going to be stalking us in the night. A US recession may indeed actually finally arrive.

But we will all likely enjoy a long period of prosperity before then.