We have been receiving an increasing number of questions about whether an inverted yield curve is predictive of recessions. We recently posted a primer on what an inverted yield curve is and below is a fuller discussion of whether an inverted yield curve is predictive of recessions.
Looking at the 10 year - 3 month yield spread (above), we see that the curve has a fairly good record of predicting recessions. It has inverted before each of the last seven recessions (no false negative signals), but has had two false positive signals in the late 1960s and late 1990s (black circles). Inversion has occurred anywhere from four to twenty months before the beginning of a recession.
If we look at the 10 year - 2 year yield spread (above), we see a similar pattern although the data doesn't go back quite as far. Inversions have occurred 13 to 24 months before the beginning a recession. There have not been any false negative signals and just one false positive (in the late 1990s). This section of the yield curve has not inverted.
If a recession begins in the next two years, it will be the first time that the 10Y-2Y has "missed" it. If a recession does not begin in the next two years, it will be the third false positive signal for the 10Y-3M yield. So what is an investor to do in the face of conflicting information?
- At a minimum, we believe that caution is warranted as both yield curves are very flat. However, the yield curve has been flat in the midst of many economic expansions. Furthermore, the typical lag between an inversion and a recession is 6-18 months. As mentioned, we do believe caution is warranted at this point, but are not fully convinced that a recession is upon us.
- Monitor all of the yield spreads, but never become overly reliant on any single yield spread, indicator, or signal. Data often provides conflicting information and indicators can provide conflicting and/or false signals.
- Look to other economic data to gain a fuller understanding of what is going on. Attempt to determine where the weight of evidence lies. Does other data support the 10Y-2Y conclusion or the 10Y-3M conclusion?
We will cover additional economic data points and "recession indicators" in future posts. Hopefully the above is helpful in filtering out the noise of media headlines regarding the inverted yield curve and its implications.