Reclaiming Common Sense

We only have three periods of inversion data for the three month and ten year. We do not have the 3 month yield data for the first Reagan Recession. We do see from the non-farm payroll data that employment fell during the recessions of 1991, 2001, and 2008. There was an inversion, more like a coupling, of the three month and ten year rates during 1991, a small inversion during 2001, and a protracted inversion during 2007-2008.

July 2000 through January 2001 Inversion. This inversion was a sustained inversion of seven months. The recession lasted from May 2001 through November 2001. The inversion started ten months before the recession and was done four months before the recession. This was the mildest recession since Reagan.

The recessions that we have had have been jobs recessions. Jobs have been growing. We have been setting Current Employment Statistics (CES) worker records and Current Population Survey (CPS) jobs records. The last recession was a Housing Recession, followed by a Jobs Recession, followed by a retail recession, which was recorded as a Gross Domestic Product (GDP) recession. It was also a 2-10 recession and a  three month ten year bond recession. New home sales are on track for their best sales year since 2007. Workers are growing at roughly 1.5%. Wages are rising at over 3%, depending upon sector.


We had back to back GDP growth of roughly  0.9% during 2015Q4 and 2016Q1. We had ten months of CPI-U deflation during 2015. We had declining existing home inventory during 2015, 216, 2017 and 2018. We were heading into a recession. Past tense. It didn't happen even though the data indicated that were were on the brink of a recession.


If the 2-10 inversion is essential as an indicator for a recession 12-24 months out, and if the 2-10 inversion did not happen, and it didn't happen, does the inversion of the three month ten year bond rates matter? Probably not.


It's the Economy.


PS. The bond rates as of the time of writing this article are 3 month (1.53%,) 2 year (1.662%,) and 10 year (1.926%.)

June 1989 through August 1989 and November to December 1989. This inversion lasted for a total of five months out of seven total. There was a reprieve during September and October, The recession officially was July of 1990 to March of 1991. This means that there was a thirteen month lead time from the beginning of the first inversion to the beginning of the recession.

August 2006 through May 2007 Inversion.  This inversion saw the 3 month rate drop to zero (0.03%) at the end of 2008 and go as low as 0.02% October 2015. This inversion lasted eight months and preceded a recession that officially lasted eighteen months. Note that the jobs level did not recover to pre-recession levels as did the other two recessions. The non-farm payroll level was 139.297 million December of 2007, We would not eclipse that level until June 2014 when we hit 139.824 million. The inversion started 16 months ahead of the recession and ended seven months before the recession started.

 If Bad News is Reported but not Recorded, did it really Happen?

First it was the 2yr/10 yr Inversion - Now its the 3 month/10 year Inversion


There were a considerable number of Chicken Little Economists and Chicken Little Reporters who were claiming that the end of the expansion was near because the 2 year bond rate was higher than the 10 year bond rate. There was/is a theory that the inversion of the rates foretells of a looming recession within 12-24 months. A lot can happen during 12 to 24 months. When that inversion "ended," the talk went to the 3 month and 10 year inversion. The article "What Bond Yield Inversion" explained that the inversion has to last for months, not moments, and that the 2-10 curve did not invert on a monthly basis this year - ever. What about the 3 month and 10 year rates? It was the shortest inversion.

June 2019 to September 2019 Inversion. This was the shortest of the four inversions. The first inversion was 5 months out of seven. The second inversion was seven months. The third inversion was ten months. This inversion was four months, comparable to the first leg of the 1989 inversion. This inversion had the smallest spread, depending what starting point you use. If you use January 2016 then the spread was 1.85%. January 2016 it was 1.83%. January 2017 was 1.91% and January 2018 was 1.15%, while the spread was just 0.29% January 2019. It was going to happen