Wednesday, November 22, 2017

The yield curve: we're not DOOMED yet


 - by New Deal democrat

Here's another item to ponder while I am away this week....

An inversion in the yield curve (i.e., shorter length bonds yielding higher than longer term bonds) has (with the exception of 1966) been a reliable sign of an oncoming recession, at least since the late 1920s. Mind you, the curve did NOT invert at all during the later 1930s, 1940s, or early 1950s, despite a number of recessions, including the big 1938 recession, taking place.

But a yield curve inversion = bad, no doubt about it.

There's been a lot of chatter recently about the yield curve flattening, with a few of the usual DOOOMish suspects calling it recessionary even now.

Hold your horses.

Here's a long term look at the 10 year minus 2 year yield curve:



Note that an inversion has been a reliable signal of a recession beginning in the next 12 to 24 months.

But we're not even inverted now!

Here's a look at the yield curve for the last 5 years:



The trend from peak to peak since the beginning of 2014 has been a slow tightening on the order of 0.4% to 0.5% per year. If that general trend were to hold, we're 18 to 24 months from an inversion.

Even if we use the steeper trend decline that began from the peak last December, we're still 10-12 months from an inversion.

But let's suppose we get at least a partial inversion 6 months from now.  That would signal a recession 12 to 24 months afterward -- in other words, 1 1/2 to 2 1/2 years from now. That's 2019 or 2020.

In short, even if the yield curve gives an accurate signal this time around, it would still mean that the expansion has a ways to go.  Sorry, but we're just not DOOOMED yet.