Why is the 10-Year Treasury Constant Maturity Minus 3-Month Treasury Constant Maturity Curve significant, but not the 5-year minus 1-year?

Question

I am new to investing and am trying to learn about economic indicators.

When the yield curve inverted this year on March 27th, the media published articles saying that this might indicate a recession. The Federal Reserve Bank of St. Louis site tracks the 10-Year Treasury Constant Maturity Minus 3-Month Treasury Constant Maturity curve, with the correlation with past recessions. It tracks a few other curves, correlating them with past recessions, perhaps because they are supposed to be significant with respect to recessions.

Thanks to Bob Baerker's answer on the Money site, I was able to get raw fed rate data. With this, I learned that the 5-year Treasury Constant Maturity Minus 1-year Treasury Maturity Curve was already inverted on 24th December, 2018, and has been almost constantly inverted since then!

So, why is it that nobody cares that this 5-year minus 1-year curve has been inverted since late last year, but everybody cares that the 10-year minus 3-month curve inverted this year? Why is it that the former is considered irrelevant to recessions, but the latter is considered relevant?

Answer 1

As a first approximation, we can think of the 3-month bill as as being the “market forecast” of the (geometric) average of the overnight rate for the next three months, while the 1-year is the average for the next year.

In a rate-hike cycle, the 1-year rate ends up much higher than the 3-month rate. This means that it will get close the the 5-year rate much faster.

Furhermore, there is presumably a greater risk premium - term premium - in the 1-year versus the 3-month maturity. This will also tend to compress the 1-/5-year slope faster.

Once the yields on the 1-year and 5-year are close, small “technical factors” in yields are much more likely to lead to an inversion. Therefore, it would be more likely to generate false positives.

The reality is that market commentators and economic researchers have searched through all the various slopes to find the “best” ones for forecasting. So it is probably not safe to say that the 1/5 slope is “irrelevant.”

In any event, yield curve slopes are hardly perfect. I discussed some of the issues in this question. Since slopes are not infallible guides to recession forecasting, one can argue that the choice of maturities (technically, tenors is a better term) is somewhat arbitrary.