Friday, November 13, 2009

Just One Super-Secular Mean Reversion?

In response to my post Where are Long Bonds Going? Late 1970's / Early 80's Edition, reader Henry Bee responds:

I'm going to replicate the study using 20-year bond in place of 30-year to see if that makes a difference because 20-year bond data goes all the way back to the 30's.
While I am definitely interested to see what Henry comes up with for the late 1950's-1970's, he'll likely run into a problem if he simply tries to replicate what I have shown going back any further. The problem? Rates were never high enough in the first place (outright), let alone on a relative basis to shorter yields.

Below we see just that, but with the 10 year Treasury bond (data goes back to the late 1800's). Interesting to see the yield bumps against the 10 year average on the way up without really penetrating the average AND then again on the way down, but in reverse.

Well, they always say it's dangerous to bet against a strong trend. From the looks of it, the decline in interest rates is more than a long trend. It is a reversion to a 100+ year mean.

Source: Irrational Exuberance


  1. Jake,

    I keep pointing out to people that long rates may be low by recent historical measures, but long rates are by no means low by long term historical measures.

    We suffer from a recency bias.

    I just don't see high interest rates going forward. Our high debt levels won't allow it. A weaker dollar - very possible. Higher interest rates - the system can't handle it.

    I was hoping a long dated treasury sell off (up to 4% yield on the 10 yr) would accompany rising gold, oil and equities. I'm starting to have doubts it will happen though.

  2. Most people take the 80s for "normal" concerning interest rates.Now is "extreme". It's the other way around.

    I wonder what really was going on in the 70s.