In this post, inspired by Crossing Wall Street, I’m going to look at the spread between AAA and BAA-rated corporate bonds and show that, generally speaking, high spreads have been bullish for the stock market, except when spreads reach extreme heights (like they have at this very moment).
A little perspective – right now the corporate bond spread is at levels not seen since 1943. See graph below. Note that AAA (blue) and BAA (green) yields are on the left axis and the spread (red) as a percentage of the AAA yield is on the right axis.
All interest rate data from St. Louis FRED Database
Conventional wisdom says that a high spread between the yields on top-rated AAA and less creditworthy BAA corporate bonds is a bearish sign for the stock market. It signals that lenders are risk-averse because they are paying a significant premium (in the form of a lower yield received) for high-quality companies.
But excluding times of severe market stress, this relationship has been contrarian (the stock market outperforms when things look particularly bad). As evidence, in the graph below I’ve divided the weekly spread since 1962 (as a % of the AAA yield) into four quartiles from lowest (purple) to highest (red), and shown the results of trading the Dow the following week.
[logarithmically-scaled, data in monthly-intervals]
Clearly, with the exception of our current crises, high corporate bond spreads have been followed by stock market outperformance.
But as I mentioned previously, in times of severe market stress, this contrarian relationship falls apart. The graph below shows the same 4-quartile test, this time using monthly data from 1928 (because more data is available in monthly intervals) in order to capture those extremely high spreads in the 30’s and 40’s.
[logarithmically-scaled, data in monthly-intervals]
At above-average spreads (green), the conclusion is still the same, the market performs well. But at extreme historic levels (red), the market has bombed. It’s noteworthy that our current bear market slipped into this highest quartile before it began its death-spiral.
Closing thoughts…
I think a similar conclusion could be drawn about a lot of contrarian indicators. To paint a picture – the market is like a little kid playing with a rubber band…stretching it back and forth, and back and forth in a predictable rhythm…until one day it breaks and pokes out his eye.
Knowing the difference between stretching and breaking is really the hardest part of being a contrarian trader.
Happy Trading,
ms
P.S. this post reminds me of my report Rising Credit Spreads and the Stock Market and makes me wonder if the results had been different had pre-1953 data been included.
Geek Note: I calculated the spread as a % of the AAA yield i.e. ((Baa – Aaa)/Aaa) to try to normalize the data and create a conclusion that held true regardless of whether all bonds were high or low at any given time. Usually the absolute spread (Baa – Aaa) is used instead. I tested the absolute version as well and came to basically the same conclusion, although less clearly defined as the normalized version.
To stay up to date with what’s happening at the MarketSci Blog, we recommend subscribing to our RSS Feed or Email Feed.
Filed under: Treasuries & Interest Rates | 2 Comments
The same is true or implied volatility, which is highly correlated with corporate bond spreads. High implied volatility is usually bullish, unless it goes super high.