Investing in the stock market just before the beginning a recession can bankrupt you but investing just after the end of the recession can produce unimaginable riches. Take the 2008 recession, investing in an ETF following the S&P 500 in October 2007 would have seen your investment drop to about half its value just a year later. But investing on that same ETF in March 2009, would have doubled your money just two years later.
As investors in the stock market, the question of when the next recession will happen should be at the top of our minds, not just to protect our portfolio from being destroyed, but also to take advantage of the huge opportunity when the dust settles.
While there are many models that attempt to predict recessions (and many more economists that try, unsuccessfully), there are a few that are historically so reliable, it’s hard to say “this time is different”.
In this article we explore one of these extremely reliable models, which is based on the spread between the 10-year US treasury note and the 3-month US treasury bills. For those of you not well versed in treasuries, a note and a bill are the same thing, a US government “bond”, i.e. US debt sold to investors in exchange for interest payments.
A vast amount of research (pioneered by Professor Arturo Estrella almost 30 years ago) has been done on the subject of predicting US Recessions based on the spread of their yields, i.e. the difference, between the interest rate paid between these two types of US treasuries. The results are amazing, to say the least.
We crunched the numbers and since the time data has been available, a recession has been predicted successfully almost all the time. We decided to create a little web application (works best with larger screens) and here’s the screen shot (and link to our app) of what it looks like.
While the actual research calls for a probability calculations (which we computed too), as you can see we decided to translate those values into more human readable (and non-math-wiz friendly) range equivalents that go from “very low” to “very high”.
And the truth is in the numbers. From 1954 (when data is first available) all the way to today -almost 70 years- the prediction of “high” probability of a recession has been correct always except for 2 occasions. First, in 2000 where the recession officially didn’t start into 2001, and second, between 1966 to 1968, where no recession actually happened.
But the most amazing thing about these two sets of dates is that while the actual recession didn’t happen until later or never at all, those dates did coincide with market peaks.
So, while the recession happened much later (or never at all), the stock market was already in panic mode. Maybe we should re-name this model from recession-predictor to stock-market-major-crash-predictor.
Take our course on Udemy!
THIS WEBSITE CONTAINS ONLY GENERAL INFORMATION, WHICH MUST NOT BE CONSTRUED AS SPECIFIC ADVICE OR RECOMMENDATIONS
