Recession models: walk forward testing

One common criticism of recession models is that they are back-fitted and therefore unlikely to continue working in real time.

RecessionAlert has an interesting walk forward analysis of their recession model, demonstrating robustness:

We achieve this by simulating a person sitting in December 1968, right before the second recession shown on all the above charts and attempting the optimization on known co-incident data to that point. We then project that model into the future to present day to see how it would have continued to perform in “out of sample” data … We repeat again including data to the 4th recession, 5th recession and so on …

We make several remarkable discoveries from the tests. The first model built hypothetically in 1969 using data from 1st July 1959 through to Nov 1969 and encompassing one recession and expansion (red line in above chart) continued to perform remarkably well into the future. Remember, once this model was constituted in November 1969, it was never modified or altered.

The second important observation is that there is not much difference between the 1st model and the 6th.

Avoiding equities during recessions is a powerful way to out-perform the market without frequent trading.

RecessionAlert also has an S&P 500 timing model in beta which combines sentiment signals (essentially buying the dips), with a recession model to avoid waterfall declines.  Presumably this method (which is similar to the model in this blog) could be improved by instrument selection using momentum.  (Although the exact trade dates required for testing are only available to subscribers).


2 thoughts on “Recession models: walk forward testing

  1. It’s certainly true that stocks are a bad idea around recessions, but how can you time a recession when all such forecasts usually fail? In the last years, ECRI was just the last example of such a model failure. Thanks, Joe

    • Hi Joe,

      Models do fail but good ones are far better than guesswork. Valid approaches are 1) walk forward testing as described in this post (method shown to work over last 50 years with no information from the future) or 2) use a successful model unchanged over a long period, for example Bob Deili’s Mr.Model:


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s