# Recession Probability Model

There are a lot of recession probability models produced by investment banks and research firms. Too many of these models give FALSE recession signals (i.e. they predict that a recession is on its way, but a recession doesn’t actually materialize in the near future).

*”Recessions” are defined by GDP, an extremely lagging economic indicator.

I’ve created my own Recession Probability Model. The concept behind this model is simple. Leading economic indicators deteriorate before recessions start. My Recession Probability Model looks for deterioration among these economic indicators. Combine the deterioration in this leading indicators, and you get the Recession Probability Model. The higher the percentage, the more likely a recession is imminent (i.e. will start within the next few months).

### # of indicators

The model after January 1971 uses 10 indicators and before January 1971 uses 9 indicators.

**A recession is imminent when the Recession Probability Model reaches 50% or above.**

- Before January 1971, this means that 5 out of 9 economic indicators show “deterioration”.
- After January 1971, this means that 5 out of 10 economic indicators show “deterioration”.

### How successful is the Recession Probability Model?

- The Recession Probability Model reached/exceed 50% in October 1969. A recession began in November 1969. Perfect timing. The model led the recession 1 month in advance
- The Recession Probability Model reached/exceed 50% in December 1973. A recession began in November 1973. This model was late by 1 month.
- The Recession Probability Model reached/exceed 50% in October 1979. A recession began in January 1980. Perfect timing. The model led the recession by 2 months in advance.
- The Recession Probability Model reached/exceed 50% in August 1981. A recession began in July 1981. This model was late by 1 month.
- The Recession Probability Model reached/exceed 50% in September 1989. A recession began in July 1990. The model was too early – it predicted the recession 10 months in advance.
- The Recession Probability Model reached/exceed 50% in February 2001. A recession began in March 2001. Perfect timing. The model led the recession 1 month in advance
- The Recession Probability Model reached/exceed 50% in February 2007. A recession began in December 2007. The model was too early – it predicted the recession 10 months in advance.

As you can see, my Recession Probability Model does a pretty good job at predicting U.S. economic recessions.

### Recession Probability Model updates

The real-time version of the Recession Probability Model will be updated once a week in the Membership Program. A 3 month delayed version of the Recession Probability Model will be available for free here on the blog.

As of early-September 2018, the Recession Probability Model is at 0%, which means that the probability of an imminent recession is extremely low.

Hi Troy.

What are the performances of this index if you use it to switch from UPRO/SSO to cash? Have you already tried this?

Hi Carlo, no I haven’t tried it yet.

But I have something better in the works: https://twitter.com/bullmarketsco/status/1039553424084361216

Do

Hi Troy

Your insights have been around 90% correct predicting market highs, small corrections etc, well done… one thing to watch with this recession is the market is looking 6 months into the future…

True