Dr. Halstead's Econometric Model

Dr. Halstead uses his econometric model to try to determine if the U.S. economy is in a recession or is not in a recession. Items such as economics news stories or comments from economists have no impact on his recession/non-recession signals. His objective econometric model uses data to generate the recession/non-recession signals.

You may wonder what specific data Dr. Halstead’s econometric model uses to generate the recession/non-recession signals. First off, trying to determine these recession starting and ending points is extremely difficult, and one must never become overconfident in believing one can consistently foresee such turning points. Secondly, Dr. Halstead is proponent of the World War II U.S. slogan: “Loose lips, sink ships”. That is, at this time, Dr. Halstead is concerned that if he disclosed the specific inputs into his econometric model, the econometric model would lose its effectiveness. Did you realize that Coca-Cola’s precise beverage formula is still a secret over 100 years later to everyone but maybe just a few individuals? Dr. Halstead, though, will disclose that his econometric model doesn’t use any of the factors from the popular Sahm Rule recession indicator created by Dr. Claudia Sahm or factor in the shape of the U.S. Treasury yield curve as popularized by Duke’s Dr. Campbell Harvey. Just about each business day, economic data are released from different organizations, including the U.S. government. Dr. Halstead has found that some data are more important than others. Dr. Halstead’s model was backtested, but has been “live” for some time. Dr. Halstead didn’t use artificial intelligence to create the econometric model.

Here's a graph of Dr. Halstead’s econometric model from 7/1/1997 to 7/2/2025:

The red line is the model, the gray shaded areas indicate the NBER recessions, and the blue line is the 50% signal trigger line. The results speak for themselves. Do you recall the rampant discussion by some in 2022 that a U.S. recession was imminent? That “imminent” U.S. recession never occurred. As you can see in the above graph, Dr. Halstead’s econometric model was nowhere close to signaling a U.S. recession during 2022. Dr. Halstead’s econometric model rarely gives a recession signal. That’s a good feature of the model since U.S. recessions are rare events.

Here are the general interpretations of Dr. Halstead’s econometric model:

If there isn't already a U.S. recession start signal triggered, a probability reading of under 50% from Dr. Halstead’s econometric model indicates that the U.S. economy is likely not currently in a recession and likely not to be in a recession within about the next 1 to 3 months.
A 50% or higher probability reading from Dr. Halstead’s econometric model indicates that the U.S. economy is likely already in a recession or is likely to enter into a recession within about the next 1 to 3 months. That recession signal would remain in effect until Dr. Halstead’s econometric model generated a probability reading under 10%.
If there is already a recession start signal triggered, an under 10% probability reading from Dr. Halstead’s econometric model indicates that the U.S. economy is likely not currently in a recession or likely not to be in a recession within about the next 1 to 3 months.

How is Dr. Halstead’s econometric model different from other recession signaling approaches? The economic probabilities coming from Dr. Halstead’s model come out just about every business day whereas some other approaches may only give signals on a monthly basis. Dr. Halstead believes that monthly models don’t provide enough timing precision for recession signaling since one would have to wait all the way until after the month is over for a monthly signal to be possibly generated. Some other recession signalers appear to rely on a non-quantitative, subjective approach toward recession signals. Dr. Halstead prefers to employ a more scientific quantitative approach.