Tuesday, January 26, 2016

More On Forecasting a Peak

Stock markets are not the only thing that delivers poor forecasts of recessions. Check out this chart:

WSJ 16-01-25 False Positives and Negatives

What these three charts show is how off its most recent maximum value each of the series is. So if a series is growing steadily, setting a new maximum each month when the new data comes out, then the curve will be pegged at zero. But if the series continues to decline off its peak, it will get lower and lower.

These charts also shade the postwar recessions.

So, what do we see. First off, all three series are off their most recent peaks.

Perhaps more importantly, none of these series appears to go down much before the business cycle peak (the left side of each shaded band). That implies, if you think these are all good variables for forecasting, that we’re already in a recession. Does it feel that way to you?

Having said that, and repeated my warnings about false positives and false negatives, the evidence here doesn’t look very good.

  • The stock price drop is a positive signal (that we are in recession). There have only been 4 large false positives since World War II.
  • Profits perform a bit worse: there have been 10 false positives since World War II. But most of those occurred when the stock market did not confirm the signal.
  • The drop in industrial production is small enough that I’m not going to say it’s sending out a positive signal. But, we just started peaking for this expansion, and we’re seeing a dropoff at the same time that corporate profits are falling. That’s not good.

This was drawn from an article entitled “Recession Warnings May Not Come to Pass” that appeared in the January 25 issue of The Wall Street Journal. Read the whole thing.

No comments:

Post a Comment