Forecasting Turning Points

This is more of an FYI post.

There are macroeconomists out there forecasting business cycle turning points, in real time, with a great deal more sophistication than the leading indicators that you hear about in the legacy media.

First some tips:

• There are others besides these, but these are some of the big ones.
• By default, each index may show a different time length along the horizontal axis. Check this and make sure you’re looking at what you think you want.
• Most of the indices show recessions as upward spikes (but not all). The threshold for a turning point may be different for each one too.
• Most of these come up with a probability that we are in recession currently.
• Recessions are caused by different things (I refer to this as long lists of pros and cons), all of them small, and some of them hard to observe. There is a strong tendency to assume that the next recession will be caused by the same thing as the last recession. This is a serious problem because this almost never happens.
• The last point is made worse by the tendency of humans to explain things with one big cause. If it were that easy, recessions wouldn’t be a problem.

Here are some links to forecasts that focus on financial conditions (because that’s what was important last time around):

• Chicago Fed National Financial Conditions Index (hosted by the St. Louis Fed)
• Kansas City Financial Stress Index (you need to click the link for the PDF of the current release)
• St. Louis Fed Financial Stress Index
• Bloomberg United States Financial Conditions Index 
• CBOE Volatility Index

Here are some links to more general indices of business conditions:

• Aruoba-Diebold-Scotti Business Conditions Index from the Federal Reserve Bank of Philadelphia.
• Smoothed U.S. Recession Probabilities, hosted by Jeremy Piger (based on the work of Marcelle Chauvet).

I tend to think the last one is the best. However, it’s an academic site rather than a commercial product, so you may need to poke around a bit to find the file/chart that’s been most recently updated (try the Figure and Excel links at the top).

FWIW: This sort of econometrics goes back to the pioneering work that James Hamilton (who is coauthor of the blog Econbrowser) published in Econometrica in 1989. Personally, I think Hamilton is on the medium-list for a Nobel Prize sometime in the next 20 years for this work.