Print this article | Return to Article | Return to CFO.com

Predicting recessions is a tricky business.

Avital Louria Hahn, CFO Magazine

February 1, 2008

What are the odds that the United States will enter a recession? No, scratch that: What are the odds that the United States will enter a recession *next month*? According to New York Fed economist Arturo Estrella, the odds are 0.40965, give or take a millionth of a decimal point. That means there's a 40 percent chance the U.S. economy is about to enter a recession (40 percent has historically been the tipping point, although a recession may lag the month in which the 40 percent threshold is crossed).

Estrella is considered an expert on the "inverted yield curve" and its power to predict recessions. The term refers to the atypical shape of a graphed line that depicts rates on 3-month versus 10-year Treasury bills. Normally, longer terms offer higher rates; when short-term notes offer a higher yield, the curve arcs down instead of up and is said to be inverted. According to Estrella, a yield curve's predictive power is most accurate 12 months out. The probability assigned to a March 2008 recession, therefore, is based on data collected in March 2007, which Estrella massages according to a fairly simple formula.

There is plenty of disagreement about the predictive power of the yield curve, and just last month two of Estrella's colleagues at the New York Fed published an article that suggests that a recession is *not* in the offing. Not, at any rate, based on yield-curve trends. They analyzed two distinct components of a yield curve, interest-rate expectations and the term premium, and found that a model built only on the former may be more accurate. But the authors, Joshua Rosenberg and Samuel Maurer, say it's unclear whether the standard model or their revised version will prevail this year.