What does an inverted yield curve mean?

This is one of those headache topics.  Daniel Gross presents a clear treatment:

...the yield curve...describes the relationship between interest rates on long-term and short-term U.S. government bonds. Interest rates on the shortest-term bonds correlate very closely with the interest rates set by the Federal Reserve Board. Long-term interest rates, by contrast, are influenced by many more factors, ranging from China's purchase of debt to investors' optimism about inflation and growth. Typically, bonds that mature further in the future pay higher yields—compensation for the risk of locking up money for a longer period.

The yield curve rarely inverts. And when it does, it usually spells trouble for the economy. It means that investors and the Federal Reserve are fretting about inflation in the short term, and that investors are pessimistic about long-term growth. According to Brian Reynolds, chief market strategist at MS Howells & Co., in the last 30 years, periods of prolonged inversion of the curve between two-year and 10-year government bonds have generally presaged recessions. The most recent period of inversion ran from February 2000 through December 2000—just before the 2001 recession.

A year ago, the yield curve was rather steep. But in the last year, the Federal Reserve Open Market Committee has taken the short-term Federal Funds rate from 1 percent to 3 percent in eight straight tightenings, the most recent one in May. (All the Fed's 2005 actions can be seen here.) Today, with two-year bonds at about 3.5 percent and the 10-year bond having fallen to about 3.9 percent, only a few dozen basis points separate the two.

Gross is an excellent economic journalist but I must differ on one key point.  The yield curve is overrated as a predictor of future output.  Here is a more cautionary and accurate analysis:

...the 1985-95 sub-sample completely reverses the results.  The yield spread becomes the least accurate forecast, and adding it to lagged GDP actually worsens the fit.

Another recent study shows that the short rate, not the yield spread, holds most of the relevant predictive power.

The bottom line: The previous power of the inverted yield curve was based on a few good predictions.  But no such predictor will stand up over time.  First, asset prices are very noisy.  Second, knowledge that we had an accurate predictor would itself change the relationship we are trying to predict. 

We face some serious economic problems today; savings may be taking the wrong form (capital gains rather than income reallocation), and perhaps we are in a housing bubble.  But observed spread in the term structure of interest rates does not add to my worries.