Industrial Production as a Predictor of Stock Returns

Does any broad measure of the state of the economy meaningfully predict financial market returns? In their May 2008 paper entitled “Time-Varying Risk Premia and the Output Gap”, Ilan Cooper and Richard Priestley investigate the output gap as a direct link between future stock returns and economic fundamentals. They define output gap as the deviation of the log of industrial production from a trend constructed from both linear and quadratic components. Using unrevised industrial production data, aggregate U.S. stock market returns and Treasury bill yields (to calculate excess returns) for the period 1948-2005, they conclude that:

• The output gap significantly predicts U.S. stock returns, most effectively at business cycle forecast horizons, both in-sample and out-of-sample.
  • The adjusted R-squared statistics for the historical relationship between output gap and U.S. stock returns are 0.02, 0.05 and 0.11 at monthly, quarterly and annual forecast horizons, respectively.
  • A one standard deviation decline in the output gap increases predicted annual excess U.S. stock returns by about 5%.
• The output gap has predictive power for excess stock returns at business cycle and longer frequencies in most other G7 countries over the period 1970-2005.
• The output gap has predictive power for excess U.S. government bond returns over the period mid-1952 through 2003.

The following chart, taken from the paper, compares expected excess return (ER) for the broad U.S. stock market, as predicted by the output gap, to the actual excess return at a quarterly frequency. The predicted return series is much less volatile than actual returns.

In summary, while of little value to traders, the industrial production (output) gap may have some meaningful predictive power for broad U.S. stock returns over relatively long periods.