Is there a reliable relationship between U.S. employment and the U.S. stock market? In their August 2009 paper entitled “The Stock Market and Aggregate Employment”, Long Chen and Lu Zhang study the interactions between the stock market and the labor market. Using quarterly returns for the S&P 500 Index and quarterly data for employment and other economic indicators over the period 1952-2007, they find that:
- Lagged payroll growth relates negatively to future excess (relative to the risk-free rate) stock market returns. In other words, an increase (decrease) in employment indicates future stock market weakness (strength).
- Adjusted R-squared (a measure of explanatory power) peaks at 9% at the four-quarter forecast horizon.
- The relationship is consistently negative, significant for forecast horizons up to eight quarters, and marginally significant at longer horizons.
- The power of payroll growth to predict stock market returns:
- Dominates that of the default spread and the relative Treasury bill yield (three-month Treasury bill yield minus its four-quarter moving average).
- Is shorter-term than that of the dividend yield and the term spread, which exhibit peak predictive power at 12 and 16 quarters, respectively.
- Is stronger than that of change in the unemployment rate.
Note that conclusions such as “strong” and “significant” in a statistical sense often do not translate into “easy pickings” in an investing context, because the preponderance of the variation in stock returns remains unexplained. Also, non-normality of the return distribution may disrupt interpretation of statistics such as R-squared.
In summary, evidence indicates that employment growth relates negatively to future stock market behavior, most strongly at a one-year forecast horizon.
