More predictors of the investment opportunity set in the ICAPM

Abstract

To derive the discrete-time version of the Intertemporal CAPM (ICAPM), Campbell (1993) summarizes the investment opportunity set determining investor’s utility as the return to aggregate wealth, and uses an aggregate stock market return to proxy for the aggregate wealth return. In doing so, state variables can be identified as the variables that have forecasting power for future stock market returns. If the stock market return is a true sufficient statistic for investment opportunities, then we should expect any predictor of stock market returns would command a significant risk premium in the cross-section. To verify his theory, we employ 40 predictors of stock market returns from the literature, and see if the risks to these predictors are well priced. We find that exposures to 27 out of the 40 predictors generate significant risk premiums. Given the noise components in the predictors, this finding suggests that the Campbell’s (1993) ICAPM choosing the stock market return as a representative measure for investment opportunities is overall satisfactory.