Abstract
In this paper we construct a proxy for the systematic component of the risk structure of corporate yields (or systematic risk structure), and test how well it predicts real economic activity. We find that the systematic corporate risk structure predicts the growth rate of industrial production 3 to 18 months into the future even when other leading indicators are controlled for. It produces more accurate out-of-sample forecasts than the model using the index of leading indicators, the autoregressive model, and the random walk model over a variety of time periods. Moreover, the predictive power of the systematic risk structure, unlike that of the other forecasting variables, has been robust over the last fifteen years. Finally, a regime-switching estimation shows that the systematic risk structure is very successful in identifying and capturing different growth regimes of industrial production. It outperforms the Index of Leading Economic Indicators in doing this. In particular, it was able to correctly predict the timing of the last slowdown in the U.S. - a daunting task for any model after a decade-long period of positive growth.