A0578
Title: Does uncertainty forecast recessions?
Authors: Michael Owyang - Federal Reserve Bank of St Louis (United States) [presenting]
Abstract: Increases in uncertainty tend to be highly correlated with declines in current and future economic activity. The literature on the real-time predictive ability of uncertainty for economic activity is smaller, but the findings are similar. It is not surprising, then, that incorporating policy uncertainty into binary outcome models has been argued to improve recession forecasts. A prior study evaluates whether the economic policy uncertainty index (EPU) and its components have predictive ability over other indicators such as the term spread and the growth rate of the S&P500. Another study uses the uncertainty index from a past study, performing their experiments in-sample. Alternative measures of uncertainty are evaluated as a predictor of business cycles using quasi-real-time experiments. A real-time estimate of the prior study's uncertainty measure is constructed for comparison with the real-time measure of another study. It is found that a combination measure of the two indices calibrated for business cycle prediction best forecasts future recessions.
