A0331
Title: The distributional predictive content of inflation expectations measures
Authors: Saeed Zaman - Federal Reserve Bank of Cleveland (United States) [presenting]
James Mitchell - Federal Reserve Bank of Cleveland (United States)
Abstract: The predictive relationship between the full distribution of future inflation and inflation expectations measures from households, firms, markets, and professional forecasters is examined. The focus is on the short-term expectations measures but also considers the marginal value of the long-term expectations. To allow for nonlinearities in the predictive relationship, quantile regression methods are used. The key findings are as follows. First, the ability of household expectations to predict future inflation, relative to professionals and the market, increases with inflation. Some households are better than others and the disagreement across household respondents has useful predictive content for turning points. Second, incorporating information about long-term inflation expectations further improves the accuracy in predicting future inflation. Third, the joint predictive ability of all three inflation expectations measures is greater than the marginal predictive content of each of the measures. However, results emphasize the importance of tracking and differentially weighting different agents' expectations of inflation when assessing inflationary pressures in a probabilistic sense. In other words, all expectations are equal, but some are more equal than others.
