Asymmetric Monetary Policy Expectations

Abstract

We document novel empirical evidence of significant time-varying skewness in the aggregate forecast distribution of the federal funds rate using the Survey of Primary Dealers. We show a New-Keynesian model with the effective lower bound can endogenously generate positive and negative skewness observed in the data. Time-varying skewness has important implications for the measurement of policy expectations. In particular, the Blue Chip Survey (BCS) forecasts track the mode more closely than the mean. The mean measure implies significantly less negative term premia on average and also outperforms the BCS forecasts based on the mean squared error, consistent with optimal forecasting.

Keywords: Skewness, Federal Funds Rate Expectations, Survey of Primary Dealers, Blue Chip Survey, Monetary Policy, DSGE

JEL Classification: E43, E44, E52, G17