Comments on "Phillips Curve Inflation Forecasts" by James H. Stock and Mark W. Watson

Abstract

Stock and Watson provide a thorough (one might say exhaustive) review of the forecasting performance of many inflation models. These include models with some economics in them and others that are purely statistical. Overall, the best of the statistical models seems to be their unobserved components-stochastic volatility (UC-SV) model. This is a two-component model of the form