Working Paper: NBER ID: w12324
Authors: James H. Stock; Mark W. Watson
Abstract: Forecasts of the rate of price inflation play a central role in the formulation of monetary policy, and forecasting inflation is a key job for economists at the Federal Reserve Board. This paper examines whether this job has become harder and, to the extent that it has, what changes in the inflation process have made it so. The main finding is that the univariate inflation process is well described by an unobserved component trend-cycle model with stochastic volatility or, equivalently, an integrated moving average process with time-varying parameters; this model explains a variety of recent univariate inflation forecasting puzzles. It appears currently to be difficult for multivariate forecasts to improve on forecasts made using this time-varying univariate model.
Keywords: inflation forecasting; univariate models; multivariate models
JEL Codes: C53; E37
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| univariate inflation process (C29) | characterized by a permanent stochastic trend component and a serially uncorrelated transitory component (C22) |
| variance of the permanent disturbance in inflation (E31) | predictability of inflation (E31) |
| time-varying trend-cycle model (C22) | decline in the root mean squared error (RMSE) of naive inflation forecasts (E31) |
| performance of multivariate forecasting models (C53) | deterioration relative to univariate models (C29) |
| fixed-parameter models (C20) | less effective in future forecasting (C53) |