A0261
Title: Trend-cycle decomposition and forecasting using Bayesian multivariate unobserved components
Authors: Mohammad Jahan-Parvar - Federal Reserve Baord of Governors (United States) [presenting]
Charles Knipp - Federal Reserve Board (United States)
Pawel Szerszen - Federal Reserve Board of Governors (United States)
Abstract: A generalized multivariate unobserved components model is proposed to decompose macroeconomic data into trend and cyclical components. The series is then forecasted using Bayesian methods. It is documented that a fully Bayesian estimation, which accounts for state and parameter uncertainty, consistently dominates out-of-sample forecasts produced by alternative multivariate and univariate models. In addition, allowing for stochastic volatility components in variables improves forecasts. To address data limitations, cross-sectional information is exploited, the commonalities across variables are used, and both parameter and state uncertainty are accounted for. Finally, it is found that an optimally pooled univariate model outperforms individual univariate specifications and performs generally closer to the benchmark model.
