Forecasts in a Slightly Misspecified Finite Order VAR

52 Pages Posted: 24 Jan 2011 Last revised: 4 Mar 2022

See all articles by Ulrich K. Müller

Ulrich K. Müller

Princeton University - Department of Economics

James H. Stock

Harvard University - Department of Economics; National Bureau of Economic Research (NBER); Harvard University - Harvard Kennedy School (HKS)

Date Written: January 2011

Abstract

We propose a Bayesian procedure for exploiting small, possibly long-lag linear predictability in the innovations of a finite order autoregression. We model the innovations as having a log-spectral density that is a continuous mean-zero Gaussian process of order 1/√T. This local embedding makes the problem asymptotically a normal-normal Bayes problem, resulting in closed-form solutions for the best forecast. When applied to data on 132 U.S. monthly macroeconomic time series, the method is found to improve upon autoregressive forecasts by an amount consistent with the theoretical and Monte Carlo calculations.

Suggested Citation

Müller, Ulrich K. and Stock, James H., Forecasts in a Slightly Misspecified Finite Order VAR (January 2011). NBER Working Paper No. w16714, Available at SSRN: https://ssrn.com/abstract=1744668

Ulrich K. Müller (Contact Author)

Princeton University - Department of Economics ( email )

Princeton, NJ 08544-1021
United States
609-258-3216 (Phone)
609-258-4026 (Fax)

HOME PAGE: http://www.princeton.edu/~umueller

James H. Stock

Harvard University - Department of Economics ( email )

Littauer Center
Cambridge, MA 02138
United States
617-496-0502 (Phone)
617-496-5960 (Fax)

National Bureau of Economic Research (NBER)

1050 Massachusetts Avenue
Cambridge, MA 02138
United States

Harvard University - Harvard Kennedy School (HKS) ( email )

79 John F. Kennedy Street
Cambridge, MA 02138
United States

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
16
Abstract Views
383
PlumX Metrics