In his article, "Vector Autoregressions as a Tool for Forecast Evaluation," Roy H. Webb proposes that VAR forecasts be used as a standard of comparison for other forecasts. He begins by explaining how conventional forecasting models are constructed and used, and summarizes a few common objections to these models. He then describes the VAR methodology and compares forecasts from a simple VAR model with those from a consulting firm that uses a conventional model and with a series of consensus forecasts. The VAR model holds its own in this competition; in fact, only the VAR model is able to predict the 1981-1982 recession one year before its occurrence.
Amanda L. Kramer