Equal weighted forecast averaging is a benchmark that has been found to provide
good forecasts of inflation (and of many other variables). Stock and Watson (2001,
2002a) indeed argue that it is the best method for predicting inflation in the US and other
G-7 countries among a wide range of forecasting methods that they consider. So, since
Bayesian Model Averaging does better than equal weighted averaging in predicting US
inflation, it should be taken very seriously as a method for forecasting inflation. I do not
mean to claim by this that Bayesian Model Averaging as I have implemented it in this
paper is necessarily the best thing that a researcher could ever do. It may be possible to
get still better forecasts by incorporating nonlinear models in the exercise, by incorporating Greenbook and private sector survey forecasts of inflation, by considering
models with more than one predictor, or by using different shrinkage techniques.