Using prior bias to improve forecast accuracy

Abstract

The study tests whether removing bias can improve out-of-sample forecast accuracy in two series of interest rates. The samples are larger than previously studied in this context, and the test is cleaner since reported interest rates are never revised.