Article ID: | iaor201526523 |
Volume: | 34 |
Issue: | 4 |
Start Page Number: | 241 |
End Page Number: | 260 |
Publication Date: | Jul 2015 |
Journal: | Journal of Forecasting |
Authors: | Ielpo Florian |
Keywords: | forecasting: applications, finance & banking |
The short end of the yield curve incorporates essential information to forecast central banks' decisions, but in a biased manner. This article proposes a new method to forecast the Fed and the European Central Bank's decision rate by correcting the swap rates for their cyclical economic premium, using an affine term structure model. The corrected yields offer a higher out‐of‐sample forecasting power than the yields themselves. They also deliver forecasts that are either comparable or better than those obtained with a factor‐augmented vector autoregressive model, underlining the fact that yields are likely to contain at least as much information regarding monetary policy as a dataset composed of economic data series.