Some Empirical Evidence on Models of the Fisher Relation: Post-Data Comparison

The Fisher relation, describing a one-for-one relation between the nominal interest rate and the expected inflation, underlies many important results in economics and finance. Although it is a conceptually simple relation, the Fisher relation has more or less complicated with mixed results. There are several alternative models proposed in the empirical literature for the Fisher relation that have different implications. We evaluate those alternative models for the Fisher relation based on a post-data model determination method. Our results for data from the U.S. Japan and Korea show that models with both regimes/periods, a regime with nonstationary fluctuations and the other with stationary fluctuations, fit data best for the Fisher relation.