Learning the Taylor Rule: Interest-Rate Predators and Inflation Prey

The Taylor rule is an influential monetary policy rule that has been used to explain central bank behavior. Although the rule has been studied extensively over the years, most discussion has centered around its ability to explain central bank behavior, primarily in backward-looking applications focused on data fitting. Although such applications are of interest to investors, they are arguably of less interest than understanding the degree to which the original rule or variants of it can aid in interest rate forecasting. To that end, in this article, the author provides a variation of the Taylor rule that adapts the well-known Lotka–Volterra equations to interest rate forecasting. He finds that several variations of the Lotka-Volterra interest rate model outperform the original Taylor rule and a rudimentary extension of it, as well as basic versions of two popular machine learning models, in their ability to effectively forecast short- and medium-term interest rates in the United States and Eurozone in cross-validations using both historical and bootstrapped data.