Two problems of the Taylor rule and a proposal: the tracking rule

Abstract

This paper deals with some problems related to the application of monetary policy following the Taylor Rule in the theoretical context of a “3-equation model”. The first problem arises if the real interest rate does not affect the equilibrium income level itself –as in the IS curve- but its rate of growth –as in the dynamic IS that we propose. Secondly, the Taylor Rule is incapable of reaching the inflation target when the central bank does not correctly estimate its parameters (the neutral interest rate and potential income) or these parameters vary. Our objective is to propose an alternative to the Taylor Rule which overcomes both problems. This alternative has been called the Tracking Rule, because instead of trying to estimate the neutral interest rate or the potential output, the central bank “tracks” these values based on the economy’s evolution, particularly on variations in the inflation and unemployment rates. After justifying the dynamic IS and explaining the logic of this rule in detail, the paper compares the Tracking Rule with the Taylor Rule, simulating both of them in the context of different types of shock in the modified three equation model. The results, measured by a loss function, show that the Tracking Rule is superior in every single case. It is particularly interesting to evaluate central bank reactions derived from the two rules when the economy suffers a large contractive shock such as the current crisis. The results show that, with the same shock, the economy is more likely to fall into the liquidity trap when the Taylor Rule is applied.