Abstract:
The paper analyses the performance of simple interest rate rules which feature a response to noisy observations of inflation, output and money growth. The analysis is based on a small empirical model of the hybrid New Keynesian type which has been estimated on euro area data by Stracca (2007). To assess the magnitude of the measurement problems regarding the feedback variables, we draw upon the real-time data set for Germany compiled by Gerberding et al. (2004). We find that interest rate rules which include a response to money growth outperform both Taylor-type rules and speed limit policies once real-time output gap uncertainty is accounted for. One reason is that targeting money growth introduces history dependence into the policy rule which is desirable when private agents are forward-looking. The second reason is that money growth contains information on the "true" growth rate of output which can only be measured imperfectly.