Determining structural breaks in central bank reaction functions of the financial crisis

Abstract

Determining breakpoints in central bank reaction functions is crucial to monitor the decisions of the governing councils correctly. This is especially true during the recent financial crisis. However, reaction functions do not necessarily change at the beginning of a crisis and so the specific breakpoints need to be estimated rather than being set exogenously. While central bank reaction functions typically include more than one exogenous variable it is possible that the breakpoints in the reaction coefficients change from one independent variable to the other. We develop an estimation strategy using real-time data that account explicitly for variable breakpoints in the different independent variables. Using Taylor reaction functions for the ECB and the Fed we show that there are indeed different breakpoints with respect to the lagged interest rate, the inflation rate and the output gap within the financial crisis, in which the reaction coefficients tend to be significantly lower. Out-of-sample forecasts underline the result of a less volatile reaction of central banks during the crisis period.