Taylor rule estimation with the presence of a ZLB-period : how the inclusion of shadow rate affect the precision of Taylor rule estimation on the federal funds rate.

Abstract

This thesis estimates monetary policy reaction functions for the United States’ economy from 1987 until 2015 using a Taylor rule. The period 2009-2015 was characterized by the federal funds rate being bounded below by zero, commonly known as the zero lower bound (ZLB). To measure the effect of monetary policy during this period, a shadow rate was proposed. A shadow rate is an estimated, theoretical interest rate not bounded by the ZLB. We compare estimations of the Taylor rule on out-of-sample data using both the federal funds rate bounded below by zero and the federal funds rate with a shadow rate. We find that the Taylor rule estimations conducted with the shadow rate fits closer to the out-of-sample effective federal funds rate, than standard estimations that includes the data bounded below by zero. Our thesis suggests that shadow rate can be used as a tool to analyze monetary policy in a Taylor rule setting, using data with the presence of a ZLB-period. Furthermore, our results suggests that there is value in including shadow rate in macroeconomic models for monetary policy.