Estimates of instrumental rules can be utilized to describe central bank's behavior and monetary policy stance. In the last decade, considerable attention has been given to time-varying parameter (TVP) specification of monetary policy rules. Constant-parameter reaction functions likely ignore the impact of model uncertainty, shifting preferences and nonlinearities of policymaker's choices. This paper examines the evolution of monetary policy reaction function in Iran via estimating a time-varying parameter (TVP) specification in the 1990:2-2014:4 period. We try to find out whether there is a significant time variation in coefficient of CBI (the Central Bank of Iran) reaction function. The main findings are threefold. First, monetary policy rules changed over time, hence making relevant the application of a time-varying estimation framework. Second, the monetary instrument smoothing parameter is much lower than typically reported by previous time-invariant estimates of policy rules. Third, CBI does not systematically follow instrumental rule to fight inflation. During the whole sample, there is no quarter in which the inflation gap coefficient is negative and significant; therefore, monetary policy has not counteracted inflationary pressures. Key words: Monetary policy, Instrumental rule, monetary policy reaction function, Time varying coefficient, Extended Kalman filter. JEL Classification: E4, E5
