Determinants of the monetary policy reaction function: An application for Turkey

After the 2008 Global Financial Crisis (GFC), it has become very important for central banks to aim not only price stability but also financial stability. For that reason, the policy makers have had to reconsider the monetary policy reaction function and its variables. The aim of this dissertation is to determine the variables of the monetary policy reaction function and then by using it to examine monetary policy reaction to the changes in its variables in Turkey after the 2008 GFC. To this end, the monetary policy reaction function of the Central Bank of the Republic of Turkey (CBRT) has been modelled. In this model, the spread between overnight borrowing and lending rates of the CBRT has been used as a dependent variable. Based on smooth transition regression (STR), the methodology proposed by Leybourne et al. (1998) is used to detrend the spread series to make it trend-stationary. This methodology is the contribution of this dissertation to the related literature. Thus, the parameters obtained by the model estimation explicitly can be used to capture the effects of the changes in the variables of the monetary policy reaction function on the dependent variable. The estimation results suggest that Turkey's monetary policy reaction function is a Taylor-type which is both forward and backward-looking (in hybrid structure) for the period between 2009:01 and 2016:12. In addition, the results of this thesis indicate that the variables of the monetary policy reaction function are the spread, the lag of spread, the actual and the expected inflation / output gaps, the actual change in the real effective exchange rate and the expected change in the gross foreign exchange reserves of the CBRT. In conlucison, this thesis provides monetary policymakers with a usable policy guide.