A VAR Framework of Exchange Rates, Interest Rates, and Inflation Through COVID-19 in Turkey: Empirical Evidence From Linear Cointegration and Causality Analysis

In this study, we investigate the impact of exchange rates and interest rates on inflation in Turkey using monthly data from January 2004 to July 2020 obtained from the Turkish Statistical Institute. Our vector autoregressive (VAR) model showed evidence of stochastic behaviour among the series. The autoregressive distributed lag (ARDL) results showed a short-run and long-run covariate relationship between exchange rates, interest rates, and inflation. Specifically, the findings showed a short-run and long-run relationship between inflation, the producer price index (PPI), and Turkish interbank offer rates (TIBOR). However, there was no connection between inflation and dollar exchange rates (DSR) or commercial banks’ interest rates (CBIR). The VAR Granger causality results revealed the variables to be exogenous except for DSR, which displayed endogeneity to other variables. Nevertheless, the results revealed unidirectional causality from the producer price index (PPI) to DSR and unidirectional causation from TIBOR to CBIR. This means an increase in production costs through raw materials importation led to the devaluation of the Turkish lira. Similarly, TIBOR rates drive CBIR higher, making domestic lending more expensive, which will inhibit loan provisions to the private sectors, resulting in an economic contraction and eventually high inflation. Our unit roots breakpoint results pointed to breaks in the dataset between 2016-2019, reflecting the effects of the Fetullah Terrorist Organization (FETO) failed coup, the 2018 U.S. embargo, and an assumed fiscal dominance as the major and direct causes of economic instability and inflation. However, Covid-19 may have acted as a contributing factor since then. Thus, we recommend that the monetary authorities articulate policy to avoid the assumed fiscal dominance.