Cointegration and causality relationship between Borsa Istanbul’s BIST100 and exchange rates

Abstract

This thesis investigates the relation between the stock prices and the exchange rates(against US dollar) of Turkey, based on the monthly data from January 2003 to February 2019.The study inspects the short-run causal relationship by employing multivariate Granger causality test, and probes the long-run relation of the aforementioned variables using Johansen cointegration test. In addition, by use of multivariate DCC-GARCH method, dynamic conditional correlation among variables are estimated & forecasted for future periods. Empirical results suggest that only one long-term relation exists between these variables, and the direction of causality runs from stock prices to exchange rates. Findings of DCC-GARCH implies that significant dynamic conditional correlation is present among the time series studied. Also, a considerable time varying correlation is established between BIST100 stock prices and real exchange rates, which is in accordance to the result of preceding methods.