Exchange Rate Volatility Models: The Turkish Case

Abstract

Exchange rate volatility models: The Turkish case This study aims to model the exchange rate volatility for Turkish economy with use of ARCH, GARCH, and SWARCH models for the period of July 2001-May 2010 with daily export-weighted exchange rate data. We find that SWARCH is the best exchange rate volatility model compared to the traditional models on the basis of the econometric selection criteria, AIC and SC. It is also found that this model accurately predicts the actual exchange rate volatility 1, 4, 8, 15 and 30 days (periods) ahead. The volatility predictions obtained from SWARCH coincides exactly with the actual exchange rate volatilities observed in Turkish economy, including the extreme exchange rate volatility in the year 2008 of global financial crisis. As a result, the study proves that SWARCH model has a better volatility predictions compare to the traditional models. It will be a particularly useful model to apply for policy makers who like to see the exchange rate fluctuations right on time and even periods ahead in their macroeconometric modeling of the Turkish Economy.