비선형 모형을 이용한 원/달러 환율 변동성이 KOSPI지수 수익률에 미치는 영향 분석

The purpose of this study is to investigate the nonlinear relationship between KOSPI index returns and volatilities of exchange rate in Korea. We adopt the logistic smooth transition autoregressive (LSTAR) model and use exchange rate volatility as the transition variable. To do this, we utilize the daily KOSPI index and Won/Dollar exchange rate, while the sample period ranges from January 2, 1999 to September 29, 2017. In this paper, the EGARCH model is used to estimate exchange rate volatility due to the asymmetry of volatility. We find from the estimation results of nonlinearity that nonlinear models are more appropriate than linear models to capture the dynamic properties of KOSPI index returns. The empirical result of the LSTAR model shows that the long-run mean of KOSPI index returns is positive (negative) in the low (high) exchange rate volatilities regime, implying that the effect of exchange rate volatility on KOSPI index highly depends on their magnitude. In addition, the result of the impulse response indicates that the impact of exchange rate shock relies on the state of the economy. This result implies that policy makers should be more concerned about minimizing the level of exchange rate voltatility during economic downturns such as the recent global financial crisis.