유가와 주가지수 사이의 비선형관계 분석

The purpose of this study is to investigate the nonlinear relationship between stock indices and oil prices in Korea. To do this, we adopt a logistic smooth transition autoregressive(LSTAR) model and a quantile regression model. To examine our purpose, we utilize the weekly KOSPI index, three market cap-index(large, medium and small stocks) and WTI prices, while the sample period ranges from January 4, 2001 to December 13, 2018. We find from the estimation results of three nonlinearity tests that nonlinear models are more appropriate than linear models to capture the dynamic properties of four index returns. The empirical result of the LSTAR model shows that the long-run mean of four index returns is positive (negative) in the high(low) oil price change regime, implying that the effect of the changes in oil prices on four indices highly depends on their magnitude. This suggests that investors interpret falling oil prices as bad news because they expect falling economic activity, while rising oil prices can be considered as a good signal as a prospering economy. In addition, the result of the quantile regression analysis indicates that the positive and negative changes in oil prices has a different effect on four stock indices, implying that it is wise to take the fluctuation of oil prices into account when structuring portfolios and diversification strategies since they produce significant effects on the stock markets regardless of the market capitalization of individual stocks.