Nonlinear Properties of Stock Price and Trading Volume

In this paper, we investigate the nonlinear properties in the dynamic process of the daily changes of stock price and trading volume in the Korean stock market by using the STAR class models. The estimation results indicate that the ESTAR model is appropriate for both the time series of price returns and volume changes, and this model is capable of capturing a substantial portion of asymmetric movements of the variables. Examining the local dynamics of variables by analyzing characteristic roots, we found that once both variables are in a specific regime, they are more likely to stay there for a period of time. Lastly, this study tests nonlinear Granger causality. Most of our test results show bidirectional nonlinear Granger causality between these two variables.