국내증시 일별 거래량의 자기상관성에 대한 고찰

This study analyzed whether the transaction volume cluster occurred in the Korean stock market. The cluster phenomenon of daily trading volume means that an abnormal surge lasts at least a short period of time if the volume of trading increases or decreases sharply compared to the average volume of trading. Based on the daily transaction volume time series data that went through the stationarity process of the daily time series data, the results showed that the daily transaction volume had time series autocorrelation at a significant level. This cluster of transactions is more pronounced with the higher the net share of individual investors or the lower the net share of institutional purchases, and it is concluded that the cluster trading pattern of individual investors affects daily trading volume autocorrelation. In addition, the cluster of transactions was more evident on the business day when the disclosure was announced, indicating that the arrival of new information in the stock market further enhances the auto-regression of daily transactions. In addition, institutional investors were more sensitive to the arrival of new information disclosed by companies than individuals. The greater the daily stock price fluctuation in the relationship between autocorrelation of daily trading volume and volatility of stock prices, the higher the degree of auto-regression in the daily trading volume time series. The stock return, which was found to be highly related to the increase and decrease in trading volume, is judged to have a limited impact on the daily trading volumen cluster in the Korean stock market.