한국주식시장의 주식 네트워크 속성에 대한 실증연구: 수익률생성모형과 Random Matrix Theory를 이용

The purpose of this study is to investigate whether the stock returns estimated by theoretical models could explain the interaction between stocks in the Korean stock market. Through this investigation, we might find the deterministic factors that significantly affect the formation process of the interactive relationship between stocks, using the stock network proposed by Mantegna(1999). We used the multi-factor model and the random matrix theory method. We have discovered empirically that the process of formation in stock network is affected by common factors. Furthermore, we found that the stocks with a large number of links with other stocks are affected more by the common factors. Also, we found that the interactions between stocks can not be explained sufficiently if the properties of the market index are removed. Thus, the property of the market index plays an important role in the process of formation of the stock networks. These results suggest that some stocks that act as a hub for clusters are affected by the common factors, especially more by the market index, while some stocks with a small number of links are affected less by the common factors. Therefore, the common factors in the stock market are important deterministic factors of the interactions between stocks.