Stock correlation analysis based on complex network

Aim at the study of the relationship of correlation between the stocks of SSE(Shanghai Stock Exchange) companies and the movement of SSE complex index, by using the mathematical statistics method and correlation coefficient analysis, select each stock's daily closing price as object, establish the complex network of stocks. Through preprocessing the data sets, we used R Studio get the mean correlation coefficient matrix, which is the foundation of the financial network. The results confirm the utility of the correlation coefficient analysis as a stability indicator for SSE stock market, since it proves that the correlation of stocks is different according to the crisis and prosperous of stock market in times. Its development is opposite to the market index, the interaction between the stocks attenuates when the complex index rises and enhances when the index falls. We also used K-means clustering algorithm to classify these nodes into different communities and found that the structures of networks vary widely according to their correlations. The method of this paper and the model it proposed is not only for the selection of our data sets, but also can be generalized to other fields of research.