Industry Volatility Spillover Effect Of Chinese Stock Market Based On Independent Component Analysis

This thesis aims to give an analysis of China’s stock market industry volatility mechanism spillover effect and it focuses on the data of five sectors of China’s market i.e. electronics manufacturing, agriculture and services, environmental engineering, communication services and logistics from August 2007 to April 2016.Based on the mathematical statistics, this thesis tests the characteristics of non-normal distribution of data as well as construct GARCH volatility model to fit these five sectors. Through parameter estimation and hypothesis test, this thesis tests spillover effect of agricultural services, environmental engineering, communications and logistics services sectors to electronics manufacturing industry. This thesis divides agricultural services, environmental engineering, communications and logistics services four sectors into separated components through PCA analysis. At the same time, this thesis deals with GARCH volatility of algorithm(FastIca) processing of agricultural services, environmental engineering, communication services and logistics of these four industries, reduce the dimension of volatility sequence data, and handle independent component. At last, the results shows that there is a significant positive correlation between volatility spillover effect and validate the volatility spillover in Chinese stock market through parameter estimation and hypothesis testing for unrelated components and separated components.Through identifying the inter-industry Chinese stock market volatility spillover effect, the results of this thesis will help investors correctly understand the risks and provide them with guidance and advice to portfolio and risk management.