Research On The Dependence Structure Of Stocks In The Industry Based On Vine Copula

The rapid development of my country’s economic level has driven the rapid development of all walks of life,resulting in the emergence of more and more leading enterprises in various industries,as well as many small and medium-sized enterprises with their own characteristics.At present,with the continuous development and growth of capital,some listed companies are not satisfied with the improvement in their own industries,and gradually begin to expand the market of other industries.This move is conducive to its own development and the prevention of financial risks,which also makes more and more industries have closer relations,and the connection between different institutions in the industry is also closer.Studying the interdependence structure of stocks in the industry will help investors make scientific and rational investments,gain a deeper understanding of the interdependence between stocks in the industry,and help market management departments effectively prevent financial risks.Therefore,it is of great significance to explore the interdependence structure among stocks in the industry.In order to explore the interdependent structure of stocks in the industry,this paper selects eight relatively popular industries as examples,which are semiconductor,real estate,Internet,automobile manufacturing,securities companies,chips,medical care and banking.Relevant data were collected and two periods of time were selected for comparison.The first sample period was from January 2,2018 to June 28,2019,and the second sample period was from January 2,2020 to June 2021.30 days.Through the AR-GJR-GARCH-EVT-Vine Copula model,the marginal distribution of each stock is described,and the tail characteristics of each stock are described.With the help of the maximum spanning tree algorithm,the copula function that can best describe the dependency structure between variables is selected,and the measure The interdependence structure of stocks in eight sectors.By analyzing the data,it is found that the logarithmic return series of stocks generally have characteristics that do not meet the normal distribution,such as sharp peaks,thick tails,and skewness.Some data have ARCH heteroscedasticity effects and residual autocorrelation phenomena.By establishing the AR(1)-GJR-GARCH(1,1)-Skew-t marginal distribution model,and processing the logarithmic return sequence of stocks,it is found that the model can well describe its autocorrelation characteristics and effectively eliminate the Conditional heteroscedasticity and leverage effects can also simulate the spikes,thick tails and skewed characteristics of the sequence,and the GPD model can further analyze and study the tail characteristics of the sequence.By constructing an R-Vine model,it is found that there is a strong interdependence between stocks in various industries.Among the eight industries,the four industries with relatively strong interdependence are real estate,automobile manufacturing,securities companies and banks.The weak ones are the semiconductor,Internet,chip,and medical and health industries;the institutions with greater influence in the first period are generally also the institutions with greater influence in the second period.In the first period,the two dependencies are Stronger institutions generally maintain stronger dependencies during the second time period.