The Spillover Effect Of G20 Stock Market Volatility Based On The Network Perspective

With the development of information technology,the globalization and liberalization of financial markets are deepening,and the links and information transmission among financial markets are constantly strengthening.Economic fluctuations in one country or region may rapidly evolve into a huge impact on other economies and stock markets through spillover effects,even triggering a global financial crisis.In order to reduce this risk,it is often necessary to study the volatility spillover effect between international financial markets to achieve risk allocation and market management.Due to the interaction among various entities,the international financial market has become a huge and complex system.It is difficult to analyze the spillover effects among various entities in a complex financial system by economic methods alone.Considering that the complex network approach is a powerful tool to study the intricate spillover relationship between financial markets,empirical analysis of volatility spillover between international financial markets using the network perspective is a hot research area in recent years.As the financial volatility spillovers in the world mainly come from some developed markets and emerging markets,and the Group of 20 covers the major developed markets and emerging markets in six continents of the world,the total gross domestic product(GDP)of the G20 member countries accounts for 85%of the global economy,and the total trade accounts for 80%of the total global trade.It has strong representativeness and wide coverage,and is of great significance to the study of volatility spillovers between international financial markets.Since the EU joined the G20 as an economy,with Britain,Germany and other developed countries joining the G20 separately,the academic community excluded the EU and studied the ties among the 19 independent countries in the G20.Therefore,this paper selects the stock price data of representative stock indexes of 19 independent countries in the G20 from January 2006 to December 2020,depicts the volatility spillover relationship between stock indexes by using the generalized variance decomposition method based on VAR model.construct.s static volatility spillover network and dynamic volatility spillover network of the G20 stock market respectively,studies the topological properties and evolut ion of volatility spillover network from node influence and network centrality analysis.and analyzes the fundamental factors that affect the volatility spillover range of the G20 stock market through panel regression analysis.The details are as follows:In the first chapter,the research background and significance of this paper are firstly described.Then,the research process of volatility spillover,complex networks and their applications in the G20 stock market are reviewed and evaluated.Finally.the main research content and specific framework of this paper are explained.The second chapter firstly introduces the development and research work of complex network theory,then introduces the basic knowledge of graph theory,including the basic concept of graph and the matrix representation of graphthen introduces in detail the topological indexes of network structure used in this paper,such as correlation degree,degree centrality and feature vector centrality,and finally summarizes the theoretical knowledge of the stock market.The third chapter analyzes the direction.size and evolution of volatility spillovers between economies from 2006 to 2020 by constructing static and dynamic networks.This chapter selects the daily closing price data of representative stock indexes of 19 countries from January 4,2006 to December 31.2020 in Wind database,and calculates the logarithmic return rate based on the unmatched transaction data deleted from the existing literature.Firstly,the static generalized variance decomposition is carried out on the logarithmic return rate of the whole sample,and the adjacency matrix of the whole sample network is obtained to depict the volatility spillover relationship among the stock indexes,and the spillover direction and magnitude among the Group of 20 are analyzed from the micro level(pairwise correlation degree)and the system level(total direct correlation degree and total correlation degree)by using the concept of node degree;Then,the degree centrality and eigenvector centrality are used to analyze the systematic importance of each economy.Then,the generalized variance decomposition is performed on the logarithmic return rate by year division,and 15 volatility spillover networks are constructed.The ranking evolution of the spread spillover value,the acceptance spillover value and the net spillover value of each economy in each year is identified by the output intensity,the input intensity and the net intensity respectively.The empirical results show that:(1)During the whole period from 2006 to 2020,it is found that the association degree among the EU member-states(France,Germany and Italy)is in the top three,and France,Germany,the United Kingdom and the United States rank first,second and third in the total direct association degree for all other economies in the system,so as to infer that they are at the core of the network.At the same time,it is also found that the total association degree of the G20 stock market is high and the overall risk level is high.(2)France,Britain,Germany and other developed countries have greater degree centrality,while France,Australia,Canada.China and Saudi Arabia have greater feature vector centrality.(3)The United States is the largest disseminator of volatility spillovers,and France is the most important recipient of volatility spillovers.In the fourth chapter,we analyze the influencing factors of volatility spillover based on the network constructed in the previous chapter.We describe the spread volatility spillover range,the tolerance volatility spillover range and the net volatility spillover range of each economy through the output intensity,the input intensity and the net intensity of the nodes,and conduct panel regression on the three to study the influence of fundamental factors on the volatility spillover range.The empirical results show that:(1)The higher the volatility of stock price,the higher the real GDP growth rate and the lower the ratio of total capital formation to GDP,the stronger the volatility spillover ability of representative stock indexes of G20 member countries.(2)The higher the real GDP growth rate,the lower the current account balance to GDP ratio and the lower the annual exchange rate,the stronger the volatility spillover ability of the representative stock indexes of the G20 member countries.(3)The higher the volatility of stock price,the lower the ratio of current account balance to GDP and the lower the ratio of total capital formation to GDP,the stronger the net spillover ability of representative stock indexes of G20 member countries.Finally,the empirical results of this paper are given from two aspects:the network correlation degree and the influencing factors of network volatility spillover in the stock market.In addition,we further look forward to the future research work.