Financial Risk Spillover Effect Metric Model Using Quantile Regression Method And Its Application Research

Measuring the Financial risk spillover effect is one of an important topic in areas of financial econometrics.In financial risk management,with the development of financial deregulation and globalization,fianacial risk spillover effect has become an important risk source of systemic financial risk.Financial risk regulation department not only need to strengthen the supervision of individual financial institutions,but also should more take into account the risk spillover effect between financial institutions to avoid the outbreak of systemic financial risk.Therefore,establishing financial risk spillover econometric model to measure systemic financial risk,analyzing the risk exposure and the contribution degree to systemic risk of individual financial instritutions,identifying important risk financial institutions and depicting the risk spillover path between financial institutions and the risk transmission network topology,are hot issues to be resolved in financial risk management.With the aid of quantile regression theory,this paper constructs different data dimension financial risk spillover effect model.The financial risk spillover measurement model using quantile regression theory can directly model the extreme quantile of financial asset returns and depict the tail behavior of financial markets.In theory,financial risk spillover measurement model will develop research ideas and methods for systemic risk management.Meanwhile,it will also provide technology and evidence for analysis and decision making in financial risk management applications.Firstly,through combine with the vectory autoregressive model and quantile regression technology,this paper proposes a quanitle vector autoregressive model.Then by setting the extreme quantile level,using the impulse response function to measure the risk shock response between financial instrituions.Moreover,employing the variance decomposition technique to construct financial risk spillover effect index used to measure the static and dynamic financial risk spillover effect.The constructed financial risk spillover measurement model is applied to analysis the G7 stock market financial risk spillover effects,and the empirical result verifies the feasibility and effectiveness of quantile vector autoregressive model.Secondly,in order to capture the nonlinear behavior of financial risk spillover,this paper considers the balance of model building adaptability and interpretability,and uses the time series variable selection technology to build semi-parameter quantile regressive model.Further,based on the semi-parameter quantile regressive model to provide nonlinear Co VaR model,which used to measure static and dynamic financial risk spillover effect and identify the financial institution's risk contribution and risk exposure.The nonlinear CoVaR model is applied to analysis the 15 industries financial risk spillover effects in Chinese stock market,and the nonlinear Co VaR model is verified to be feasible and effective.Thirdly,from the perspective of high-dimensional data,this paper builds a factor CAViaR model that can measure the high-dimensional financial system risk spillover effect.The factor CAViaR model uses factor quantile regression method to estimate model coefficient,and applies singular value decomposition to identify financial risk factor contribution and factor load.Through constructed tail risk behavior analysis graph and global risk comovement analysis graph,exploring the asymmetry of risk spillover and identification of financial risk important institutions.The factor CAViaR model is applied to analysis the 172 stock returns risk spillover effects in Chinese stock market,and the factor CAViaR model is verified to be feasible and effective.Finally,based on the financial risk spillover effect and network toplolgy theory,this paper from the systematic perspective to construct financial risk network that more intuitive and image analysis risk spillover effect.Employing variance decomposition technique of quantile vector autoregressive model to measure the financial risk spillover effect,then designs financial risk weighting matrix to build the directed and weighted financial risk network.Moreover,identification of risk network topology characteristics and using node link scatterplot to visualize financial risk network.In the empirical analysis,constructing the bank shares financial risk network in Chinese stock market.