Systemic risk measurement in Chinese stock market based on FDG Copulas: from global financial crisis to Covid-19 pandemic

Abstract

Compared with most of the financial markets in developed countries, China's stock market has a short history of 30 years. Though developing rapidly, it is not mature and faces a lot of challenges. The Global Financial Crisis(GFC) in 2008 brought huge losses to the Chinese stock market. In the ten years after the GFC, it has experienced four bull markets and bear markets. In 2020, the COVID-19 pandemic struck financial market struck it again, but China's stock market ended 2020 on a high note, with a key benchmark up 27% and at a multiyear high. Considering high volatility of the financial market, this thesis aims to analyze systemic risk in China covering a period from the GFC to COVID-19 pandemic. To have a thorough understanding, this study measures the financial risk from both macro and micro perspective. This study first focuses on financial sector, including banks. insurance companies. securities companies, and other companies that provides financial services. FDG copulas is applied to data of 42 institutions and companies in financial industry throughout the whole sample period. By measuring the dependence coefficients, it is found that dependence is the strongest during the GFC. Securities companies are highly correlated with other companies. Our findings imply that dependence between financial institutions must be taken into consideration in portfolio management. Thus, in the following portfolio management problem, the author uses FDG copulas which are able to simulate data with dependence structure being considered to forecast risk. Our sample data includes stock prices of 47 financial institutions. To provide optimal weighting techniques, this study measures the portfolio risk by Component Expected Shortfall (CES) and put forward four strategies for people with different degrees of risk tolerance. This study also computed the joint extreme risk probability (JERP) and found that the probability of losing money from more than 60% of the asset is over 90% when the market risk is high. Finally, this study measures systemic risk of the entire stock market. Risk contributions of 29 industrial sectors and risk spillover effect are examined. The results show that financial sector contributes the most to systemic risk. From the GFC to the COVID-19 period, Banking sector gradually lost its dominant position to Non-bank-Finance Sector. The COVID-19 pandemic has improved the importance of some industries, such as Computer and Pharmaceuticals. Risk spillover effect is time-varying but its tendency is consistent with that of systemic risk. This thesis proposes the FDG copula-based CES method, which is ideal for risk forecasting when high-dimensional modelling is needed. Our research not only helps investors to better allocate their portfolio, but also help supervision departments to monitor key sectors and the co-movements of systemically important sectors (SISs) and other sectors, thus timely and effective measures could be taken to manage risk.