A1048
Title: Bayesian estimation of systemic risk in financial markets with dynamic networks
Authors: Mike So - The Hong Kong University of Science and Technology (Hong Kong) [presenting]
Abstract: In financial markets, systemic risk is a kind of risk that the failure of one stock in the market triggers a sequence of failures. The study proposes a Bayesian decision scheme to dynamically keep track of the systemic risk under any preference and restriction in financial risk management. We begin with capturing the moving correlations of stock returns. The correlation represents the strength of the relationship among stocks. Then, we construct a dynamic financial network to link together those stocks with a strong relationship. Making use of the concept of financial space, which is related to the position of stocks on the network plot, we locate two stocks in the financial space at a closer distance when the relationship between these two stocks is strong. Using the distance between stocks in the financial space, together with the preference and restriction in financial risk management, we propose a systemic risk measure. We demonstrate the use of the proposed systemic risk measure in producing early warning signals to the global financial instabilities using financial in 2020 to 2022.
