Research Article
Risk Quantification and Dynamic Guarantee Proportion Optimization Based on Improved Var-GARCH Model
@INPROCEEDINGS{10.4108/eai.2-12-2022.2332277, author={Xuemin Zhu and Sheng Liu and Xuelin Zhu}, title={Risk Quantification and Dynamic Guarantee Proportion Optimization Based on Improved Var-GARCH Model}, proceedings={Proceedings of the 3rd International Conference on Big Data Economy and Information Management, BDEIM 2022, December 2-3, 2022, Zhengzhou, China}, publisher={EAI}, proceedings_a={BDEIM}, year={2023}, month={6}, keywords={margin trading; dynamic guarantee ratio; historic simulation approach; var model}, doi={10.4108/eai.2-12-2022.2332277} }- Xuemin Zhu
Sheng Liu
Xuelin Zhu
Year: 2023
Risk Quantification and Dynamic Guarantee Proportion Optimization Based on Improved Var-GARCH Model
BDEIM
EAI
DOI: 10.4108/eai.2-12-2022.2332277
Abstract
In margin financing and securities lending business, customers always want a lower guarantee ratio, while securities companies and regulators require a high guarantee ratio. Obviously, the fixed margin system can no longer meet the needs of many parties. At this time, the dynamic margin system came into being. By establishing a risk control model, the dy-namic margin system can deduce a reasonable margin collection ratio, which can greatly improve the capital use efficiency of investors. Based on the above background, the paper uses optimized VaR-GARCH model to quantify the risk of margin financing and securities lending and set a dynamic margin ratio. This method combines GARCH model with historical simulation method, so that the VaR index calculated by historical simulation method can be adjusted with market fluctuations. In order to reflect the analysis process, the paper takes China Ping 'an Stock as an example to make a detailed analysis. The results show that in the 1067 trading days of the back test, the actual losses of China Ping 'an stock are all smaller than the calculated VaR value except for the actual losses in 10 trading days, which shows that the VaR-GARCH model proposed in this paper has high accuracy in risk measurement of margin financing and securities lending, covering almost all the actual loss ratios, and its ratio value is far smaller than the current fixed margin ratio of margin financing and securities lending.
