基于t-Copula GARCH模型的投资组合风险测度研究

Research on Investment Portfolio Risk Measurement Based on t-Copula GARCH Model

摘要: 目前，各个行业相依性变得越来越高，金融风险传染加剧，投资组合管理的重要性尤为突出。本文选择五家公司股票2019年1月1日~2023年6月1日的每日收盘价为研究对象，对数据进行处理后，选择t-Copula-GARCH模型，利用蒙特卡洛模拟方法对不同置信水平下的投资组合风险价值VaR进行测度。结果表明，t-Copula GARCH模型具有刻画真实的金融资产分布的能力，不同资产之间相依关系是非对称的，在不同的置信水平下使用Copula函数构建的投资组合能显著降低投资风险。因此，基于本文的结论，建议考虑各行业间相依性的强弱以及相依结构的特点进行理性的投资，并将自身的风险偏好与预测结果相结合，对投资组合进行适当的调整。

Abstract: Currently, the interdependence between various industries is becoming increasingly high, and the contagion of financial risks is intensifying. The importance of portfolio management is particularly prominent. This article selects the daily closing prices of five companies’ stocks from January 1, 2019 to June 1, 2023 as the research object. After processing the data, the t-Copula GARCH model is selected to measure the VaR of investment portfolios at different confidence levels using Monte Carlo simulation method. The results indicate that the t-Copula GARCH model has the ability to characterize the true distribution of financial assets, and the interdependence between different assets is asymmetric. Investment portfolios constructed using the Copula function at different confidence levels can significantly reduce investment risk. Therefore, based on the conclusions of this article, it is recommended to consider the strength of interdependence between industries and the characteristics of interdependence structures for rational investment, and combine one’s own risk preferences with prediction results to make appropriate adjustments to the investment portfolio.