基于Copula-GARCH的投资组合风险度量的实证应用

Application research of Copula-GARCH method in portfolio risk measurement

通过Va R模型对我国的上证指数和深成指2010年1月25日至2012年12月28日的收盘价的波动性加以实证分析,进行模型检验,从而得出模型有效的结论。用Copula函数刻画上证综指和深成指之间的相关结构进而得到投资组合收益率的联合分布,建立了GARCH(1,1)-t模型,利用各项资产收盘价、成交量的历史数据,通过Monte Carlo模拟生成具有Copula相关结构的收益率分布,得到组合的风险价值。应用Copula-Va R-GARCH模型通过Monte Carlo模拟计算投资组合的Va R值,对探讨股票市场波动(风险)以及与预期收益之间的关系具有重要的理论意义和实用价值。

In empirical aspects, this paper is based on Shanghai stock index and Shenzhen index component portfolio for research object, and takes closing price from Jan.25th, 2010 to Dec.28th, 2012 as the valid data. By comparing the empirical results,it can draw conclusion that the two kinds of Copula Function selection methods are feasible, indeed using the preferred Copula Function to establish GARCH(1,1)-t model, also for Copula applications in portfolio risk research provides a new idea. The paper use Copula-VaR-GARCH model to imitate financial time series and compute VaR by Monte Carlo simulation, studying on stock market volatility(risk)and the relationship between expected return has important theoretical significance and practical value.