| Article ID | Journal | Published Year | Pages | File Type |
|---|---|---|---|---|
| 1144027 | Systems Engineering Procedia | 2012 | 7 Pages |
In the paper, firstly, we fit the returns series with GARCH-EVT method, construct dynamic correlation coefficient matrix of elliptical copulas with DCC method, then analyze the dynamic dependency between stock index returns, finally calculate the CVaR of QDII funds through Monte Carlo simulation. Empirical study indicates that: while compared with dynamic Copula method, the static Copula model assuming the correlation between the assets is unchanged has overestimated the risk. CVaR value calculated by dynamic t Copula method is smaller than the one calculated via dynamic Gaussian Copula method. Accordingly, to optimize the investment portfolio, QDII fund manager should decrease the investment proportion in the developed market and emerging market while more money should be invested in Chinese A shares and the Hong Kong market.Kewords: QDII Fund; Dynamic Copula; CVaR
