Portfolio management using time-varying vine copula: an application on the G7 equity market indices
研究了在七国集团股票市场指数投资组合中，使用时变藤蔓连接函数捕捉资产间依赖结构的结构性断点，相比传统均值方差和等权重策略，在最小条件风险价值和收益风险比上表现更优。
We consider structural breaks and use vine copulas to hierarchically model the underlying assets’ dependence structure of the portfolio of G7 equity market indices (1998–2019). This framework is noticed for its flexibility in capturing asymmetry and non-linearity in a time-varying style. We compare the portfolio performance in terms of the minimum Conditional Value-at-risk (CVaR) and the maximum return-to-CVaR ratio criteria with the traditional mean-variance framework and the equal-weighted strategy. The outcomes show the outperformance of our method across subperiods. Canonical vine copula marginally outperforms drawable vine copula in terms of return-to-risk ratio. Our proposed vine copula models better capture the risk-return tradeoff especially during critical market moments.