Statistical Inference of Stochastic Copulas and Their Application to Finance

结果表明，粒子滤波技术是一种有效的时变相关结构随机copuls统计推断方法。在金融应用方面，我们开发了基于copula的随机波动率（SV）模型、多变量SV模型、随机静脉copula模型等。此外，我们还提出了基于尾风险平价/预算方法的新型投资组合优化。

We find out that the particle filtering technique is an effective method for statistical inference of stochastic copulas with time-varying dependence structure. As financial applications we have developed copula-based stochastic volatility(SV) model, multi-variate SV model,stochastic vein copula model, and so on. Furthermore, we have proposed new types of portfolio optimization based on the tail-risk parity/budgeting approach.