Box-Cox-SV模型及其对金融时间序列刻画能力研究

Research on Box-Cox-SV model and its abilities to describe financial time series

利用Box-Cox变换构造的随机波动模型,即Box-Cox-SV模型,较好地概括了现有文献中出现的一些常用SV类模型,避免了模型选择中的困难.论证了该类模型的矩属性和平方序列的自相关特征,利用MCMC估计方法对上证指数收益序列建立了Box-Cox-SV模型,并与EGARCH模型对金融时间序列的刻画能力在理论上和实证上进行了对比研究.

This paper uses the Box-Cox-SV models to avoid the difficulty in selecting suitable model for Stochastic Volatility（SV） models, Box-Cox-SV models include some commonly used SV models appeared in existing literatures. The character of the model＇s moments and auto-correlation of squared series are demonstrated. Using MCMC methods, the parameters of Box-Cox-SV model have been estimated. Under theoretical and empirical significance, a comparative research between Box-Cox-SV models and EGARCH models on their capability of describing financial time series has been conducted.