基于SKT-ARFIMA-HYGARCH-VaR模型的股票型基金投资风格漂移风险测度研究

发布时间：2018-03-18 04:21

本文选题：基金投资风格　切入点：投资风格漂移　出处：《中国管理科学》2011年05期 　论文类型：期刊论文

【摘要】：投资风格漂移是把双刃剑,基金在获取短期超额收益的同时,其背后也折射出巨大的漂移风险。本文以2004年成立的8只开放式股票型基金为研究样本,在量化基金投资风格漂移收益及分析其序列呈尖峰、厚尾与有偏特征的基础上,通过引入skt分布来刻画新生变量的分布,构建ARFIMA-HYGARCH-VaR模型来测度基金投资风格漂移风险值,并与skt分布下的RiskMetrics及GARCH族等5种VaR模型的风险测度能力做了比较实证分析,同时对各种VaR模型进行失败频率回测检验与动态分位数测试。研究结果表明:在不同显著性水平下,skt分布下的各种模型基本都有较好的风险测度能力,但ARFIMA-HYGARCH模型的VaR风险测度更加精确与稳定;Person吻合度检验也证实了skt分布能较好刻画投资风格漂移日收益序列的分布。本研究为控制较严重的投资风格漂移及规范基金产品创新设计与发行无疑具有重要的理论价值与现实意义。
[Abstract]:The investment style drift is a double-edged sword. While the fund obtains short-term excess income, it also reflects the huge drift risk behind it. This paper takes 8 open-end equity funds established in 2004 as the research sample. On the basis of quantifying the investment style drift returns and analyzing the peak, thick tail and biased characteristics of the series, the paper introduces skt distribution to describe the distribution of new variables, and constructs the ARFIMA-HYGARCH-VaR model to measure the risk value of fund investment style drift. And compared with five VaR models under skt distribution, such as RiskMetrics and GARCH family, the ability of risk measurement is analyzed empirically. At the same time, failure frequency test and dynamic quantile test are carried out on various VaR models. The results show that all the models with different significant levels have good risk measurement ability. But the VaR risk measure of ARFIMA-HYGARCH model is more accurate and stable / person coincidence test also proves that skt distribution can depict the distribution of daily return series of investment style drift. This study is to control the serious investment style drift and the criterion base. There is no doubt that the innovative design and distribution of gold products have important theoretical value and practical significance.
【作者单位】：华南理工大学经济与贸易学院;华南理工大学工商管理学院;广东商学院金融学院;
【基金】：教育部人文社科基金2010年度规划项目(10YJA630131)
【分类号】：F224;F832.51

【参考文献】