基于配对策略的基金动态资产配置

发布时间：2018-03-01 05:00

本文关键词： 配置策略指数奥恩斯坦-乌伦贝克过程配对交易 HJB方程　出处：《系统管理学报》2017年05期 　论文类型：期刊论文

【摘要】：作为市场中性策略的一种,配对交易早已被广泛应用于各类基金的投资实践。考虑基金管理者将总资产分别配置于风险资产与无风险资产,其中风险资产采用配对策略,假设风险资产价差服从指数O-U过程,基于最优投资组合理论,研究了不同市场环境、风险偏好下的最优投资组合动态资产配置策略。以投资到期效用期望最大化为目标,运用动态规划原理,得到了最优动态资产配置策略所满足的HJB方程,通过Legendre转换和分离变量法求得了该非线性方程的显式解,并证明了该显式解的相关性质。为了检验该配置策略的有效性,对我国A股市场的"中国太保"和"国海证券"进行了策略的情景模拟,通过最小二乘法和最大似然估计得到模型中的各项参数,并对2016-03-16~2016-04-29的实盘数据进行模拟交易,得到了该配对下的最优资产配置策略。最后,对A股各行业股票进行了全量模拟,计算了各行业的平均配对收益率,验证了本模型结果的稳健性。
[Abstract]:As a kind of market-neutral strategy, pairing transaction has been widely used in the investment practice of all kinds of funds. Based on the optimal portfolio theory and the exponential O-U process, this paper studies the optimal portfolio dynamic asset allocation strategy under different market environment and risk preference. By using the principle of dynamic programming, the HJB equation satisfied by the optimal dynamic asset allocation strategy is obtained. The explicit solution of the nonlinear equation is obtained by means of Legendre transformation and the method of separating variables. In order to test the effectiveness of the allocation strategy, the scenarios of "China Taibao" and "Guohai Securities" in China's A-share market are simulated. The parameters of the model are obtained by least square method and maximum likelihood estimation, and the real position data of 2016-03-16 / 2016-04-29 are simulated and the optimal asset allocation strategy is obtained. The average paired return rate of each industry is calculated, and the robustness of the model results is verified.
【作者单位】：上海师范大学商学院;上海师范大学数理学院;
【基金】：国家自然科学基金资助项目(71471117) 教育部人文社会科学研究青年基金资助项目(17YJCZH044) 上海市教委科研创新重点项目(13ZZ107) 上海师范大学自科项目(SK201506)
【分类号】：F832.51

【相似文献】