基于成交量分解模型的改进VWAP策略

发布时间：2018-02-14 20:02

本文关键词： 算法交易 VWAP策略成交量分解日内周期结构　出处：《运筹与管理》2017年02期 　论文类型：期刊论文

【摘要】：传统VWAP(交易量加权平均价格)策略通过拆分大额委托订单,跟踪市场成交均价,达到最小化冲击成本的目的,而准确预测成交量日内分布是运用VWAP策略的关键。通过详细考察现有的改进VWAP策略中成交量预测模型的建模方式和预测结果,发现由于无法分离成交量日内周期结构,现有模型样本依赖性较大且难以适用于多数股票。因此,本文从个股与市场成交量变化趋势的关系角度出发,推导个股成交量与市场趋势的关系,通过构造个股成交量关于市场因素的因子载荷,将日内成交量分解为市场共同部分和个股特殊部分,预测成交量日内分布并构建动态VWAP策略。实证结果表明新的成交量分解模型可以有效分离个股的成交量日内周期结构,在此基础上构造的改进VWAP策略不仅具有较为广泛的适用性,且跟踪误差减少幅度比现阶段同类型的改进VWAP策略更大,能更好的降低市场冲击成本。
[Abstract]:The traditional VWAP (transaction volume weighted average price) strategy can minimize the impact cost by splitting large entrusted orders and tracking the average transaction price in the market. Accurate prediction of intraday distribution of trading volume is the key to the use of VWAP strategy. Through a detailed study of the modeling methods and prediction results of the existing improved VWAP model, it is found that the intraday cycle structure of trading volume cannot be separated. The existing models are highly dependent and difficult to apply to most stocks. Therefore, this paper derives the relationship between individual stock turnover and market trend from the point of view of the relationship between individual stock and the trend of market turnover. By constructing the factor load of individual stock trading volume on market factors, the intra-day trading volume is decomposed into common market share and individual stock special part. Forecasting intraday distribution of turnover and constructing dynamic VWAP strategy. The empirical results show that the new decomposition model can effectively separate the intraday cycle structure of individual stock trading volume. On this basis, the improved VWAP strategy not only has more extensive applicability, but also reduces the tracking error more than the same type of improved VWAP strategy at present, and can better reduce the market impact cost.
【作者单位】：电子科技大学经济与管理学院;
【基金】：教育部人文社会科学研究项目(14YJA790062)
【分类号】：F830.91

【相似文献】