Abstract: The traditional VWAP trading strategy trading volume depends on the distribution of split single transaction, and the forecast of the distribution of trading volumes is based on the interval volume proportion of the total volume, and this prediction method does not consider the stock price changes. Therefore, by means of the time series factor decomposition the paper first predicts the distribution of trading volumes, then based on the stock price changes in the distribution of trading volumes adjust the trading volume distribution of dynamic ranges and builds the stocks selling strategy by the trading volume distribution, and finally, through an empirical test in this paper, tests the predictive validity and effectiveness of the distribution of trading volumes. Numerical results show that the prediction method of the distribution of dynamic volumes given in this paper is better than the traditional VWAP method and the results of the new method are much closer to the distribution of the actual transaction volume. And compared with the traditional VWAP trading strategy, the strategy given in this paper is more gainful.

Key words: algorithmic trading, VWAP selling strategy, trading volume distribution prediction, dynamic adjustment

摘要： 传统的VWAP交易策略通过预测区间交易量分布进行拆单交易,对于交易量区间分布的预测是基于区间成交量占总成交量比例进行的,这一预测方法没有考虑股票价格变动因素。因此,本文首先通过时间序列因素分解方法进行区间交易量分布预测,进而根据股票价格变化对区间交易量分布进行动态调整,并构建了基于动态区间交易量分布的股票卖出策略,最后通过实证检验了本文给出的动态区间交易量分布预测的有效性和交易策略的有效性。数值结果表明,本文所给动态区间交易量分布预测方法比传统VWAP方法预测结果更加接近于实际的交易量分布,且本文所给交易策略与传统VWAP交易策略相比,具有更大的收益。

关键词: 算法交易, VWAP卖出策略, 交易量分布预测, 动态调整