报告题目：Ensemble learning enhanced VWAP execution

报 告 人：  程雪 北京大学金融数学系

报告时间：2018年11月3日下午16:00-17:00

报告地点：金融工程研究中心（本部览秀楼）105学术报告厅

报告摘要：

Volume Weighted Average Price (VWAP) strategy is a commonly used benchmark for the execution of meta orders. To dynamically impound real time market information such as traded prices and volumes into the execution of a meta order, in this article we propose to enhance the plain VWAP strategy by incorporating the ensemble learning inferred probability of next price move and the Kelly principle into a pre-assigned VWAP execution. The resulting strategy is termed as the {\it ensemble learning enhanced VWAP (eVWAP)} strategy. The eVWAP strategy is implemented to the component stocks in the SSE50 Index of Shanghai Security Exchange and its performance is investigated with the plain VWAP strategy. Joint work with Yuhao Lu and Tai-Ho Wang.