基于机器学习的改进TWAP、VWAP算法的研究及应用
投稿时间:2020-01-10  修订日期:2020-02-12  点此下载全文
引用本文:
摘要点击次数: 74
全文下载次数: 0
作者单位E-mail
郑继翔 招商证券股份有限公司 zhengjx@cmschina.com.cn 
陈卓 招商证券股份有限公司  
柯军 招商证券股份有限公司  
黄钰 招商证券股份有限公司  
洪轩儒 招商证券股份有限公司  
李怡洁 招商证券股份有限公司 liyijie@cmschina.com.cn 
中文摘要:算法交易是一种全新的交易方式,备受机构投资者青睐,近年来市场份额占比持续上升,算法交易的快速发展对资本市场的总体效率与微观结构影响深远。本文对传统TWAP与VWAP算法进行改进,利用滚动的1分钟粒度高频实时资金博弈数据,引入机器学习的方法训练量价模型,对股票分钟价格走势进行预测,并将短期价格预测结果代入优化模型生成最优委托价格和数量。该算法经测试可以稳定的跑赢市场均价,具备推广应用的可行性。
中文关键词:算法交易 短期价格预测 机器学习 逻辑回归 TWAP VWAP
 
Research and Application of Improved TWAP and VWAP Algorithms Based on Machine Learning
Abstract:Algorithm Trading is a new trading method that is favored by institutional investors. In recent years, the market share has continued rise, the rapid development of algorithm trading has a profound impact on the overall efficiency and microstructure of the capital market. This paper improves traditional TWAP and VWAP algorithms, uses rolling one-minute high-frequency real-time financial data, and introduces machine learning methods to train the price-and-volume model to predict asset minute price trend, then substitutes short-term predication results into the optimization model to generate optimal order price and quantity. This algorithm has been tested to outperform the market average and has feasibility to promotion and application.
keywords:algorithm trading  short-term price predication  machine learning  logistic regression  TWAP  VWAP  
查看全文   查看/发表评论   下载pdf阅读器