基于算法交易的机构投资策略研究分析

发布时间：2018-05-21 09:54

本文选题：算法交易模型 + 执行成本　；参考：《复旦大学》2013年硕士论文

【摘要】：在美国,算法交易已经成为基金业界交易策略的主流。我国算法交易尚处于萌芽状态,一些金融交易软件提供商和金融投资机构也正在尝试着进入这一领域。一些券商也开发了自己的算法交易平台,但由于种种原因,没有进行大范围的推广和使用。大部分中国机构投资者运用的现有系统还没有加入算法层,由基金经理发出指令,由交易员人工下单执行。这一执行过程受限于对市场实时变化数据的敏感性和及时性。当市场已经改变了方向,而往往策略却滞后于市场。不能及时根据市场的历史信息和动态实时信息进行精确分析和迅速的拆单、下单。基于以上传统下单的不足之处,本文提出了在交易层和数据层中加入了算法层的解决方案。执行下单时加入了算法策略,大大的提高了执行效率,提高了执行效果并且减少了冲击成本。此外,目前的算法交易研究多限于简单的VWAP算法。本文在此基础上还研究了IS算法的市场表现,加入这种在欧美市场同样流行的算法使得本文可以更加完整地反映算法交易对降低成本的作用。并且对于VWAP和IS两种算法的利弊进行比较。最后本文还加入了算法交易主体并检验其对市场质量有何影响,提起了算法交易对于市场的影响更多的关注。本文得出的结论主要有以下两点：从算法交易对执行成本的影响看：算法交易确实能够通过减少大额订单的市场冲击,在证券市场上为投资者降低交易成本、控制交易风险。VWAP算法在平均执行成本低于机构投资者的情况下,保证了更加稳定的执行效果；IS算法同样能够为投资者大幅节约交易成本,帮助投资者获取更高的投资收益率。虽然IS算法保证了机构投资者的交易能够更快更早的完成,但其绩效表现的波动较大,执行成本的标准差大于VWAP算法,执行效果及其稳定性均逊于VWAP算法。从算法交易对市场质量的影响看：算法交易能够通过减小大额订单对市场的冲击降低证券市场的波动性,并且算法交易所生成的实时更新的限价订单流能够为市场带来更好的流动性,本文研究表明算法交易的发展对提高证券市场的质量起着积极的作用。
[Abstract]:In the United States, algorithmic trading has become the mainstream of the fund industry's trading strategy. In China, algorithm trading is still in its infancy, and some financial transaction software providers and financial investment institutions are also trying to enter this field. Some securities companies have also developed their own algorithm trading platform, but due to various reasons, they have not carried out a wide range of promotion and use. Most of the existing systems used by Chinese institutional investors have not yet been added to the algorithmic layer, with fund managers giving orders and traders manually placing orders. This implementation process is limited by the sensitivity and timeliness of real-time market change data. When the market has changed direction, often the strategy lags behind the market. Unable to accurately analyze and quickly disassemble orders according to market historical information and dynamic real-time information in time. Based on the shortcomings of the traditional order, this paper proposes a solution to add the algorithm layer to the transaction layer and the data layer. An algorithm strategy is added to execute the order, which greatly improves the execution efficiency, improves the execution effect and reduces the impact cost. In addition, the current research on algorithm transaction is limited to simple VWAP algorithm. This paper also studies the market performance of is algorithm. Adding this algorithm, which is also popular in European and American markets, this paper can reflect the effect of algorithm transaction on cost reduction more completely. The advantages and disadvantages of VWAP and is are compared. Finally, this paper also adds the algorithm trading agent to test its influence on the market quality, and points out that the algorithm transaction has more attention to the market. The main conclusions of this paper are as follows: From the impact of algorithmic trading on execution costs, it is true that algorithmic transactions can reduce transaction costs for investors in the securities market by reducing the market impact of large orders. Under the condition that the average execution cost is lower than that of institutional investors, the control of transaction risk. VWAP algorithm ensures a more stable execution effect. The is algorithm can also save transaction costs for investors and help investors obtain higher investment returns. Although the is algorithm ensures that the transaction of institutional investors can be completed faster and earlier, its performance fluctuates greatly, the standard deviation of execution cost is larger than that of VWAP algorithm, and the execution effect and stability are inferior to that of VWAP algorithm. From the perspective of the effect of algorithm trading on market quality: algorithm trading can reduce the volatility of securities market by reducing the impact of large orders on the market. And the real-time updated limited order flow generated by the algorithm exchange can bring better liquidity to the market. The research shows that the development of the algorithm trading plays a positive role in improving the quality of the securities market.
【学位授予单位】：复旦大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F832.51;TP301.6

【相似文献】