基于沪胶主连的程序化交易系统开发与应用

发布时间：2018-03-16 19:12

本文选题：程序化交易系统　切入点：系统设计　出处：《长江大学》2017年硕士论文　论文类型：学位论文

【摘要】：随着国内首只期货期权(豆粕期权)的推出,金融市场上的衍生产品更加的丰富。当然这种发展趋势是为了保护广大中小投资者的权益,但同时给交易者带来了巨大挑战。投资标的繁多使得交易者会错失很多良机,但是在计算机技术与互联网技术迅猛发展的今天,程序化交易最终成为人们一致的选择。程序化交易之所以为广大交易者所接受是因为其自身所具有的系统客观性,风险分散性所决定的。在进行交易之前编写好入场规则,出场规则,资金管理规则并交予计算机严格执行。最大程度上避免了人为主观性的干预及人工下单不具备的分散风险的劣势。首先文章从实战出发,基于上海期货交易所的品种沪胶主连,对商品交易过程中的持仓量,成交量与价格波动的关系做VEC模型,对价格做方差分解,从长期来看持仓量和成交量对于价格解释的贡献水平在0.05%和0.06%左右。由此,在进行程序化交易系统信号触发设置时,除了要考察历史价格的变动外,还要考察持仓量和成交量对于价格波动的影响。其次基于概率神经网络,数量化各种交易形态,将其分为9种主要交易形态,利用PNN网络进行分类识别,判断趋势的重合率为多少。试验数据结果显示分类结果良好,识别准确度达到91.67%,好于预期。利用MATLAB试验结果做出程序化交易系统趋势信号指标TI,并根据趋势指标进行程序化的交易信号优化与决策。最后介绍了程序化交易系统的开发与设计步骤,并且在实践当中检验程序化交易系统的实战性,给出了测试数据的报告,并对测试报告进行系统的分析与解读。
[Abstract]:With the introduction of the first futures option (soybean meal option) in China, the derivative products in the financial market have become more abundant. Of course, this trend of development is aimed at protecting the rights and interests of the majority of small and medium-sized investors. But at the same time, it brings great challenges to traders. A lot of investments make traders miss many opportunities, but today, with the rapid development of computer technology and Internet technology, Procedural trading eventually becomes a unanimous choice. Procedural trading is accepted by traders because of its systemic objectivity and dispersion of risks. The rules of entry and management of funds are strictly enforced by the computer. To the greatest extent, it avoids the disadvantages of artificial subjective intervention and the decentralized risk that the orders do not have. First, the article starts from the actual combat. Based on the variety of Shanghai Futures Exchange, the VEC model is used to analyze the relationship between the position, the volume and the price fluctuation, and the variance of the price is decomposed. In the long run, the contribution of position and trading volume to price interpretation is about 0.05% and 0.06%. Secondly, based on probabilistic neural network, quantification of various trading patterns, it is divided into 9 main trading patterns, using PNN network to classify and identify. The results of the test show that the classification results are good. The recognition accuracy is 91.67, which is better than expected. The trend signal index tii of programmed trading system is made by using the result of MATLAB test, and the program transaction signal optimization and decision is made according to the trend index. Finally, the programmed trading system is introduced. The steps of development and design, In practice, the paper tests the effectiveness of the program trading system, gives the report of the test data, and analyzes and interprets the test report systematically.
【学位授予单位】：长江大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：O212.1

【相似文献】