证券投资统计套利策略实证分析

发布时间：2018-02-21 13:02

本文关键词： 统计套利策略协整主成分分析证券市场　出处：《山东财经大学》2012年硕士论文　论文类型：学位论文

【摘要】：2011年在中国金融史上是不平凡的一年，，融资融券的正式推出不仅为中国证券市场带来了做空机制，也使得中国证券市场彻底告别了单边市的时代。中国的证券市场也将进入机会更多、更有效率和市场化程度更高的新时期。很多以前在发达国家得到广泛应用的投资理念和策略，也将在中国落地生根大放异彩。这其中就包括统计套利的投资策略和技术应用。本文正是在这种大背景下，通过借鉴金融发达国家的统计套利投资理念，针对中国内地证券市场做实证分析，也算是统计套利中国化的初步探索。通过长期的证券投资实践，人们总结出很多规律和投资策略，这其中就包括了利用股票与股票价格之间的长期均衡关系和在短时间内打破这种均衡关系的“异象”之间进行套利的统计套利策略。本文通过对统计套利技术的介绍、分析和实证检验，证实了这种投资策略在中国内地证券市场上应用的可行性。统计套利策略有很多具体的方法，本文采用了应用最为广泛的协整策略和主成分分析策略。通过对这两种方法的对比得出了主成分分析方法效率更高，在国内市场更加实用的一般性结论。在实证分析中，我们使用了很多残差处理模型和样本期外时间窗口外推方法。经过对比分析我们得出了使用正态分布和均值回归模型处理残差序列，使用每日平移时间窗口外推方法收益率最高的结论。
[Abstract]:2011 was an extraordinary year in China's financial history. The formal introduction of margin financing not only brought about a short selling mechanism for China's securities market, It also made China's securities market bid farewell to the era of one-sided market. China's securities market will also have more opportunities. Many of the investment ideas and strategies that were widely used in developed countries will also take root in China. This includes the investment strategy and technology application of statistical arbitrage. It is under this background that this paper makes an empirical analysis on the securities market in the mainland of China by referring to the statistical arbitrage investment concept of the developed financial countries. It is also a preliminary exploration of statistical arbitrage in China. Through long-term securities investment practice, people have summed up a lot of laws and investment strategies. This includes the statistical arbitrage strategy which uses the long-term equilibrium relationship between stock and stock price and breaks the "vision" of the equilibrium relationship in a short time. The analysis and empirical test confirm the feasibility of applying this investment strategy to the securities market in mainland China. There are many specific methods for statistical arbitrage strategy. This paper adopts the most widely used cointegration strategy and principal component analysis strategy. By comparing these two methods, it is concluded that the principal component analysis method is more efficient. A more practical general conclusion in the domestic market. In empirical analysis, We use a lot of residual processing models and extrapolation methods of time window outside the sample period. Through comparison and analysis, we get that we use normal distribution and mean regression model to deal with residual sequences. Use the daily translation time window extrapolation method to get the highest rate of return.
【学位授予单位】：山东财经大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：F224;F832.51

【引证文献】