基于β系数优选的股票动态投资组合分析

发布时间：2018-04-30 14:06

本文选题：β系数 + 量化投资策略　；参考：《重庆工商大学》2017年硕士论文

【摘要】：自从马科维茨的投资组合理论提出到现在已经有了六十多年,该理论在很多方面已经取得了很大的进步和发展。比如:从单期研究到多期的研究,通过各种方法简化均值-方差模型的求解过程,对风险衡量方法的修正以及对原假设条件的逐渐放松等方面。在实践方面,投资组合理论问世以后,它开始为金融机构和投资者所广泛采用;金融学也开始了它在实际投资中的量化阶段。本文给出了基于定量的投资组合的管理方法,该方法主要包括两个阶段,一是行业和股票优选阶段,二是对股票进行投资组合,主要是确定所选股票最优投资权重的。上述两个阶段主要在本研究的第3章和第4章进行论述。第3章选取申银万国一级分类行业指数2014年6月3日至2015年9月30日这样一个涵盖市场上升阶段和下跌阶段的完整投资周期数据为样本。通过对这些样本数据的检验,得出了这些一级分类行业指数受市场态势的影响;并计算了行业指数的上升β系数(up-marketβ,+?)和下降β系数(down-marketβ,?-)。最后,根据上升β系数和下降β系数的比值构建了一个指标并从中选出了6个优质的行业。第4章引入参考时间窗口长度L和持有期限窗口长度H两个外生的时间参数构建了动态的均值-方差投资组合模型,并利用遍历法求解最优时间窗口参数。然后采用了第三章优选行业的代表性股票进行动态投资组合实证。在假定投资者风险容忍水平一定的情形下,以投资者效用最大化为衡量标准,利用MATLAB软件程序来寻求收益最优的外生参数及每次调整资产的权重。最后,利用计算所得最优参数进行投资,并通过多项业绩评价指标(包括投资期年收益率、风险调整收益率和预测收益率等)对比分析动态投资组合策略和被动投资的收益情况。结果表明本文的动态投资组合策略在风险调整后的收益率以及预测收益率等方面表现均优于被动投资。总之,本文研究为投资者提供了一种定量的投资组合管理方法,具有一定的理论意义及实用价值:一方面,通过实证分析验证了我国股市的非有效性;另一方面,为投资者如何分配各股票投资权重提供了有益的借鉴。
[Abstract]:It has been more than 60 years since Markowitz's portfolio theory was put forward. It has made great progress and development in many aspects. For example, from single-period study to multi-period study, the solution process of mean-variance model is simplified by various methods, the risk measurement method is modified, and the original assumptions are gradually relaxed. In practice, portfolio theory began to be widely used by financial institutions and investors, and finance began its quantitative stage in actual investment. In this paper, a quantitative portfolio management method is presented. The method mainly includes two stages, one is the industry and the stock selection stage, the other is the stock portfolio, which is mainly to determine the optimal investment weight of the selected stock. The above two stages are mainly discussed in chapters 3 and 4 of this study. The third chapter selects the whole investment cycle data from June 3, 2014 to September 30, 2015, which covers both the rising and falling stages of the market. Based on the test of these sample data, the influence of market situation on the industry index is obtained, and the rising 尾 coefficient of industry index is calculated by up-market 尾. And decreasing 尾 -market. Finally, according to the ratio of rising 尾 coefficient and decreasing 尾 coefficient, an index was constructed and six high quality industries were selected. In chapter 4, the dynamic mean-variance portfolio model is constructed by introducing two exogenous time parameters: the reference window length L and the holding term window length H, and the optimal time window parameters are solved by traversal method. Then we use the third chapter to select the representative stocks in the industry to carry out dynamic portfolio demonstration. Under the assumption that the level of investor risk tolerance is constant, taking the maximization of investor utility as the criterion, the MATLAB software program is used to find the optimal exogenous parameters of income and the weight of assets adjusted every time. Finally, using the calculated optimal parameters to invest, and through a number of performance evaluation indicators (including the annual rate of return on the investment period, Risk adjusted rate of return and forecast rate of return are compared to analyze the dynamic portfolio strategy and the return of passive investment. The results show that the dynamic portfolio strategy performs better than passive investment in terms of risk-adjusted return rate and prediction rate of return. In short, this paper provides a quantitative portfolio management method for investors, which has certain theoretical significance and practical value: on the one hand, it verifies the non-validity of China's stock market through empirical analysis; on the other hand, For investors how to allocate the weight of each stock investment provides a useful reference.
【学位授予单位】：重庆工商大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F224;F832.51

【参考文献】