基于多因素收益预测的均值—方差投资组合模型及实证研究

发布时间：2018-05-07 08:44

本文选题：投资组合 + 多因素　；参考：《东北大学》2012年硕士论文

【摘要】：上世纪80年代末90年代初,我国的证券市场开始慢慢形成。二十年来,我国的证券总体规模逐步提升,形成了一个较为完善的系统。然而金融市场规模庞大,极为复杂,其中每一项投资活动都具有较高的风险性,各项资产收益也具有较高的不确定性。许多投资者不能够正确的认识投资的风险,盲目入市,最后导致投资失败。其失败的重要的原因是投资者本身缺乏科学的投资策略,风险意识淡薄。如何将投资的资金进行合理的资产投资,能够在股市出现价格的较大波动时,投资者不至于遭受重大损失,或者能够保证一定的收益,成为众多投资者投资的核心问题。鉴于上述,本文在Markowitz均值-方差理论的基础上,结合Goldfarb和Iyengar是出的多因素因子模型,建立了本文的基于多因素收益预测的均值-方差投资组合优化模型,并选取一家业绩较好的封闭式基金作为研究对象进行实证分析。首先,本文就经典的投资组合理论进行概述,并同时对本文运用的主成分分析方法和多元线性回归方法进行了概述。其次,利用主成分分析的方法对选定的六个大盘指数进行整合,得到第一因子与第二因子,并应用到多因素因子模型当中,得到参数期望收益μ与影响系数V的估计值,同时,运用无偏估计的方法得到收益的残差方差D。之后利用2008-2011年的大成优选的资产组合的数据,得到资产组合的优化权重,并将其与基金的实际权重进行对比分析,得到优化的收益率要比实际收益率高。最后,总结全文并分析本文研究不足,确定下一步研究方向。
[Abstract]:In the late 80s and early 90s of last century, the securities market of our country began to form slowly. In the past twenty years, the overall scale of our securities has been gradually improved, and a more perfect system has been formed. However, the scale of the financial market is huge and very complex, and every investment activity has high risk, and the income of each asset is also high. Uncertainty. Many investors can not correctly understand the risk of investment, blindly enter the market, and finally lead to the failure of investment. The important reason for the failure is that the investors themselves lack scientific investment strategy and the risk consciousness is weak. How to invest the capital in a reasonable asset investment can be put into the stock market when the price fluctuates greatly. Investors will not suffer heavy losses, or they can guarantee certain profits, which will become the core issue for many investors.
In view of the above, on the basis of the Markowitz mean variance theory and the multi factor factor model of Goldfarb and Iyengar, this paper establishes the mean variance portfolio optimization model based on the multi factor income prediction, and selects a closed basis gold as the research object. The classic portfolio theory is summarized, and the principal component analysis method and multiple linear regression method used in this paper are summarized. Secondly, the first and second factors are obtained by using the method of principal component analysis to obtain the first and second factors, and apply to the multi factor factor model, and get the reference of the six factors. The expected value of the number of expected returns and the estimated value of the influence coefficient V. At the same time, using the unbiased estimation method to get the residual variance D. of the income, the optimum weight of the portfolio is obtained by using the data of 2008-2011 years' large optimal portfolio, and the comparison analysis is made with the actual weight of the fund, and the optimized return rate is better than the actual income. Finally, summarize the full text and analyze the deficiencies of this study, and determine the next research direction.

【学位授予单位】：东北大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：F832.51;F224

【参考文献】