多基准组合下跟踪误差最小化投资策略有效性：实证检验

发布时间：2018-04-13 01:08

本文选题：投资策略 + 跟踪误差最小化　；参考：《电子科技大学》2017年硕士论文

【摘要】：1952年Markowitz提出了均值-方差模型,为投资者提供了一种全新的思路,以此为基础,学者提出大量的投资策略,构建了现代投资理论体系[1]。随着中国金融市场日趋成熟,中国学者也越来越注重对投资策略的研究,希望寻求更有效的资源配置方式,获得更高的收益,承担更小的风险。本文在简要阐述均值方差投资策略,等权重投资策略及单一基准组合下跟踪误差最小化投资策略的基础上,将对一种能对原有投资策略进行优化的策略进行深入研究,即多基准组合下跟踪误差最小化投资策略。首先详细阐述策略模型并求出模型的解,运用数学方法论述该投资策略的优势。其次,将该策略运用到股票市场中,用模拟和实证的方法验证模型,分别以均值方差投资策略,等权重投资策略和单一基准组合下跟踪误差最小化投资策略为基准进行比较。通过实证检验本文得到结论主要有以下几点:1.当以模拟法进行实证检验后,多基准组合下跟踪误差最小化投资策略收益率较高,波动性较低,夏普比率普遍大于风险厌恶系数不同的均值方差投资策略。用现实数据进行实证检验时多基准组合下跟踪误差最小化投资策略依然展现出较好的投资效果,并随着原投资策略限制的增加,多基准组合下跟踪误差最小化投资策略优势越为明显。2.多基准组合下跟踪误差最小化投资策略与单一基准组合下跟踪误差最小化投资策略比较,其平均收益率在大多数情况下高于单一基准组合下的投资策略,具有较高的夏普比率和较低的波动性,展现出较好的投资效果。3.多基准组合下跟踪误差最小化投资策略在所有比较投资策略中换手率处于较低水平,现实可操作性强。4.运用实证数据绘制投资者财富线,多基准组合下跟踪误差最小化投资策略财富线处于较高位置,充分证明了多基准组合下跟踪误差最小化投资策略能为投资者获得更高的收益,有效提升投资者财富。综上所述,多基准组合下跟踪误差最小化投资策略能得到较好的投资业绩,是一项有效的投资策略。同时该投资策略为中国投资者及投资经理提供了一个新的思路。
[Abstract]:In 1952, Markowitz put forward the mean-variance model, which provided a new way of thinking for investors. On this basis, scholars put forward a large number of investment strategies, and constructed a modern investment theory system [1].With the maturation of Chinese financial market, Chinese scholars pay more and more attention to the research of investment strategy, and hope to find more effective way of resource allocation, gain higher income and take less risk.On the basis of a brief description of mean variance investment strategy, equal-weight investment strategy and tracking error minimization investment strategy under a single benchmark combination, this paper makes a thorough study on a strategy that can optimize the original investment strategy.That is, the investment strategy of tracking error minimization under multi-benchmark combination.Firstly, the strategy model is explained in detail and the solution of the model is obtained, and the advantage of the investment strategy is discussed by mathematical method.Secondly, the strategy is applied to the stock market, and the model is validated by simulation and empirical methods. The comparison is based on mean variance investment strategy, equal-weight investment strategy and tracking error minimization investment strategy under a single benchmark combination.Through the empirical test, the conclusion of this paper is as follows: 1.After the empirical test by simulation, the investment strategy with tracking error minimization under multi-benchmark portfolio has higher yield, lower volatility and higher Sharpe ratio than the mean variance investment strategy with different risk aversion coefficient.With the empirical test of real data, the investment strategy of tracking error minimization under multi-benchmark portfolio still shows good investment effect, and with the increase of the original investment strategy restrictions,The investment strategy advantage of tracking error minimization under multi-benchmark combination is more obvious. 2. 2.Compared with single benchmark portfolio tracking error minimization investment strategy, its average return rate is higher than that of single benchmark portfolio investment strategy in most cases.With a higher Sharp ratio and lower volatility, showing a better investment effect. 3.The tracking error minimization investment strategy under multi-benchmark combination is in the lower level of all comparative investment strategies, and the practical maneuverability is strong. 4.Using the empirical data to draw the investor wealth line, the tracking error minimization investment strategy wealth line is in a higher position under the multi-benchmark portfolio, which fully proves that the tracking error minimization investment strategy under the multi-benchmark portfolio can obtain higher returns for investors.Effectively enhance the wealth of investors.To sum up, the tracking error minimization strategy under multi-benchmark portfolio can obtain better investment performance, and it is an effective investment strategy.At the same time, the investment strategy for Chinese investors and investment managers to provide a new thinking.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F832.51

【相似文献】