投资收益评估与投资收益的非线性研究

发布时间：2018-05-05 22:42

本文选题：资产 + 资金管理　；参考：《东北师范大学》2013年博士论文

【摘要】：为了更好地比较投资前景(如投资组合,固定资产等)的表现,在本文中我们首先探索并发展了剔除背景风险影响的马科维茨均值-方差比率统计量。我们考察了马科维茨均值-方差比率的大样本和小样本检验的理论和表现,在非渐进框架下(小样本情况下),获得了一个基于马科维茨均值-方差比率统计量的一致最优无偏检验。我们讨论了基于马科维茨均值-方差比率统计量的大样本检验和小样本检验的应用性,并且实证分析了剔除以美国标准普尔500股指收益为背景风险后的韩国股指收益和新加坡股指收益的差异,从而展现了我们获得的基于马科维茨均值-方差比率统计量的一致最优无偏检验在实际应用中的优越性。特别要提出的是,我们发现了基于马科维茨均值-方差比率的小样本检验能够比较出两个不同投资收益表现的差异,但是常用的基于夏普比率的检验以及我们获得的基于马科维茨均值-方差比率的大样本检验却不能比较出这些收益表现的差异。另一方面,为了选择好的投资项目来提高收益,对于单个的投资收益序列的研究,尤其是近些年来对其非线性特性的研究,越来越引起人们的兴趣。在本中,我们提出了一个高效,便捷的方法来探测一个时间序列是否存在非线性性质。我们的非线性检验的一个优点是,不需要假定非线性的具体形式。模拟结果显示我们的检验方法在稳定性和势方面都有很好的表现。应用在太阳黑子数据非线性检测得到了与现存的优秀工作一致的结果；应用于检验股票价格的随机游动假设方面也证实了最近发现的结果。
[Abstract]:In order to better compare the performance of investment prospects (such as portfolio, fixed assets, etc.), we first explored and developed the Markowitz mean variance ratio statistics that eliminated the impact of the background risk. We examined the theory and performance of the large sample and small sample test of the Markowitz mean variance ratio in the non progressive framework. Under the case of small sample, a uniform optimal unbiased test based on the Markowitz mean variance ratio statistics is obtained. We discuss the applicability of the large sample test and the small sample test based on the Markowitz mean variance ratio statistics, and the empirical analysis is taken to eliminate the background wind of the US Standard & Poor's 500 stock index return. The difference between the return of the South Korean stock index and the earnings of the Singapore stock index shows the superiority of the uniform optimal unbiased test based on the Markowitz mean variance ratio statistics. In particular, we find that the small sample test based on the mean variance ratio of the Markowitz can be compared. There are two differences in the performance of different investment returns, but the commonly used test based on the SHARP ratio and the large sample test based on the Markowitz mean variance ratio can not compare the differences in the performance.
On the other hand, in order to choose a good investment project to improve the income, the study of a single investment return sequence, especially the study of its nonlinear characteristics in recent years, has attracted more and more interest. In this paper, we have proposed a efficient and convenient method to detect the existence of nonlinear properties of a time series. One advantage of the nonlinear test is that there is no need to assume the specific form of the nonlinearity. The simulation results show that our test method has a good performance in both stability and potential. The application of the nonlinear detection of sunspot data is consistent with the existing excellent work; it is used to test the random walk of the stock price. The hypothesis has also confirmed the recent findings.

【学位授予单位】：东北师范大学
【学位级别】：博士
【学位授予年份】：2013
【分类号】：F830.59;F224;O212.1

【相似文献】