基于分层线性模型的投资组合分析  

Portfolio Analysis Based on Hierarchical Linear Model

作　　者：何晓群[1,2] 刘达通 

机构地区：[1]西京学院应用统计科学研究中心,陕西西安710123 [2]中国人民大学统计学院,北京100872 [3]中国工商银行数据中心,北京100096

出　　处：《当代经济科学》2015年第2期57-61,126,共5页Modern Economic Science

摘　　要：投资就是投资者当期投入一定金额的资金,以期在未来获得能够补偿某些因素的回报,这些因素包括投资资金被占用的时间、预期的通货膨胀率和未来收益的不确定性,其中不确定性被视为风险。在经济飞速发展的今天,越来越多的人们开始熟悉股票、债券、基金、期权等金融产品,即使对其中的运作原理不甚明白,也清楚这些金融产品能够带来收益,同时伴随着风险。不管是机构投资者还是个人投资者,都希望在追求高收益的同时能够有效控制风险。This paper conducts portfolio analysis based on hierarchical linear model and Sharpe single-index model. The model describes the relationship between each kind of asset and market index and seeks overall optimal solution by using simulated annealing algorithm through the target function to be optimized of Sharpe ratio structure model. This paper undertakes empirical analysis by using the data of actual stock market and draws the following main conclusions: 1. The models of this paper can accurately picture the relationship between a single stock and 300 index of Shanghai Stock Exchange and Shenzhen Stock Exchange. 2. The portfolio Sharpe ratio of hierarchical linear model is higher than that of linear regression model in training set and test set and this advantage is more significant in test set.

关 键 词：分层线性模型 夏普单指数模型 组合优化 

分 类 号：F830.59[经济管理—金融学] F224