基于资产选择的投资组合模型与中国股市实证检验

Portfolio selection based on asset selection and the empirical study in Chinese stock market

Lars-Lasso回归算法是近年来统计选元的一个新兴方法.针对资产数量众多的投资市场,本文将Lars-Lasso方法运用到资产配置的第一步资产选择中,针对已选择的小数量资产进行资产组合配置.对中国沪深股市股票进行实证分析.结果证明,在Lasso选元下得到的投资组合总体表现优于市场指数.

Lars-Lasso regression algorithm is a popular statistical element method. For the investment markets with large amount of assets,we apply the Lars-Lasso method to the asset selection,which is the first step of portfolio selection,and then use the portfolio optimization model.The utility of this approach is illustrated by empirical studies on Chinese stock market,and it is verified to have better performance than the market index.