摘要
在借鉴国外有关研究成果的基础上,建立了基于CPLC方法的Tobit模型构建指数跟踪,在目标函数中增加一个与组合权重绝对值总和成比例的惩罚参数,并利用上证180历史数据进行实证检验,通过跟踪基准指数能力、比较样本外绩效和样本外优越性指标,表明新模型不但调整稳定了优化过程,而且提高了样本外的跟踪效果.
This paper proposes constant penalty L_1 constraint method of Tobit based on index tracking model according to foreign literatures, and adds to the objective function a penalty proportional to the sum of the absolute values of the portfolio weights. Subsequently it implements this methodology with historical data of Shanghai Index 180 and tests achievements of our portfolio by tracking benchmark and comparing performance and superiority of out-of-sample data. New model not only regularizes and stabilizes the optimization process, but also improves the effect of out-of-sample tracking.
出处
《中南民族大学学报（自然科学版）》
CAS
2010年第4期127-130,共4页
Journal of South-Central University for Nationalities：Natural Science Edition
基金
国家自然科学基金资助项目(70672015)
关键词
指数跟踪
跟踪误差
惩罚参数
index tracking
tracking error
penalized parameter
