考虑到投资者通常采取安全第一的准则,采用跟踪偏差的下半概率作为跟踪风险的度量;而为在恰当描述证券收益分布的厚尾特性的同时克服机会约束对模型求解所造成的困难,假设风险资产的收益服从多元t分布,由此建立了新型金融指数跟踪模型.在分析所建立模型结构特性的基础上,文中还导出了该模型的解析最优解.实证结果表明了新模型的有效性和实用价值.
Considering that investors usually adopt the safety first principle, the lower partial probability of the tracking deviation is taken as the risk measure; to overcome the computational intractability induced by the chance constraint while properly reflecting the fat-tall phenomenon of security return distributions, it is assumed that the joint distribution of stocks' return follows the multivariate t-distribution. The new financial index tracking model is then derived under the above framework. After analyzing the structure property of the established optimization problem, we derive its explicitly optimal solution. Empirical results demonstrate the efficiency and practical value of our new tracking model.