报告题目：Semiparametric Dynamic Portfolio Choice with Multiple Conditioning Variables

报告人：英国南安普顿大学教授Zudi Lu

时间：2015年12月24日  下午1：30

地点：汇贤楼122教室

个人简介：

ZudiLu is a Professor/Chair in Statistics, in Mathematical Sciences Academic Unitand Southampton Statistical Sciences Research Institute (S3RI) at University ofSouthampton, UK. His research interest includes nonlinear financial time seriesmodelling and financial statistics / econometrics; statistical inference &computation for nonlinear spatial/temporal modelling and prediction; appliedtemporal/spatial modelling for financial, environmental and socioeconomicrisks; non-parametric/semi-parametric modelling and statistical learning.

报告摘要：

Dynamicportfolio choice has been a central and essential objective for investors inactive asset management. In this paper, we study the dynamic portfolio choicewith multiple conditioning variables, where the dimension of the conditioningvariables can be either fixed or diverging to infinity at certain polynomialrate of the sample size. We propose a novel data-driven method to estimate theoptimal portfolio choice, motivated by the model averaging marginal regressionapproach suggested by Li, Linton and Lu (2015, Journal of Econometrics). Morespecifically, in order to avoid the curse of dimensionality associated with themultivariate nonparametric regression problem and to make it practicallyimplementable, we first estimate the marginal optimal portfolio choice bymaximising the conditional utility function for each univariate conditioningvariable, and then construct the joint dynamic optimal portfolio through theweighted average of the marginal optimal portfolio across all the conditioningvariables. Under some regularity conditions, we establish the large sampleproperties for the developed portfolio choice procedure. Both the simulationstudy and empirical application well demonstrate the finite-sample performanceof the proposed methodology.

（Joint work with Jia Chen,Degui Li and Oliver Linton）