[12月4日] 计量金融与大数据分析workshop

发布日期:2020-12-01 05:35    来源:

时间:2020年12月4日(周五)10:00-11:30 am

地点:北京大学经济学院107教室

主持人: (国发院)沈艳、黄卓、孙振庭、张俊妮、胡博

              (经济学院)王一鸣、王熙、刘蕴霆

主讲人:王斌

题目:Approximate Semiparametric Maximum Likelihood Estimation of Recurrent Diffusion Models

报告摘要:

In this paper we consider semiparametric estimation of recurrent diffusion models, where drift functions are parametric and contain unknown finite dimensional parameters, while diffusion functions are left unspecified. Our estimate is an approximate MLE using Euler scheme, with some appropriate first-step kernel type estimate of diffusion function plugged in. We establish the consistency and asymptotic distribution of our semiparametric estimator. Our asymptotic results are developed under the sampling scheme with sampling interval shrinking to zero and time span increasing to infinity, and without assuming stationarity of the underlying diffusion process. A simulation study is provided to examine the finite sample performance of our estimator.

主讲人简介:

王斌,助理教授,哈尔滨工业大学(深圳)经济与管理学院