主题: Generalized Functions and robust M-Estimation with Application on Cointegrating Multiple Index Time Series Models

Abstract：

This paper studies robust M estimation for an additive single-index model for cointegrating time series where usually a non-smooth loss function with kinks are used; examples include least absolute deviation estimator, quantile estimator, and Huber's estimator. We derive asymptotic properties of M estimators while nonsmoothness of the loss functions have been a source of technical difficulties in developing a unified method for deriving the properties of the estimators. The family of generalized functions offers an elegant way of approximating nonsmooth functions by smooth functions, and thereby overcoming much of the technical difficulties. We evaluate the finite-sample performance of the proposed estimation method and theory by both simulated data and an empirical analysis of predictive regression of stock returns.

报告时间：2023年6月28日下午14:00

线下地点：威廉希尔6-210

腾讯会议ID:  794-284-973

嘉宾简介：

董朝华，中南财经政法大学统计与数学学院教授，主要研究领域为高维计量理论、非参数与半参数方法、非平稳时间序列和面板数据模型、微观计量应用，其研究成果发表在Journal of Econometrics、Journal of Business and Economic Statistics、Econometric Theory、《系统工程理论与实践》等国内外一流经济学期刊，主持多项自然科学基金，中国系统工程学会金融系统工程专业委员会理事，并担任Journal of Econometrics, Econometric Theory, Econometric Reviews, Annals of Statistics, Journal of Business and Economic Statistics等期刊匿名评审员。