Kun Ho Kim, Yeshiva University
Project Description/Abstract
In this paper, we conduct simultaneous inference of the non-parametric part of a partially linear model in time series when the non-parametric component is a multivariate unknown function. To this end, we propose a method to construct a simultaneous confidence region of the multivariate function. Relevant asymptotic results are derived and the finite sample performance is examined in a simulation study. The developed methodology is applied to two examples in time series: the forward premium regression and a factor asset pricing model.
Co-Authors
Likai Chen, Washington University in St Louis
Tianwei Zhou, Washington University in St Louis