Seminar | Kunpeng Li, Capital University of Economics and Business

Seminar Vol. 148

Title: Threshold Spatial Autoregressive Model

Speaker: Kunpeng Li, Capital University of Economics and Business

Time: April 29th, 2019 13:30–15:00

Venue: Conference Room 106B, Zhonghui Building (IESR, JNU College of Economics)

About the speaker:

Kunpeng Li is a Professor and Associate Dean of the International School of Economics and Management, Capital University of Economics and Business (China). He earned a Ph.D. in Economics from Tsinghua University in 2011. Kunping Li's main research interest lies in the field of Econometrics, and his works have been published in a number of international journals including Annals of Statistics, Journal of Business & Economic Statistics, Journal of Econometrics, and Review of Economics and Statistics. The economist is currently an Associate Editor of the Journal of Business & Economic Statistics.

Abstract:

This paper considers the estimation and inferential issues of threshold spatial autoregressive model, which is a hybrid of threshold model and spatial econometric model. We consider using the quasi maximum likelihood (QML) method to estimate the model. The asymptotic theory of the QML estimator is established under the framework that the threshold effect shrinks to zero along with an increasing sample size. Our analysis indicates that the limiting distribution of the QML estimator for the threshold value is pivotal up to a scale parameter which involves the skewness and kurtosis of the errors due to the misspecification on the distribution of errors. The QML estimators for the other parameters achieve the oracle property, that is, they have the same limiting distributions as the infeasible QML estimators, which are obtained supposing that the threshold value is observed a priori. We also consider the hypothesis testing on the presence of threshold effect, and the hypothesis testing on the threshold value equal to some pre-specified one. We run Monte carlo simulations to investigate the finite sample performance of the QML estimators and find that the QML estimators have good performance.