Speaker：LI Kunpeng（Professor, International School of Economics and Management, Capital University of Economics and Business）
Description：This paperconsider the estimation and inferential issues of threshold spatialautoregressive model, which is a hybrid of threshold model and spatialeconometric model. We consider using the quasi maximum likelihood (QML) methodto estimate the model. The asymptotic theory of the QML estimator isestablished under the setup that the threshold effect shrinks to zero alongwith an increasing sample size. Our analysis indicates that the limitingdistribution of the QML estimator for the threshold value is pivotal up to ascale parameter which involves the skewness and kurtosis of the errors due tothe misspecification on the distribution of errors. The QML estimators for theother parameters achieve the oracle property, that is, they have the samelimiting distributions as the infeasible QML estimators, which are obtainedsupposing that the threshold value is observed a priori. We also consider thehypothesis testing on the presence of threshold effect, and the hypothesistesting on the threshold value equal to some prespecified one. We run Montecarlo simulations to investigate the finite sample performance of the QMLestimators and find that the QML estimators have good performance.
Time：May 14，2019（Tuesday）,14:00-15:30
Venue: Shahe Main Teaching Building, room 501
