Nonlinear Cointegration, Misspecification and Bimodality

We show that the asymptotic distribution of the ordinary least squares estimator in a cointegration regression may be bimodal. A simple case arises when the intercept is erroneously omitted from the estimated model or in nonlinear-in-variables models with endogenous regressors. In the latter case, a solution is to use an instrumental variable estimator. The core results in this paper also generalises to more complicated nonlinear models involving integrated time series.

MoreLess

Year of publication:	2010
Authors:	Medeiros, Marcelo C. ; Mendes, Eduardo ; Oxley, Les
Publisher:	Rio de Janeiro : Pontifícia Universidade Católica do Rio de Janeiro (PUC-Rio), Departamento de Economia
Subject:	Cointegration \| nonlinearity \| bimodality \| misspecification \| instrumental variables \| asymptotic theory.

More details

Series:	Texto para discussão ; 577
Type of publication:	Book / Working Paper
Type of publication (narrower categories):	Working Paper
Language:	English
Other identifiers:	635887398 [GVK] hdl:10419/176060 [Handle] RePEc:rio:texdis:577 [RePEc]
Source:	EconStor

Persistent link: https://www.econbiz.de/10011807401