Cointegration: Its Fatal Flaw and a Proposed Solution

23 Pages Posted: 8 Jan 2020

Multiple version iconThere are 2 versions of this paper

Date Written: December 17, 2019

Abstract

It has been argued that whenever regression models involve nonstationary and trending variables, the estimation methods appropriate to stationary series cannot be applied to such models and instead require cointegration techniques. Unfortunately, the extant methodology applied to cointegration is a trap: if the error term of a cointegration regression is made up of omitted relevant regressors, then, even though they are integrated to the same order, the dependent and the independent variables of the regression are not cointegrated! This paper presents a way out this dilemma by proposing a remedy based on time-varying coefficient (TVC) modeling that over-comes all the shortcomings described in the paper.

Keywords: integrated variable, cointegrated variables, cointegrating vector, nonstationary variable, time-varying coefficient model

JEL Classification: C20, C22

Suggested Citation

Swamy, P.A.V.B. and von zur Muehlen, Peter, Cointegration: Its Fatal Flaw and a Proposed Solution (December 17, 2019). Available at SSRN: https://ssrn.com/abstract=3505463 or http://dx.doi.org/10.2139/ssrn.3505463

P.A.V.B. Swamy (Contact Author)

Federal Reserve Board ( email )

20th Street and Constitution Avenue NW
Washington, DC 20551
United States

Peter Von zur Muehlen

Federal Reserve Board ( email )

10435 Hunter View Road
Vienna, VA VA 22181
United States
5713281783 (Phone)

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
61
Abstract Views
485
Rank
612,654
PlumX Metrics