Kointegrasi dalam ekonometrlka

Abstract

Economics researchers use frequently OLS regression for time series variables without considering if the time series are stationary. These time series would provide many problems (challenges) in regression analysis. We may avoid the problems by transforming the series 10 be stationary and then use the appriate analysis. This paper explains the problems and how to solve it. The conclusion of this paper is that we must consider the stationary of time series before using OLS regression. If there is stationary, we must transfor it first and then use cointegrated regression.