An Application of an Error Correction Model with Higher Order Cointegrated Variables to the Demand for Money

Résumé

Abstract

This paper applies the maximum likelihood technique developed by Krishnakumar J. and E.H. Gueye (1998) to estimate the parameters of a money demand equation for Switzerland in which there are variable integrated of different orders and particularly of order greater than one (1). The procedure was implemented in MATLAB for estimating our empirical model. Our results turn out to be satisfactory with interesting economic interpretations. This confirms the relevance of the methodological approach considering the integration and cointegration of macroeconomic time series.