مهدی مصطفوی

The purpose of this study is estimation of narrow money in the long run via ARDL and Johansen procedures. In doing so first econometric model has been investigated. For modeling narrow money two studies have been applied namely Rother (1999) and Bahmani Oskooee (1996). The applied model in this study deals with GDP, inflation, rates of return on foreign exchange and cars. In order to investigate the long run relationship among involved determinants in the model, cointegration test has been applied via ARDL and Johansen procedures. According to ltrace adjusted there is only one cointegration vector for the RM1 model. Then we derived the long run coefficients for RM1 model and concluded the same results as those of the ARDL test in the light of the theory basis. Although the ARDL and Johansen use very different techniques in estimation, the former employs ordinary least squares while the latter uses the maximum likelihood estimation method. Thus, one-to-one comparison of the magnitudes of the coefficients may not be appropriate and requires some caution. Hence there is a bit difference between two mentioned results in determinants parameters magnitudes. Explanation of the coefficients of determinants (GDP and inflation) in Johansen is more real and closer to theory in comparison to ARDL procedure.   JEL classification: C52