Estimating the demand for money in Canada: Does including an own rate of return matter?

Traditional studies estimating the long-run demand for real money in Canada assume that narrow money, or M1, bears zero interest. However if implicit interest has been paid, such interest should be taken into account in determining the opportunity cost of holding money. Using quarterly data over the period 1961:1-2000:3 we construct and employ a competitive own rate of return variable. Over 1961:1-1982:1, the conventional money demand model which omits an own rate of return performs well. Over the period 1982:2-2000:3, where the degree of competition in the banking industry increased, the conventional money demand model does not perform well, whereas inclusion of the own rate of return yields correct parameter estimates.