A nonlinear smooth transition regression (STR) model of the demand for narrow money in Colombia is specified using monthly data for cash, prices, the scale variable (industrial GDP), the interest rate and the rate of depreciation, within the single equation framework allowed by the data. In comparison with linear error correction model, the nonlinear specification is highly superior according to the statistics. The dynamics described by this model matches both the magnitudes and the behavior of the aggregate demand for narrow money in Colombia during the sample period (1980.5-1998.11).
