Estimating and Forecasting Modified Money Demand Function in Iran Given Economic Sanctions

Money demand is one of the key variables in the economy which is considered by the policymakers in determining monetary policies. In fact, the impact of monetary policies is transmitted from the money demand channel by the private sector to the real sector of the economy. Therefore, more accurate estimation and forecasting of this variable in terms of environmental factors could be helpful for the monetary policymakers. In the present study, the adjusted function of real money demand with respect to the variables of economic sanctions, economic uncertainties and underground economy by using Markov Switching Model for the period of 1979 to 2018 with two regimes of high money demand (a regime with greater y-intercept) and lower money demand (a regime with less y-intercept) were estimated. The Artificial Neural Network (ANN) method was used in order to forecast the money demand function. Then, in order to ensure the high predictive capability of the ANN method, the forecast was performed by Markov Switching Model. The results demonstrated that Measure of National Income and Output has a positive effect, housing returns (the proxy for interest rates) has a negative effect, exchange rate in both regimes has a negative effect, the volume of the underground economy in both regimes has a positive effect, economic uncertainties in both regimes have a negative effect and economic sanctions in both regimes have a negative effect on real money demand. Furthermore, the results of the forecast indicated that the ANN method has a higher predictive capability in comparison with Markov Switching Model.