Forecasting the Occurrence of Business Cycles Using Band-Pass Filter in Iran's Economy

The aim of this paper is to study and forecast the business cycles in Iran’s economy in the period 1370-1392. For this purpose, quarterly data are used, and the business cycles that have occurred are extracted by using the Band-Pass filter. Then, in order to predict the business cycles, logit and probit regression methods are used. According to the nominal and real indices that affect the business cycles, the variables used in this research include oil revenues, government spending, inflation, issued building permits, and the imports of intermediate and capital goods. The results indicate that if oil revenues and inflation increase, the probability of economic boom increases too. However, an increase in the number of issued building permits reduces the probability of a boom in the economy. Finally, imports of capital and intermediate goods increase the probability of a boom. A sample prediction showed that the model could classify the observations correctly in 95% of the cases. It is also concluded that, the sample predicting ability of all the models is the same. The model was also applied to the out-of-sample prediction for the period 1394-1392. The results indicate the ability of the model to deal with this kind of prediction.