Shahid Beheshti UniversityJournal of Economics and Modelling2476-5775414-1520131122Forecasting Iranian’s Economic Growth Using Mixed Frequency Data Sampling TechniqueForecasting Iranian’s Economic Growth Using Mixed Frequency Data Sampling Technique12353502FAMohammad Noferestiassociate Professor of Economics, Shahid Beheshti UniversityMahboubeh BaiatM. A. Student, Shahid Beheshti UniversityJournal Article20150214Economic growth, measured by Gross Domestic Product rate of growth in this paper, is the most single indicator revealing the overall performance of the economy. Forecasting economic growth helps the policy makers to visualize the future state of the economy and undertake some policy actions if necessary. In this paper we use the recently introduced method by Ghysels, Santa-Clara and Valkanov (2004) to forecast economic seasonal growth rate. This method, which is named Mixed frequency Data Sampling technique (MIDAS), facilitate the use of variables with different high and low frequencies in on regression. The presence of high frequency variables in the regression equation allows us to revise the previous forecasted value as soon as new data for the high frequency variable is released. Comparing the forecasts made by the regression with that of the retained growth rate observations, indicate that the forecasted values are very accurate. An earlier prediction of economic growth rate for the fall of 1393 by the model is to be 1.8%. But as new monthly data for the high frequency explanatory variables became available, the forecasted value was revised to be 1.5%. We predict the GDP growth rate for the winter of 1393 to be -0.11%. In such a case the Iranian annual economic growth rate would not be more than 2.3 percent relative to the year 1392. Economic growth, measured by Gross Domestic Product rate of growth in this paper, is the most single indicator revealing the overall performance of the economy. Forecasting economic growth helps the policy makers to visualize the future state of the economy and undertake some policy actions if necessary. In this paper we use the recently introduced method by Ghysels, Santa-Clara and Valkanov (2004) to forecast economic seasonal growth rate. This method, which is named Mixed frequency Data Sampling technique (MIDAS), facilitate the use of variables with different high and low frequencies in on regression. The presence of high frequency variables in the regression equation allows us to revise the previous forecasted value as soon as new data for the high frequency variable is released. Comparing the forecasts made by the regression with that of the retained growth rate observations, indicate that the forecasted values are very accurate. An earlier prediction of economic growth rate for the fall of 1393 by the model is to be 1.8%. But as new monthly data for the high frequency explanatory variables became available, the forecasted value was revised to be 1.5%. We predict the GDP growth rate for the winter of 1393 to be -0.11%. In such a case the Iranian annual economic growth rate would not be more than 2.3 percent relative to the year 1392. https://ecoj.sbu.ac.ir/article_53502_726dde94c9992fce4d194687bda40e5f.pdf