Forecasting Ammonia Price Based on Fundamental, Technical Analysis and Neural Network

In order to get future information about economic variables, economists have discovered new methods. Large reserves of oil and gas in the Middle East and especially Iran have obtained advantages for these countries. Therefore, the purpose of this study is modeling and forecasting the price of petrochemicals in the Middle East through GMDH neural network method. We focus on a single product, ammonia, which is a petrochemical with natural gas feedstock for which two approaches are taken: fundamental analysis for long-term and technical analysis for short-term. The findings suggest that the natural gas price have important effects on the price of ammonia in the Middle East, despite low natural gas price for ammonia producers in this region. Modeling through technical analysis is satisfactory because of smooth trend line of ammonia price. In the end, it is shown that the GMDH neural network has better predictive power than ARIMA method in predicting ammonia price based on error criteria.