Analysis of Most Important Variables Affecting TEPIX and Modeling Them with Artificial Neural Networks and Comparing Results with Technical Analysis and Elliott Waves

Subject Areas : Financial engineering

Mohammad Kamravafar ¹ (Student of industrial management (MA) at science and research branch, Islamic azad university, Tehran, Iran)
S. Zabihollah Hashemi ² (Assistant Professor at industrial management, Tehran central branch, Islamic azad university, Tehran, Iran)

Received: 2016-10-01 Accepted : 2016-11-14 Published : 2017-04-21

Keywords: Artificial neural networks (ANN), Technical Analysis, TEPIX, Elliot Waves,

Abstract :

The main goal of this research is to studying an identifying the main influencing variables on the TEPIX (Tehran Stock Price Index) and modelling them using artificial neural networks and comparing results with technical analysis and Elliot waves. Independent variables used are dollar exchange rate, inflation, GDP, unemployment and liquidity and dependent variable is TEPIX. In this study, artificial neural networks (NLP and GMDH), technical analysis tools (Elliot waves and regression channel) are used that they show between independent variables in GMDH, unemployment is unneeded variable and have low influence, but others have high influence in the model. Further the study shows that technical analysis and artificial neural networks may have same results, but ANN have more power to predict the TEPIX