Title: Application of GMDH-type neural network to stock price prediction of Iran's auto industry

Authors: Vahab Bashiri; Hamidreza Salehi; Aein Ghorbani Ghashghaeinejad; Masoumeh Bashiri

Addresses: Department of Accounting, Imam Khomeini International University, Ghazvin, Iran ' Department of Accounting, Tehran University, Tehran, Iran ' Department of Management, Islamic Azad University, Firoozkuh Branch, Tehran, Iran ' Department of Petroleum Engineering, Petroleum University of Technology, Abadan, Iran

Abstract: There are many backward and forward linkages between automotive industry and other industries because of technical and structural properties of that. After the oil industry, the automotive industry is one of the largest and most profitable industries in Iran. Also, this industry has received government extensive attention in order to become competitive. Thus, development of Iranian economy is closely related to the competitive advantage of Iranian automotive industry. So, owing to the importance of this matter, we have tried to develop a Group Method of Data Handling (GMDH) type neural network for stock price prediction of automotive industry. For stock price prediction by the GMDH-type neural network, we are using earnings per share, prediction earnings per share, dividend per share, price-earnings ratio, earnings-price ratio as input data and stock price as output data. For this work, data of eight automotive companies are gathered from Tehran Stock Exchange. The GMDH-type neural network is designed by 80% of the experimental data. For testing the appropriateness of the modelling, reminders of primary data were entered into the GMDH network. The results are very encouraging and congruent with the experimental results.

Keywords: GMDH; group method of data handling; neural networks; stock price indicators; stock price prediction; stock markets; stock price forecasting; automobile industry; automotive firms; modelling.

DOI: 10.1504/IJBFMI.2016.080131

International Journal of Business Forecasting and Marketing Intelligence, 2016 Vol.2 No.4, pp.359 - 378

Received: 30 Jun 2016
Accepted: 29 Jul 2016
Published online: 03 Nov 2016 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article