Predicting the Stock Return Direction Using Artificial Neural Network: The Case of Amman Stock Exchange

This study aims at measuring the ability of artificial neural networks (ANNs) to predict the direction of one-yearahead stock return. It applies fundamental analysis on the Jordanian industrial and service corporations over the period 1999 to 2009. Five independent variables are used to achieve the objective of this study, which are earnings change, earnings level, the ratio of market value of equity to book value of equity, dividend yields, and return on equity. Descriptive measures, correlation analysis, and regression analysis are also used to explore data before (ANNs) are used to predict stock return. The above variables were modelled, as lag variables, with stock return. The findings of the study show that (ANNs) are able to predict stock return with accuracy of 80.2% in the training sample and 58.1% in the test sample. Keywords: .