Forecasting earnings of firm’s listed in ASE using ARIMA model

Abstract

The study aims to estimate and forecast earnings of the firms listed in Amman Stock exchange (ASE) using a time series data of earning per share (EPS) for the period from 1978 till 2016.  The data has been extracted from firms’ annual reports.  A wavelet Transform (WT) decomposes the data and detects the fluctuations and outlay values.  The parameters p, d, and q are estimated using the ARIMA model, the results show that the ARIMA models accuracy criteria MASE and RSME have the lowest values (0.7089 and 0.0709) respectively, thus the forecasting accuracy is high.  It is concluded that firms’ earnings show slow increasing trend for the upcoming 38 financial years.