Forecasting financial time series data base on wavelet transforms and ARIMA model

Abstract

This article suggests a novel technique for forecasting the financial time series data based on Wavelet transforms and ARIMA model. The financial data are decomposed via Haar Wavelet transforms. Then, the future observations of this series are forecasts using a suitable and best fitted ARIMA model. Daily prices from Amman Stocks Market (Jordan) from 1992 until 2008 are used in this study.