Forecasting Gold Price: Evidence from Pakistan Market

Abstract

In our day to day life, predictability of gold’s prices is significant in many domains such as economics financial and political environment. The main objective of this research is to forecasts the prices of gold, making use of ARIMA and two distinct versions of wavelet scheme. The monthly data consists of 221 observations starting from Dec 2005 to April 2013, has been used in this research. After evaluating the accuracy of these models by mean absolute error and mean square error, it turns out that wavelet neural transformation has better prediction accuracy than rest of the models. Also, this study utilizes the return forecasts from the above mentioned different models in a simple trading strategy and compare pay offs in order to know as to which framework serves a better forecasting model.

Keywords: Gold Price, ARIMA, Wavelet, Multiple Regression, Wavelet Neural Transform, Error Analysis