Gold price prediction using the Box-Jenkins methodology

Abstract

Information on the speculation and trading of gold abounds. Investors are attracted to moving their funds to gold as guaranteed storage of wealth, while traders capitalize on the dynamism of the market to build capital. The ups and downs in the price of gold and other precious metals can be predicted with proven mathematical and artificial intelligent algorithms. This study used machine learning algorithm in the price prediction of gold over a ten-year period. Autoregressive Integrated Moving Average (ARIMA) model was used in the experiment, while Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE) evaluation metrics were used in the evaluation of the performance of the various ARIMA models. The results obtained in the study proved that ARIMA could achieve high prediction performance over the entire period of prediction. The best prediction outcome of 98.23% was obtained during the 52-week period.