Forecasting gold prices in India using ARIMAX and machine learning algorithms

Abstract

Forecasting is a function in management to assist decision making. It is also described as the process of estimation in unknown future situations. In a more general term it is commonly known as prediction which refers to estimation of time series or longitudinal type data. The main object of this paper is to compare the traditional time series model with machine learning algorithms. To predict the gold prices based on economic factors such as inflation, exchange rate, crude price, bank rate, repo rate, reverse repo rate, gold reserve ration, Bombay stock exchange and National stock exchange. Two lagged variables are taken for each variable in the analysis. The ARIMAX model is developed to forecast Indian gold prices using daily data for the period 2016 to 2020 obtained from World Gold Council. We fitted the ARIMAX (4,1,1) model to our data which exhibited the least AIC values. In the mean while, decision tree, random forest, lasso regression, ridge regression, XGB and ensemble models were also examined to forecast the gold prices based on host of explanatory variables. The forecasting performance of the models were evaluated using mean absolute error, mean absolute percentage error and root mean squared errors. Ensemble model out performs than that of the other models for predicting the gold prices based on set of explanatory variables.