Multiple Regression Time Series Model for forecasting Gold Prices

  • Tanmay Aggarwal, Debapriya Nayak, Somya Agarwal, Parul Gupta, Gauri Anand


Prediction of price is an important aspect of the finance domain. This urge to create commercial value makes developing fitter predictive model’s necessary. Gold Price is considered an easy liquid investment and is a popular investment by people in rural regions. Gold is a commodity investment which requires a Global outlook for safe investment and better returns. Multiple linear regression is statistical tool which helps to determine relationship between many variables. In the regression one variable acts as a dependable variable while all the other variable act as an independent variable. The forecasting based on multiple linear regression helps in the prediction on values of the dependent variable. The study considers Gold price as the dependent variable and the objective is to develop a multiple linear regression-based forecasting model based on financial factors such as - Crude oil price, interest rate (repo rate), Indian currency price in dollars, Sensex (BSE), Inflation rate and US Dollar index. Any sudden increase in demand can be traced with the help of such model as it will be being an odd condition as compare to the structure and pattern of the forecasted gold price. This study will help us in developing such a model for the gold price.