Academics have found it particularly challenging to develop parsimonious models that can accurately model and forecast gold price. One of the reasons could be the complex nature of gold market fundamentals. Arguing that the key to forecasting gold prices lies in analyzing the factors that generate investment demand (as opposed to gold supply and fabrication demand), we empirically validate a model that factors in consumer sentiments, interest rates, returns on stock market, and oil prices. Our findings indicate that although gold price has significant correlation with all these four variables, interest rate and consumer sentiments are only significant predictors of the gold price.
Advances and Applications in Statistics,
Vol. 13, no. 2 (Dec 2009), pp. 225-239