Empirical Modeling for the Spot Price of Gold Based on Influencing Factors

Abstract

In light of the special roles of the price of gold on the technological and economic development as well as social aspects of human society, it is of great importance and necessity to develop a series of statistical models that, based on sound reflection of the current structure of the gold market, are able to provide valuable references for trends of the gold price. In fact, the gold price is influenced by a variety of economic factors. For forecasting purposes, it is useful to study the short- and long-run effect of direct and indirect economic factors towards the supply and demand of gold and thus the spot price of gold. To this point, this paper focuses on analyzing the individual and mutual impact of influencing factors on the gold price and, simultaneously, providing a short-term forecast of the price of gold based on the relationships with other major macroeconomic variables. While these relationships are modeled with multiple regression, forecasts of individual factors are obtained under multivariate time series models. The forecasting results are found to be accurate in a relative sense, confirming the significant impact of the factors chosen while indicating the validity of the modeling idea applied.