Supply and demand for gold.

摘要

The purpose of this dissertation is to analyze the major fundamental factors influencing the physical gold market. What distinguishes this study from other gold studies is its scope. One of the largest and most extensive historical databases of gold demand and supply variables has been compiled.;The dissertation is also unique in that it pays special attention to the supply side of the gold market, investigating how growing production has and will impact the market. A model of South African production is developed and unlike other gold studies, supply variables are incorporated in the reduced-form gold pricing models. In addition, the dissertation theoretically and empirically identifies the major factors influencing gold demand.;Supply and demand curves for each of the supply and demand components are estimated using a method similar to the Fair Method. The Fair method is a two stage least squares technique with an adjustment for autocorrelation.;Supply curve estimation results indicate there is a tendency towards low adjustment speeds. This is reflected in long run elasticities being greater than short run elasticities. In the case of South African production and Net Communist sales, both long and short run supply elasticities were inelastic. It was found that the long run supply curve for Other Country production is elastic.;On the demand side, adjustment speeds were found to be extremely high, meaning that there is very little difference between long and short run elasticities. In general, the demand elasticities were low in both the long and short runs.;In conclusion, the dissertation found that as Other Country gold production increases, as a percentage of total world production, this will cause the long run total supply curve to become more elastic. With demand being relatively inelastic, this implies that shifts in supply will impact the gold price rather than the physical quantity of gold. In contrast, with supply becoming more elastic, shifts in demand will impact the physical quantity of gold rather than the dollar gold price. In addition, it is shown that the introduction of supply variables in the reduced form gold pricing models significantly improves each model's explanatory performance.