While researching how Dow Jones re-weights its AIG Commodity index, we came across the following. It’s a discussion of the merits and demerits of using production data and liquidity as measures of a commodities relative importance in the global marketplace. We’ve included it for educational purposes, and its surprising but not unwelcome recognition of gold’s importance in the global monetary system (emphasis added is ours):
Liquidity, or the relative amount of trading activity centered on a particular commodity, is an important indicator of the value placed on that commodity by financial and physical market participants. Liquidity provides a window on the commercial significance of a commodity. The DJ-AIGCI relies on data that is both exogenous to the futures markets (production) and endogenous to those markets (liquidity) in determining relative weightings.
Production data, although a useful measure of economic importance, may underestimate the economic significance of storable commodities (e.g., gold) at the expense of relatively non-storable commodities (e.g., live cattle). Production data alone also may underestimate the investment value that financial market participants place on certain commodities.
Gold clearly illustrates the potential shortcomings of exclusive reliance on production data and the greater balance provided by reliance on liquidity data. Since time immemorial, gold has played a unique role in the world economy, which is not effectively captured by current production data. For example, although only 247 metric tons of gold were produced in 2004, approximately 32,000 metric tons were held as official government reserves. Of the approximately 150,000 tons of gold mined since the dawn of recorded history, approximately 85%, 127,500 metric tons, is still held by central banks and non-governmental entities in bullion, coin, and jewelry form.
Based on this data, a production-based ranking of commodities would contain a small amount of gold (2.5%), and large amounts of non-storable commodities, such as live cattle (11%). A doubling of the price of gold may be a more significant global economic event than a 25% increase in cattle prices, yet the two events would have a similar impact on a production-weighted index. Primary reliance on liquidity data as a weighting measure reduces this type of potential distortion.