old and silver have been long considered close substitutes in precious metal G markets. This study, based on rational expectation framework, demonstrates that there is a short-term panty between the gold and silver prices. The stability of this panty allows one to earn above-average returns on a frequent basis before transaction costs.
Traders of precious metals commonly believe that there is an equilibrium parity between the gold and silver prices over the long run.' Assuming markets are consistently moving toward equilibrium, profitable strategies would generally follow this ruie: If actual parity drops below the equilibrium level,* buy gold and sell silver; if the parity is currently above the equilibrium level, sell gold and buy silver.
The practioners' (McAlvany, 1981) rationale for such a fixed parity is as follows:
Since gold and silver are regarded as close substitutes in a typical portfolio, the demands of both metals are affected by similar factors. The long-term equilibrium price is therefore determined by the ultimate fixed stock reserves of the two metals.
Even if the rationale of this all-time fixed parity, for example, 30: 1, is intuitively appealing, the caveats are two-fold. The ex-post derivation of this particular level is based on historical observations, which cannot be used to justify the applicability Throughout history, gold and silver have fluctuated in relationship to one another. In ancient Egypt, the gold silver ratio was 1: 1. During a subsequent 2000-year period, the ratio was 13.5: 1. In 1837 the ratio was officially set by the US Congress at 16.1; in 1896, when William Jennings Bryan w a s making his impassionate plea for free coinage of silver, the world ratio was 32.5:l. For a more detailed historical review of the ratios, see McAlvany (1981). For the implied trading strategy, also see Lang (1983).
T h e parity is defined by the ratio between the spot gold price per ounce and the spot silver price per ounce.
The author would like to thank the helpful comments from Jeny Wisler, Lawrence Conway, M. E. Ellis, and two anonymous reviewers. The usual disclaimers apply.