On the Efficiency of the Markets for Gold and Silver. 1981.

This study is set in the period of the Hunt Brothers silver debacle and provides evidence on the question gold and silver market efficiency in the 1970’s. They provide early evidence that the variance of precious metals prices is not constant and that they contain a unit root, making OLS regression analysis unsuitable for examining their prices. This has been found to be the case in many studies between the 80’s and today.

In terms of market efficiency the two precious metals are shown to suffer from autocorrelation, that their prices are affected by past changes. However, when trading strategies are tested to see if it is possible to beat the market the positive risk adjusted returns are small. Even modest trading costs make the strategies unprofitable.

The authors contend that based on their findings the assets are more speculative than investment, especially during certain parts of the sample. With hindsight and what is known now about the Hunt Brothers activities they appear to have been correct.

Method: Goldfeld-Quandt, Runs Tests, Trading rules.

Data: Weekly Friday spot prices for gold (PM fix) and Silver (Handy and Hardman) January 1971 – December 1979

Full Citation: Solt, Michael E., and Paul J. Swanson. “On the Efficiency of the Markets for Gold and Silver.” Journal of Business (1981): 453-478.

Abstract: Gold and silver are precious metals that have come to be viewed as possible investment assets. From the investor’s point of view, it is important to know that gold and silver prices reflect the relevant information set at every point in time. A sequence of price changes for each metal is analyzed in this paper in order to establish if these markets are efficient in this sense. The distributions generating the price changes appear to exhibit characteristics of nonnormality and nonstationarity in the means and variances. There appears to be some positive dependence in the series, though it does not seem that investors can easily exploit this dependence. These results have important implications for the role of gold and silver as investment assets.