Knowledge that Transforms

This paper is directed to the goldexchange world which existed before the advent of the two-tier gold market in March 1968. It considers the choice between dollars and gold as international reserves and views the economic forces underlying the substitution of dollars for gold as significant in determining the viability of a dollar-exchange standard. Gold and dollars are both viewed as nonsterile assets with an expected return on gold and a certain return on dollars. In the case of dollars and gold, gold is often labelled as the sterile asset. This is incorrect since a substantial and variable forward dollar premium on gold futures has been paid in an active Zurich market, indicating a positive expected return on gold. This market quoted a price of gold six months forward, in terms of an annual premium on the market price, from 1961 to 1968. An individual willing to pay a premium in order to purchase sterile gold at about $35 an ounce six months in the future indicates that he attaches a positive probability to the chance of a rise in the dollar price of gold. Otherwise, why forego the opportunity to invest in interest-yielding assets? The average premium, from the second quarter of 1961 through the first quarter of 1968 was 6.39 percent. The forward premium on gold as an index of the expected return on gold is a crucial factor, in the empirical analysis below, of the choice by central banks between dollars and gold. It proves to be a significant independent variable. If we restrict consideration to short-term dollar liabilities (ninety days) then we have an asset (dollars) with a certain positive return (neglecting capital gains on ninetyday dollar assets) compared with an asset (gold) with an expected positive return. The choice between dollars and gold is cast in terms of independent, utilitymaximizing portfolio behavior. This is contrary to the view expressed by Fritz Machlup among others, that the holdings of reserves by officials are entirely politically determined: