The gold standard is typically analyzed as a fixed exchange rate regime. While the existence of a gold-point spread is recognized (if only implicitly), the location of the exchange rate is deemed irrelevant as long as the rate remains within the spread. The common procedure, then, is simply to represent the exchange rate by the mint parity, a practice justified by long-term automatic adjustment processes that eliminate balance of payments disequilibria that can exist only at (or beyond) a gold point. In particular, the interwar dollar/sterling gold standard is universally viewed as involving a pound substantially overvalued at the mint parity but with corrective processes stymied by altered circumstances and policies compared to the prewar period, eventually leading to the demise of the gold standard. So the existing studies of the stability of the interwar gold standard are long-term in nature, making no reference to the gold-point spread or the movement of the exchange rate within it.

In this chapter a different focus is adopted. The functioning of a gold standard in general – with the dollar/sterling interwar experience as the case study – is analyzed from the standpoint of short-run efficiency in which the precise location of the exchange rate plays the central role. Two concepts of “efficiency” are utilized: market efficiency, involving the behavior of private parties in response to the profit opportunities afforded by the gold standard, and regime efficiency, pertaining to the probability of maintenance of the gold standard with the existing mint parity.

Gold-point arbitrage versus gold-effected transfer of funds

Both gold-point arbitrage (GPA) and gold-effected transfer of funds (GTF) involve purchasing specie in one country and transporting it across the ocean to be sold in the other country. Under GPA the indirect foreign-exchange transaction via the medium of gold is combined with a direct foreign-exchange transaction in the opposite direction, resulting in a profit. Under GTF the indirect foreign-exchange transaction via gold substitutes for a direct foreign-exchange transaction in the same direction. The gold and foreign-exchange transactions are complements under GPA, substitutes under GTF. The objective under GPA is to make a sure profit, that under GTF is to transfer funds from one country to the other cheaper than via a direct foreign-exchange transaction.

It is ironic that contemporary accounts of such operations almost always pertain to GTF, whereas modern textbooks deal exclusively with GPA. It is not so much that GTF and GPA are confused in authors' minds; rather, the two operations have somehow become merged as one. Only Einzig (1931, p. 18) distinguishes GTF from GPA:

Gold-point spread and asymmetry of gold points

Following chapter 9, the gold-export and gold-import points are, respectively, the cost of specie export (CX) and the negative of the cost of specie import (−CM), expressed in percentage terms (percent of parity). Then the gold-point spread (more generally, specie-point spread), also in percentage terms, is the difference of the gold points: CX − (−CM) = CX + CM, the sum of the specie-shipment costs. If the gold points are expressed as dollars per pound, GX and GM, then the gold-point spread again is derived as the difference between the gold points, this difference taken as a percentage of (true mint) parity: 100[(GX − GM)/M] = CX + CM again, following the notation of section 4 of chapter 8. This equivalent formulation of the spread justifies anew the representation of the gold-import point as the negative of percentage cost.

Of course, the gold points differ for GPA and GTF, whether for export or import. However, for both operations, the export point always differs in magnitude from the import point. The implication is that the gold points are not equidistant from mint parity. The mid-point of the gold-point spread, then, differs from mint parity. The spread mid-point (SM), expressed as the percentage deviation from parity, is half the sum of the algebraic gold points: SM = (CX + (−CM))/2 = (CX − CM)/2. Of course, parity itself in this configuration is zero [100.{(M − M)/M} = 0].

The model

In 1929 the greatest historian of the dollar–sterling exchange rate wrote:

Since that date, the issue on which researchers have concentrated using dollar–sterling exchange-rate data is integration in the foreign-exchange market. This type of analysis was considered in chapter 7. Historians have used gold-point data to explore the same issue, but in a compartmentalized fashion, without reference to the exchange rate. This work was examined and developed in chapter 10.

One problem with the research of previous authors is the use of inappropriate exchange-rate or gold-point data – a defect corrected in chapters 7 and 10 through the employment of the data generated in this study. Other limitations, however, are methodological in nature and retained in the earlier chapters.

First, it is not stated why the gold-point spread itself rather than some transformation is the appropriate measure of external integration. Second, internal integration – the logical complement to external integration – is not even considered conceptually, let alone measured.

Unit of account

The pound, shilling, and pence schemata as the unit of account began prior to the Norman Conquest, in the Anglo-Saxon period, when the only coin in existence was the silver penny. With rare and temporary exception, the penny was the only coin in circulation until the late thirteenth century. The term “penny” is believed to emanate from a coin issued by King Penda of Mercia in the seventh country. In the ninth century the penny was in all the Saxon kingdoms and 240 pennies always constituted a pound. Even though the weight of the penny varied across kingdoms and over time, “the intention was that the pound weight of silver and the pound of money should be the same and that the pound of silver should be minted into 240 pennies” (Feavearyear, 1963, pp. 7–8). The abbreviation for penny, d., is from the Latin denarius (a Roman coin and unit of account), which until the fourteenth century meant both a pennyweight and the face value of the penny coin.

According to Feavearyear, the Saxons may have applied the term “scilling,” later “shilling” and meaning “a piece cut off,” to pieces of broken silver added to the scale to compensate for underweight coin. Over time, the shilling became equal to a definite number of pennies, eventually 12, the preference of William the Conqueror.

The main theme of this volume is that, under a gold (or any metallic) standard, the precise location of the exchange rate is of importance – and this is so whether the exchange rate is within the gold-point spread (the range of the exchange rate delimited by the gold points) or outside the spread. This viewpoint is contrary to the traditional approach, which is dichotomous in nature, checking only whether the exchange rate is inside or outside the spread. The traditional approach typically analyzes the gold standard as a fixed exchange-rate regime. While the existence of a gold-point spread is recognized (if only implicitly), the location of the exchange rate is deemed irrelevant as long as the rate remains within the spread.

The common procedure, then, is simply to represent the exchange rate by the mint parity, a practice justified by long-term automatic adjustment processes that eliminate balance-of-payments disequilibria that can exist only at (or beyond) a gold point. For example, the interwar dollar–sterling gold standard is almost universally viewed as involving a pound substantially overvalued at the mint parity but with corrective processes stymied by altered circumstances and policies compared to the prewar period, eventually leading to the demise of the gold standard. So the existing studies of the stability of the interwar gold standard are long run in nature, making no reference to the gold-point spread or the movement of the exchange rate within it.

In this book a different focus is adopted.

Institutions

1. A country's domestic and international monetary standards – be they metallic (gold, silver, or bimetallic) or paper – can differ, as can its effective and legal domestic standard (sections 1 and 4 of chapter 2).

2. A gold (or other metallic) standard involves both domestic and international conditions (sections 2 and 3 of chapter 2).

3. The monetary system of colonial times continued to exist in the United States to the mid-1790s, with no effective American mint until 1794 (sections 1 and 2 of chapter 3).

4. The gold value of the dollar was unchanged from 1837 to 1934 (section 2 of chapter 3).

5. The United States was formally bimetallic until 1873–1874, when silver was legally demonetized (section 3 of chapter 3).

6. Several kinds of US government or central bank paper currency convertible into coin on demand were authorized, between 1861 and 1913 (section 4 of chapter 3).

7. The important US Mint Acts, that determined coinage of private bullion and provision of gold bars, were in 1792, 1834, 1837, 1853, and 1882 (sections 5 and 6 of chapter 3).

8. Banks were a late development in American history, and were of four types: federal, private, state, and national (section 7 of chapter 3).

9. Disregarding paper standards, the United States was on an effective silver standard to 1834 and a gold standard thereafter (section 8 of chapter 3).

10. […]

Unit of account

Until the mid-1790s the monetary system of colonial times continued to exist in its basic form in the United States. The most important feature of that system was the dichotomy between the medium of exchange and the unit of account. The “Spanish dollar” – a silver coin produced in Mexico and Peru as well as Spain, and therefore sometimes called the “Mexican dollar” – was the dominant coin in transactions. Known in Spanish as the peso or piastre, it was termed the “dollar” or “piece of eight” in England and the colonies. The word “dollar” is a corruption of “thaler,” an abbreviation of “Joachimsthaler,” a silver coin produced in 1517 in a Bohemian county of the same name. The smallest subdivision of the peso was the “real,” one-eighth of a dollar. The term “piece of eight” flowed naturally for the entire peso.

In contrast, the unit of account was based on the English system of pounds (£), shillings (s.), and pence (d.), where £1 = 20s. and 1s. = 12d. The phrase “based on” rather than “equivalent to” is used advisedly, because “a shilling from the British mint was not a shilling in any colony” (Carothers, 1930, p. 34). Instead of a national, homogeneous, standard of value, the individual states, as the colonies before them, had their own units of account.