INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand corner and continuing from left to righti n equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

UMI A Bell & Howell Information Company 300 North Zeeb Road. Ann Arbor. Ml 48106-1346 USA 313/761-4700 800/521-0600

Order Number 9512517

Banking before the bank: London's unregulated goldsmith-bankers, 1660-1694

Quinn, Stephen Francis, Ph.D. University of Illinois at Urbana-Champaign, 1994

UMI 300N.ZeebRd. Ann Arbor, MI 48106

BANKING BEFORE THE BANK: LONDON'S UNREGULATED GOLDSMITH-BANKERS, 1660-1694

BY

STEPHEN FRANCIS QUINN

B.A., Rockford College, 1988 M.S., University of Illinois, 1994

THESIS

Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Economics in the Graduate College of the University of Illinois at Urbana-Champaign, 1994

Urbana, Illinois UNIVERSITY OF ILLINOIS AT URBANA-CHAMPAIGN

THE GRADUATE COLLEGE

MAY 1994

WE HEREBY RECOMMEND THAT THE THESIS BY

STEPHEN FRANCIS QUINN

ENTITLED. BANKING BEFORE THE BANK:

LONDON'S UNREGULATED GOLDSMITH-BANKERS, 1660-1694

BE ACCEPTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR

THE DEGREE OF_ nnr.TOR OF PHTT,OSOPHV

Head of Department

Committee on Final Examination-]-

Chairperson \cuJL

t Required for doctor's degree but not for master's

0-517 ® Copyright by Stephen Francis Quinn, 1994 To my parents, whose love is my greatest blessing.

iii Acknowledgements

I must thank my advisor Larry Neal for assistance at every step. Likewise, I wish to thank my two other committee members, Charles Calomiris and Paul Newbold. Little of the original research in the thesis would have been possible without the generous support provided by Alison

Turton and Philip Winterbottom, both archivists at the Royal Bank of Scotland, the Social Sciences

Research Council and the Institute for Humane Studies. Additionally, I am indebted for comments, corrections, quibbles and other aid supplied by Lee Alston, Jeremy Atack, David Mitchell, Leslie

Pressnell, Forrest Capie, Duncan Ross, D.W. Jones, William Kennedy, Dudley Barnes, George

Selgin, Gerald Dwyer, Joel Mokyr, Eugene White, Lynne Kiesling, Will Hausman, Diane Owen,

Case Sprenkel, Bill Bryan, Paul Johnson, Rod Long, Mary Gregson, Kyle Kaufman, Joe Mason,

Tom Nonnenmaker, Rich Sicote, Mike Maciosek, the Northwestern University Economic History

Workshop, the Indiana University Economic History Workshop, the Quantitative Methods in

Economic History Workshop at the London School of Economics and the Centre for Metropolitan

Studies, London. Special thanks go to Shelly Frank. Finally, all remaining are the author's.

iv Table of Contents

Banking Before the Bank:

London's Unregulated Goldsmith-Bankers, 1660-1694

Chapter 0, An Introduction to Goldsmith-Banking 1

Chapter 1, Why Bills of Exchange were Endorsed and Bank Notes were Not;

Transferability of Debt Contracts in the Early Modern Era 27

Chapter 2, Uncleared Balances and Goldsmith-Bankers:

The Endogenicity of Decentralized Inter-banker Debt Clearing 60

Chapter 3, How a New Fashioned Goldsmith Avoided the Ancient Course of the Exchequer:

A Story of Private Bank Notes Paying Public Taxes 111

Chapter 4, Tallies or Reserves? Sir Francis Child's Balance Between

Reserves and Extending Credit to the Crown, 1685-1695 143

Chapter 5, Gold, Silver and the Glorious Revolution:

International Arbitrage and Goldsmith-Bankers 164

Chapter 6, Conclusions 217

Bibliography 223

Appendix 231

Cirriculum Vitae 264

v Chapter 0

An Introduction to Goldsmith-Banking

Out of shops located primarily on Lombard Street and in the West End, a few of London's goldsmiths specialized in offering banking services during the second half of the seventeenth century. In addition to traditional goldsmith activities such as selling silverware and exchanging coins, goldsmith-bankers supplied sophisticated financial intermediation between customers.

Goldsmith-bankers accepted deposits, payable on demand, called "running cash." Goldsmith- bankers also accepted deposits under conditional withdrawal but with greater interest. The same goldsmiths loaned deposits, managed fractional reserves, issued bank notes and offered checkable accounts. Goldsmith-bankers supplied virtually any intermediary service that might define a bank.

This thesis, however, seeks a larger story for London's goldsmith-bankers. The following chapters examine how goldsmith-bankers connected their customers to the various monetary and financial institutions of late seventeenth-century England. The intermediation that an individual banker provided to his depositors and borrowers formed a core of services. Goldsmiths-bankers grew into a network by joining together through a system of inter-banker clearing. The net widened as many goldsmith-bankers branched out into tax collection and funding government debt. Strands of intermediation even reached overseas as goldsmith-bankers orchestrated international credit and bullion movements. Goldsmith-banking wove a web of intermediation that filled the interstices between fellow bankers, specie, the English Treasury, and overseas capital markets.

Connecting these institutional nodes created positive network externalities. The public could deposit bank notes with competing banks, which enhanced the viability of all bank debt as a means

1 of payment. Goldsmith-banking made the use of bank notes and checks common. Moreover, goldsmith-banking made taxes easier to pay and budget deficits easier to fund. The intermediation of goldsmith-bankers contributed to the expanding financial capabilities of the English Treasury and played a critical role in the origins of the Financial Revolution. Goldsmith-banking integrated

London's financial and monetary institutions and laid a framework of which later innovations, like the Bank of England, would take full advantage.

London's goldsmith-bankers connected so many institutional nodes by turning obstacles into opportunities. After the Restoration of Charles II in 1660, the state of the coinage was terrible. In response, goldsmith-bankers supplied paper debt as a medium of exchange. Also, most financial assets were not marketable, so depositors delegated the investment of their funds to goldsmith- bankers in exchange for interest. Such delegated lending permitted goldsmith-bankers to intermediate between suppliers and demanders of capital. As with other assets, government debt was too illiquid for merchants' tastes, so goldsmith-bankers pooled deposits (short run liabilities) to finance investment in Treasury debt (long run liabilities). Goldsmith-bankers also joined separate forms of intermediation to unblock financial bottlenecks. By supplying paper debt and investing in government debt, goldsmith-bankers induced tax collectors to accept bank notes and checks.

Goldsmith-bankers lowered transactions costs, overcame differences in information, reduced risks and accessed international markets as opportunities required.

The overall picture demonstrates that goldsmith-banking integrated a number of critical seventeenth century institutions. Chart 0.1 maps the range of institutions that goldsmith-banking reached. Each connecting line in Chart 0.1 represents an interstice that goldsmith-bankers bridged.

Each intermediation is examined within the correspondingly numbered chapter. For example, this

introductory chapter (0) examines the relationship between two basic services that coexisted within the businesses of individual goldsmith-bankers: the supply of bank debt used as a medium of

2 J3 Z 3 (D c o ^ m =, o i 3 3 o 1 P (D 3 CO r* (D CO a o o a o 0) i CO —i 3" •T§ !8 Ix- W oi "S *< ! (D i__ ^ ^ 2. CO i 3 co o (D (D i 3 8- Si 0) &—« l o 3. ?(D 7T o oi (D % r-* Q g_ CL en 3 CO o 3 o ! < < (D 0) —T (D «-*- (D 5~ Q CD 0 0 0) 3 3 3 t-t- 7T CD CO -i 3 Q. CD. sr (Q S-" 0 -t O a Q m CO 1 jn (D o oi m 3 O 3 3 3 o

Id. 3 3 CO * 5 8-

3 exchange and the supply of delegated lending. The chapters that follow describe how goldsmith-

bankers provided intermediation that connected such a wide net of economic activities by building

off of these two core banking functions.

Our picture of goldsmith-banking is not based on any single banker; rather, the findings of

this thesis are a composite. The extant records of three different goldsmith-bankers have been

tapped, which takes advantage of the strengths of available sources. For example, only Stephen

Evance's bullion book has survived, and only Francis Child offers a tabulation of his specie

reserves. Use of multiple sources also expands our view to cover both the Restoration and Glorious

Revolution eras. Collecting the experiences of individual goldsmith-bankers into one framework

necessarily suspends some details for the sake of generalization. The benefit, however, is a new­

found awareness that goldsmith-banking integrated England's financial and monetary institutions far

more completely than widely supposed. As Professor Henry Roseveare, author of The Treasury and

The Financial Revolution among other important works, has noted,

Too often it has been assumed that the developments which transformed England's financial potency in the eighteenth century took their departure from the expulsion of James II and the arrival of Dutch William, with Dutch banking, Dutch stock-jobbing, Dutch taxes and all the other paraphernalia of "Dutch finance" in his baggage. That is, at best, a half-truth, and after a reign of 300 years as the whole truth it is more than time that it made way for the other half - that seventeenth century England had laboured long to produce a radical transformation of its financial system which would make it superior of any of its foreign models, and that some of the most important stages in that effort took place in the reign of Charles II.(Roseveare 1991, p.3)

Goldsmith-banking's web of intermediation formed an essential stage in the transformation of England from an early seventeenth century financial backwater into a great eighteenth century

European power that could tax and borrow more than her competitors. Goldsmith-banking emerged well before 1688, and goldsmith-bankers played a decisive role in introducing modern banking to the English public. Moreover, the network of goldsmith-bankers established linkages between

4 money, banking and the English Treasury. Later innovations in state finance, like the national debt and the Bank of England, were truly momentous in the development of long-run government debt, but the earlier development of private commercial banking fostered both the common use of transferable short run debt and laid down connections between private banking and government finance. Decades before the explosion of public debt and of Bank of England notes, London had become familiar with bank notes and checks and had grown accustomed to bank deposits funding

Treasury borrowing, through the services of goldsmith-bankers.

Moreover, the Treasury and the Bank of England came to mimic the proven practices of goldsmith-bankers. An implication of this dissertation is that the Bank of England largely displaced private bankers, rather than forging new connections between the Treasury and the public. For example, a bitter attack on the Bank of England printed in January of 1694 noted,

[The Bank] may retard all Returns of Receivers into the Exchequer who may pay in money by other hands into the Bank, and keep the money due to the King out of the Exchequer to the last day, which will make the King pay interest for money for a longer time.1

Chapter 3 shows that goldsmith-bankers and tax collectors had enjoyed such cooperative arrangements decades before the founding of the Bank of England in May of 1694. The pamphlet's accusation against the future Bank of England applied equally to the goldsmith-bankers.

The breadth and sophistication of London's system of goldsmith-bankers clearly demonstrates that the Financial Revolution was stirring in England well before the arrival of

William III in 1688 or the founding of the Bank of England in 1694. Rather than forming in a vacuum, the great innovations in the state's financial institutions that followed the Glorious

1 The publication, entitled Reasons Humbly Offer'd to the Honourable House of Commons by Eminent Merchants and Citizens of the City of London: Showing the Inconveniences that May Arise by the Bank, is anonymous but is held by Hoare & Co. of London and was likely written by the goldsmith-banker Sir Richard Hoare.

5 Revolution of 1688 were planted in a thriving financial environment. The richness of goldsmith- banking as a nexus of intermediation begins, however, with two core banking services: supplying media of exchange and supplying delegated lending. This introductory chapter considers the

interaction between these two activities - how they reinforced and impinged on each other. The

investment of delegated funds in non-marketable assets created asymmetric information between

goldsmith-bankers and their depositors. As a complement to delegated lending, demandable debt

provided a way for depositors to monitor their goldsmith-banker; however, the risk associated with

non-marketable asset portfolios impaired the circulation of that banker's debt. The trade-off

between liquidity and lending encouraged goldsmith-bankers to develop mechanisms to mitigate

these tensions while opening new opportunities. Recent work by Hoffman, Rosenthal and Postel-

Vinay (1994) have found that mid-eighteenth century Parisian notaries also overcame the problems

of intermediation with asymmetric information. Building on influential works by Calomiris and

Kahn (1991), Gorton and Pennacchi (1990), and others, this introduction presents a theoretical

framework for understanding the supply of banking services by goldsmith-bankers that formed the

center of the web of intermediation called goldsmith-banking.

I. Lending and Liquidity

A striking feature of the late seventeenth century was the extensive use of credit.

International payments were financed by foreign bills of exchange.(van der Wee 1977) Remittances

within Britain made extensive use of inland bills of exchange.(Kerridge 1988) London's wholesale

trade and much of the city's retail trade was a network of debt.(Earle 1989) For decades before the

Glorious Revolution in 1688 and the Bank of England that followed in 1694, debt, acting as a

medium of exchange, was the blood in London's economic veins.

6 During the second half of the seventeenth century, the port of London was a booming entrepot between Europe and the rest of the world.(Davis 1954, Zahedieh 1994) At the same time, silver coin was scarce, and often counterfeit or clipped when available. Using debt as a medium of exchange allowed merchants to diminish their reliance on specie. Moreover, London's merchants desired liquid debt.(Jones 1972) A demand existed in late seventeenth century London for highly liquid debt that could mimic money in transactions. With London's demand for liquid debt came a number of suppliers. Wholesalers, great merchants and money-scriveners all specialized in extending credit.(MeIton 1986, Earle 1989) The English Treasury even attempted offering near- money debt called orders.(Roseveare 1962) The most renowned suppliers of debt suitable for various transactions, however, were the goldsmith-bankers. Individual goldsmith-bankers offered a spectrum of debt contracts ranging from promissory notes and checkable demand accounts to international bills of exchange.2 As a network, goldsmith-bankers reduced the incidence of dealing with specie further by coordinating a system of inter-banker debt clearing.(Chapter 2) Use of specie was further averted by arranging the acceptance of goldsmith-banker debt for the payment of taxes.(Chapter 3)

Besides providing highly liquid debt to facilitate transactions, goldsmith-bankers acted as delegated lenders of deposits. Goldsmith-bankers did not act as brokers when investing depositors' money. Rather, a depositor delegated the lending of his money to the goldsmith-banker with no surety of how the banker would in turn invest those funds. Such delegated lending was a break from the early modern tradition of wealthy merchants lending their own money.(Ashton 1960)

Other early modern financial intermediaries, like money-scriveners or notaries, also assumed

2 For example, in May of 1668, the goldsmith-banker Edward Backwell began keeping an account titled "Bearer" for notes he issued payable to bearer and not to any named individuaI.(BackwelI Ledger "Q", EB/6, 1668) Although never amounting to more than a few hundred pounds each year, this goldsmith-banker had introduced bank currency three decades before the founding of the Bank of England.

7 delegated functions when providing brokerage services.(Melton 1986, Hoffman, Rosenthal and

Postel-Vinay 1994) Once a money-scrivener arranged a mortgage, however, the customer took ownership of the debt. In contrast, the goldsmith-banker owned the debt created from investing deposits. Goldsmith-banking divorced liabilities from assets, so goldsmith-bankers became substantial debt holders in their own right.

Through the extended forms of intermediation examined in later chapters, goldsmith-bankers advanced the economic value both of bank debt as a medium of exchange and of deposits as a means of saving. Chapter 2 demonstrates how goldsmith-bankers expanded the circulation of bank debt (notes and checks) through inter-banker clearing. Chapter 3 examines the institutional arrangements necessary to allow goldsmith-bankers' debt to pass in payment of taxes. Both the system of clearing and the paying of taxes increased bank debt's ability to substitute for specie.

Chapter 4 probes the difficulties a goldsmith-banker faced as an investor of deposits in large amounts of government debt. While offering strong returns, government debt also exposed holders to potential default and ruin. Chapter 5 explores how investing in government debt could be tied to bullion arbitrage. While each chapter considers its own issues in detail, a working goldsmith-banker had to balance all these issues simultaneously.

This introduction builds a framework for placing the examples of goldsmith-banking visited in the body of this thesis within a wider range of banking issues. As unregulated, fractional reserve bankers of issue, London's goldsmith-bankers operated with few restrictions. Beyond usury laws, which were often circumvented in early modern Europe, commercial contract law and the state provision of coined money provided the only two notable exogenous institutional constraints on

London's early bankers. Moreover, the practice of goldsmith-banking even helped to shape these constraints. Chapter 1 of this dissertation surveys how legal recognition of transferable debt within

English common law expanded to meet the practices of bankers and merchants. Chapter 5 examines

8 how goldsmith-bankers facilitated the international flow of bullion in and out of England which contributed to eventual monetary instability.

Study of such unconstrained bankers requires an analytic framework that encompasses how delegated lending and the supply of liquid bank debt are both complementary and at odds. The forms of intermediation highlighted in this dissertation stress that London's goldsmith-bankers developed institutions that allowed delegated lending and debt issue to complement each other.

Chapter 2 explains how goldsmith-bankers accepted the debt of rival bankers from the public, and, through the process of clearing such debt between each other, the goldsmith-bankers monitored each other for illiquidity and abusive debt issue. Similarly, Chapter 3 describes how goldsmith-bankers combined the lending of deposits to the Treasury with securing the acceptance of bank debt for settlement of taxes which increased the liquidity of goldsmith-banker notes and checks.

Banking scholars have considered how delegated lending that specializes in non-marketable

(illiquid) loans interacts with the supply of near-money substitutes. For example, Calomiris and

Kahn (1991) note that asymmetric information between depositors and bankers would make demandable debt less liquid, ceteris paribus. However, the same paper demonstrates that demandable bank liabilities assist delegated lending by allowing informed debt holders to monitor a bank on behalf of relatively uninformed debt holders. Similarly, Gorton and Pennachi (1990) explain that demandable debt allows less-informed debt holders to have first lien on a banks assets relative to better informed equity holders. In both stories, illiquid assets are made into liquid deposits by banker intermediation; moreover, demandable debt minimizes the problems arising from the banker knowing more about his portfolio's risk than depositors (asymmetric information).

Focusing on the theoretical challenges overcome by London's goldsmith-bankers draws attention to the achievements that goldsmith-banking accomplished in addition to those topics discussed in the following chapters. Goldsmith-bankers were entrusted with deposits under varying

9 interest rates and terms of withdrawal. They lent those deposits in their own name rather than acting as brokers. They overcame severe informational asymmetries between themselves and depositors because few financial assets in seventeenth century London were marketable. Moreover, goldsmith-bankers drafted and discounted bills of exchange to and from the European continent and supplied notes and checks as media of exchange in business transactions. The chapters below on inter-banker clearing, tax collection, war financing and international bullion flows all build on these everyday successes of goldsmiths as unregulated bankers. The literature on banking surveyed in this introduction demonstrates that the unheralded underpinnings of high finance were substantial advancements in their own right.

II. A Conceptual Development of Unregulated Banking

The potential superiority of unregulated or "free" banking has been a recurrent subject in the banking literature. The attractiveness of unregulated banking is a relative consideration whose

answer requires appraising the merits and demerits of both free and regulated banking systems.

Attention to how bankers fare collectively is essential. Invariably, arguments in favor of central

banks and regulations as well as unfettered banking hinge on the stability of monetary and financial

systems. For example, Chapter 2 of this dissertation examines the goldsmith-bankers' system of

inter-banker clearing. Monetary and financial history, however, is thick with examples of systemic

crises attributable to governmental abuse of fiat money, poorly conceived constraints such as unit

banking laws and the unintended consequences of impositions like mandated deposit insurance.

Instead, the controversy of the free banking debate really turns on whether unregulated monetary

and financial markets would (or did) do worse than the alternatives. As one commentator noted, the

free banking approach's, "ability to sway or convince us must ultimately rest on the efficiency and

stability, or safety, of laissez faire in comparison to the conceivable alternatives." (Palasek 1989, p. 399)

10 The free or laissez-faire school of banking theory provides a useful conceptual development of early bankers that is not based on any one historical experience.(Selgin 1988, White 1989,

Glasner 1989, Dowd 1993) The story of banking can begin with the evolution of commodity money. Money might spontaneously originate from some easily-priced commodity that becomes commonly used as an intermediary good which circumvents the ubiquitous lack of a coincidence of wants between parties.(Glasner 1989, p. 4-5; Selgin 1988, p. 17-18, Dowd 1993, p. 26-27)

Concerns of transport, storage, divisibility, uniform quality and high value by volume favored precious metals over rival commodities such as cattle in becoming widely accepted media of exchange. The private minting of coins to ensure quality and weight in turn followed the use of precious metals as money.

A conceptual development of banking begins with the fact that coins might be the optimal commodity money, but the costs of storage, assay and movement of commodity money remain.

Warehousing and money changing were the earliest economic opportunities for banks to lessen the costs of transacting in specie. The word "bank" derives from the word bancum, meaning the table behind which money changers operated in twelfth and thirteenth century Italy.(DeRoover, p. 200-1)

Early banks could also provide low-cost storage.(Dowd 1993, p.27) Goldsmiths, who traded in the commodity on whichthat money was based on, enjoyed economies of scale in protection and so could store other people's money for a competitive fee. Moreover, goldsmiths' large, personal stock of the monetary commodity, familiarity with the market prices, and specialization in assaying facilitated the exchanging of one form of coin for another.(Glasner 1989, p. 9-10) Every London goldsmith-banker examined to date participated in the exchange of coins, foreign and domestic.(MitcheIl 1994, Chapter 5)

Besides storage, exchange and assay, a goldsmith could facilitate the avoidance of moving specie. In the simplest case, two transacting parties could go to the goldsmith's shop where the

11 purchasing party had a deposit. If the selling party trusted the goldsmith and wanted to continue

storing the sale's proceeds, the ownership of the money could be transferred from the buyer to the

seller without moving the coins.(Selgin 1988, p. 19) Transfer of ownership without the physical

movement of specie out of the warehousing goldsmith's possession opened a tremendous vista of

potential savings in transaction costs when combined with a legal system that recognized relatively

sophisticated forms of debt transfer.(Selgin 1988, pp. 20-21; Glasner 1989, pp. 1-12) For example, the buyer could write an order (a check) telling the goldsmith to transfer the money's ownership to the seller. Similarly, the goldsmith could have given the original depositor a receipt (a note) saying the goldsmith promised to pay whoever the depositor (who becomes the buyer) gave the receipt.

Both developments entirely eliminated any need for the buyer to go to the goldsmith's shop. Such simple examples required a legal system that allowed the transfer of debt (the goldsmith first owed the depositor but then owed the seller/note holder) without all parties being present or agreeing with the transfer of ownership. The goldsmith lost control over, and even knowledge of, whom he owed.

Chapter 1 of this thesis reveals that this legal concept called transferability or negotiability was a complicated institutional development and should not be assumed lightly in theoretical works.

Again, the marriage of deposit-holding bankers with negotiable debt instruments like notes

(promises to pay) and checks (orders to pay) greatly expanded people's ability not to move specie when settling transactions. Such private currency offered additional savings if the paper debt could be used multiple times before being presented for repayment in hard money. The savings gained by creating and then transferring paper debt in lieu of acquiring and moving specie, however, was not unique to bankers and was enjoyed by any reputable merchant. Negotiable debt instruments and commodity money warehousing were separate, complementary strains of financial development that the successful banker combined. To the degree that warehouses of commodity money attracted

12 available specie into their strongrooms by low fees and widely circulating debt, such bankers

became the dominant suppliers of negotiable debt instruments used as media of exchange.

Free banking theory identifies the next important conceptual innovation in the merchant's

efforts to minimize the movement of specie for settling transactions as the clearing of offsetting

debts.(Selgin 1988, pp. 23-25; Dowd 1993, pp. 29-30) A deposit customer might have come across

another banker's note in trade, and out of convenience might have wanted to deposit the value of

the note with his regular goldsmith-banker. If the goldsmith was a less costly collector of another

banker's note than the depositor, then a reasonable fee could be negotiated for the clearing service.

The fee could be reduced even more if the goldsmith expected the deposit to remain in his care for

an extended period of time. Moreover, economies of scale in collection developed as goldsmiths

accepted larger amounts of rival bankers' notes from customers. Even larger savings occurred,

however, when two bankers compared their offsetting debts and only physically transferred the net

clearings. Moreover, competition among bankers would drive down the fee for clearing debt

between major issuers.

Chapter 2 explains how the goldsmith-bankers' system of inter-banker debt clearing arose

endogenously out of competition between bankers and without centralization through a

clearinghouse. In 1680, Sir Dudley North commented on the results of goldsmith-banking by noting

that, "running cash," the seventeenth century term for demand accounts, "by almost all merchants was slid into goldsmiths' hands; and they themselves [the merchants] paid and received only by

bills; as if dealings were in banco [in bank money]."(North 1826, p. 102) Sir Dudley North applied the term in banco to goldsmith-banker debt as a medium of exchange, but the term came to London

from the Continental exchange banks whose credits had long been referred to as in banco and were cleared within a bank rather than between banks.

13 We do not know precisely how goldsmith-bankers developed, but the warehousing of precious metals, the exchanging of coins and issuance of transferable debt were all common to the earliest known goldsmith-bankers. Reducing the movement of specie saved transactions costs and formed a principal value-adding service goldsmith-bankers could provide. This introductory chapter has differed from many other stories, however, by developing the economics of supplying an alternate medium of exchange without introducing fractional reserves. The lending of deposits is not necessary or even sufficient for the provision of services that minimize the transaction costs of commodity money. For a fee, note issuing or deposit taking "banks" with one hundred percent reserves, such as Amsterdam's Exchange Bank, were capable of supplying media of exchange and clearing services.3 The question in the next section becomes if and how fractional reserve banks

would displace fully-backed debt issuers.

III. Fractional Reserves and Delegated Lending

So far, the functional conception of a goldsmith-banker has evolved from a keeper of a

strongroom and a money changer into a supplier and clearer of paper debt, which spared customers

having to move commodity money themselves. Moreover, competition between bankers would keep

the provision of private currency to a minimal fee. Goldsmith-bankers, however, also loaned

money. Whether through pawns or discounting bills of exchange, goldsmiths becoming bankers

3 In contrast to the goldsmith-bankers, the great Continental exchange banks of early modern Amsterdam, Antwerp, Rotterdam, Hamburg, Middleburg, etc. were government chartered banks that reduced transaction costs by clearing debt.(van der Wee 1977) The author is not arguing that the Continental exchange banks were optimal, but they were very successful at facilitating trade. For example, Amsterdam's Exchange Bank, called the Wisselbank, was founded in 1609 and cleared all bills of exchange over a small amount.(van Dillen 1934) The Wisselbank charged a fee when specie was deposited or withdrawn, otherwise debts were transferred by offsetting ledger entries at no charge. The Amsterdam Exchange Bank was not a lending (fractional reserve) institution except for a few extraordinary loans to the city of Amsterdam and the Dutch East India Company.(van Dillen 1934) When Stuart era pamphleteers called for the establishment of a "bank" in London, they meant a state sponsored exchange bank on the Dutch example.

14 could parlay their expertise in assaying metal and following exchange rates into profitable lending opportunities.

Warehousers of commodity money would eventually notice that over the course of the daily in- and out-flow of specie, a goodly pile of money rested idle. With little initial risk, goldsmiths could lend some of their depositors' money. As bill and pawn brokers, goldsmiths had traditionally lent their own capital, so the lending of deposits might be difficult to monitor initially. Both the fungibility of money and the common use of negotiable debt would also encourage the switch to fractional reserves. Competition between goldsmiths would facilitate this evolution by lowering fees until positive interest rates were needed to attract deposits. As an anonymous 1676 pamphlet The

Mystery of the New Fashioned Goldsmiths or Bankers, which was hostile to goldsmith-banking, explained:

And having thus got Money into their [goldsmith-bankers'] hands, they presumed upon some [depositors] to come as fast as others was paid away; and upon that confidence of a running Cash (as they call it) they begun to accommodate men with moneys for Weeks and Months upon extraordinary gratuities, and supply present Money for their Bills of Exchange, discounting sometimes double, perhaps treble interest for the time as they found the Merchant more or less pinched.(Anderson and Cottreil, p. 161)

With fractional reserves, banks also began accepting each other's debt at par without a fee.

As White and Selgin (1989) explain,

Each bank has an additional incentive to accept rival notes: larger interest earnings. If the notes acquired are redeemed sooner than the notes issued, interest-earning assets can be purchased and held in the interim.(White 1989, p. 227)

Inter-banker clearing, as developed fully in Chapter 2, followed naturally from mutual banker acceptance of debt. Unlike a clearinghouse system, London's goldsmith-bankers had a decentralized clearing system that reduced the movement of specie between bankers by moving debt instead. The voluntary retention of rival's uncleared debt meant that clearing occurred less often and transaction costs were saved because more debt was settled by off-setting debt rather than specie. Goldsmith-

15 bankers willingly accumulated uncleared balances because rivals' notes acted as precautionary reserves against potential adverse net clearing positions in the near future.(Chapter 2) Effective inter-banker clearing aided the expansion of par acceptance and the circulating range of bank debt.

By offering a positive interest rate on demand balances, goldsmith-bankers drew more commodity money into their vaults.

Because economies of scale in banking can only be realized when the demand to hold deposits is increasing, a fall in the cost of holding deposits, which induces people to increase their deposit holdings, makes further reduction in banks' costs possible. Here is another example of the self-reinforcing processes that drive the evolution of monetary institutions.(Glasner 1989, p. 13)

Glasner, along with other advocates of laissez-faire banking holds that with the positive returns generated by fractional reserve banking, private bank notes and checkable deposits (often called inside money) would displace commodity money (outside money) in nearly all local circulation.

Much of London's specie did fall into the hands of the goldsmith-bankers as investments as well as running cash (demand accounts). In 1678, Sydney Godolphin advised William of Orange on the question of how to invest a large sum in England that,

There is a goldsmith in London, one Mr Duncomb, who is a man in great creditt, that will take this sum of money at 6 pr cent, and give your Highness his bond, to pay it at any time upon three months warning. This is the best private security that can bee gotten; if your Highness likes better to accept of security upon the Kings revenue, you may have 8 per cent; but you cannot bee at liberty to have your money at soe short warning. . ."(Roseveare 1991, p. 84)

Chapter 4 of this thesis deals explicitly with how goldsmith-bankers like Charles Duncombe acted as delegated lenders that exploited the illiquidity of the King's debt mentioned in Godolphin's advice

(above). Moreover, Chapter 3 explains how goldsmith-bankers used Treasury debt to arrange bank debt to pass in payment of taxes. The illiquidity and institutional peculiarities of government debt assisted goldsmith-bankers in supplying debt that the public valued both as savings and as a substitute means of payment.

16 IV. Note Discrimination and Asymmetric Information

An insight of free banking theory that applies to goldsmith-banking is the insight that bank debt used as a medium of exchange will rarely be priced with an interest rate or discount.(Selgin

1988, White 1989, Dowd 1993) Such debt gains value by mimicking the usefulness of the monetary numeraire, but with fewer transactions costs. Adjusting the price of a £100 note at the moment of transaction through calculating the value of interest due, or a suitable discount if non- interest bearing, adds transactions costs and dissipates debt's value as a substitute for specie.(White

1987) Goldsmith-banker debt was widely used as a substitute for coin money.

A tension exists between bank debt's value as a circulating medium of exchange and bank debt's value as a return earning deposit. This trade-off has a distributive impact between bankers.

Imagine a shopkeeper accepts a goldsmith-banker's note in payment for wares sold. Regardless of whether the retailer holds the note or deposits the debt with a banker, the total volume of existing bank debt does not change. The disposition of a circulating note, however, matters very much to the individual bankers involved. The issuing banker would prefer the retailer hold the note rather than deposit it for interest; however, the banker would also prefer having the deposit to having the retailer take the note to a rival banker.

For the individual banker, a trade-off occurs because the same risk that accompanies returns on assets predisposes debt that circulates as a medium of exchange to be redeemed. A simple example, however, will suffice to show how the two banking activities, delegated lending and note issue, can impinge on each other. Banker A and Banker B both issue £100 of their own bank notes every day to satisfy customers' demands for cash holdings. Assume Banker A is widely perceived as riskier than Banker B. The public will tend to retain Banker B's notes rather than Banker A's notes. Note discrimination might occur as one half of all of Banker A's circulating notes being redeemed each day while only one quarter of all of Banker B's outstanding notes are redeemed

17 daily.4 With an outflow of £100 each day from both bankers, Banker A will soon have £100 worth of notes in continuous circulation.5 Banker B, however, will enjoy £300 in notes remaining in the public's hands.6 Banker B's £300 circulating debt is an unredeemed liability that could be invested in interest bearing assets and represents the returns to supplying bank debt that is valued as a substitute for specie. The opportunity cost to Banker A's higher risk portfolio would be the lost returns from the otherwise loanable reserves, in this case £200 relative to Banker B. Banker A is balancing having fewer pounds to lend against greater return and greater risk per pound lent. Note that a merchant might prefer Banker A for interest-bearing deposits but prefer to hold the notes of the less risky Banker B.

Goldsmith-bankers faced the same potential trade-offs between the circulative life of debt and the perceived riskiness of assets because few financial assets were readily marketable. In the seventeenth century, risk to depositors from a lack of information about bankers' assets was unavoidable. Bills of exchange, East India Company stock and perhaps the shares of a few other trading companies were the few commonly priced and traded financial assets before the Glorious

Revolution. As Chapter 4 will expand on, even government debt was not conveniently acquired or transferred before the founding of the Bank of England. Underdeveloped markets for assets meant that goldsmith-bankers provided especially valuable delegated lending services, but depositors had

little knowledge of the risk of a banker's assets, if depositors even knew in which merchants, companies and revenue streams a banker had invested. A goldsmith-banker's fiduciary

responsibility would have precluded any detailed public presentation of his portfolio, for goldsmith-

4 In this hypothetical example, no assumption is being made about any individual's preferences; rather, the hypothetical redemption rates represent the complicated summation of various note holders' discriminatory behavior.

5 {£100 .lew notes + £100 existing notes} * 0.50 redemption rate = £100

6 {£100 new notes + £300 existing notes } * 0.25 redemption rate = £100

18 bankers regularly invested deposits armed with specialized information produced by customers' financial histories. For example, extending overdrafts on demand accounts was a common form of lending that built on the goldsmith-banker's experiences with a customer. Goodhart summarizes,

The implications of such information asymmetry, however, are that there is still no well-functioning market for bank assets of the kind that banks very largely provide; after all, in my view it is the basic rationale for banking that they provide loans that are substitutes for non-existing markets.(Goodhart 1991, p. 5)

This definition of a bank as a delegated lender of deposits emerges as a separate, potentially related, form of value adding economic activity from our previous supplier of media of exchange, but one that also applies to London's goldsmith-bankers. Unfortunately, Goodhart does not consider a bank that might supply both "substitutes for non-existing markets" and private currency.7

Calomiris and Kahn (1991) and Gorton and Pennacchi (1990), however, show that the very use of debt instead of equity as a contractual form for deposits helps overcome the lemons market created by asymmetric information by sorting investors.8 Equity holders only receive their returns after debts are paid. Deposit by debt allows banks to, ". . . split the cash flows of underlying [bank] assets," so the uninformed small investor has an easily marketable security and the well-informed large investor holds a residual equity claim.(Gorton and Pennacchi 1990, p. 50) With only unlimited liability ownership, the goldsmith-banker was always the equity investor and clearly had

7 Goodhart's The Evolution of Central Banks assumes note issue ends up in the hands of central banks because of, "their privileged legal position, as banker to the government and in note issue."(Goodhart 1988, p. 5) These privileges arise to handle, "a somewhat chaotic system of note issue, to centralize, manage, and protect the metallic reserve of the country, and to facilitate and improve the payments system. While these latter functions were seen as having beneficial economic consequences, the ability to share in the profits of seignorage and the greater centralized control over the metallic (gold) reserves had obvious political attractions as well."(Goodhart 1988, pp. 4-5)

8 Unlike the story that asymmetric information between bankers and depositors causes runs, the Diamond and Dybvig approach considers randomly high withdrawals leading to a run given sequential servicing.(Diamond and Dybvig 1983) Diamond and Dybvig define a bank as essentially a risk sharing intermediary. Numerous authors have shown that introducing equity and the trading of equity to the Diamond and Dybvig style story circumvents the problem of runs and even the existence of banks.(Gorton and Pennacchi 1990, Dowd 1993)

19 better information regarding his own assets than deposit holding investors. In a related approach,

Calomiris and Kahn (1991) argue that demandable debt was the optimal means of monitoring a bank. Given first-come, first-served payment of creditors (the sequential service constraint), note holders with the greatest personal stake would monitor a bank closely and redeem debt when they lost confidence. The disaffection of key monitors signaled the bank's weakness to other debt holders and could precipitate an overall run. To avoid this, banks had to satisfy a few monitoring debt holders while most of the bank's other holders enjoyed the benefits of large informational asymmetries. The goldsmith-bankers participating in inter-banker clearing and retaining the uncleared debt of their rivals as precautionary reserves provided monitoring along the lines of the

Calomiris and Kahn story.

As work by Gorton (1985), Gorton and Mullineaux (1987), Calomiris (1993), Calomiris and

Gorton (1991), and Kaufman (1994) have demonstrated, the potential for "contagion" to move between similar banks, especially if their affairs are related, creates a negative externality. One bank's troubles, justified on solvency grounds or not, might well spread to others. Other banks have an incentive to stop the contagion from spreading. Given that bank panics arise quickly, other banks may have no time to differentiate themselves in the public eye after panics begin. Banks, fearing contagion, could actually desire a fellow bank suffering a run to survive and send a signal of systemic health to the public. This is in contrast to the normal desire to see competitors fail. The potential incentive to see a competitor survive a run combines with the fact that the banks most at risk for contagion effects are likely the best judges of whether an illiquid competitor is solvent.

Contagion tends to spread to similar banks, and banks in the same location and with comparable loan portfolios would be better at estimating the non-marketable assets that make a solvent bank vulnerable to runs. Moreover, if a group of banks do not know which of them might be a panic's

20 first victim before the fact, a credible, public commitment to support each other can allay depositor

fears and reduce the initial occurrence of unfounded runs.

A system in which bankers are able to estimate each other's solvency, provide immediate

support, and offer credible evidence of co-monitoring is greatly facilitated by regular inter-banker

clearing. Recall that even without fractional reserve lending, bankers would act as debt brokers

among each other, capitalizing on the economies of scale associated with redeeming debt. With

fractional reserves, each banker has an incentive to accept competitors' notes at par without a fee.

Over time, the flow of clearing between bankers would give them important information about each

other. As Gorton and Mullineaux (1987) comment about the nineteenth century New York

commercial bank clearing house, "While the clearing house was organized to produce a simple

product, check clearing, it was also capable of producing a by-product - information."(Gorton and

Mullineaux 1987, p. 460) In particular, other bankers would know if any bank faced negative net

clearings consistently, which would indicate that the bank was overissuing its notes. With special

information and an incentive to distinguish "good" from "bad" banks, networks of banks could engender public confidence by monitoring and expelling members.(Gorton and Mullineaux 1987)

As Chapter 2 explains, goldsmith-bankers developed an extensive inter-banker clearing system that allowed bankers to reveal to the public their trust in competitor's solvency. Clearing arrangements also allowed goldsmith-bankers to co-insure in the event of a run or panic. Banks could accept each other's debt and delay demanding payment which would support troubled colleagues and sterilize a potential contagion affect. Holding the uncleared debt of a threatened

banker directly bolstered public confidence by acting as a credible signal that fellow bankers believed in the solvency of the suffering bank. For example, in the summer of 1668, the goldsmith- banker Edward Backwell held over thirty thousand pounds of uncleared debts payable by his colleague Sir Robert Vyner. Vyner was suffering a run because he held one and a half million

21 pounds in government debt created during the Second Dutch War. The English had just lost that war, and Backwell had himself just weathered a run because of the government debt he held.

Backwell might well have feared that Vyner's collapse would spark runs on all holders of government debt. Clearing arrangements between goldsmith-bankers facilitated both the supply of media of exchange and the helped reduce the systemic risk associated with depositors' lack of knowledge of loan portfolios. The institutional form under which bankers cleared each other's debt also could counteract the potential for runs and panics inherent in the asymmetry of information produced when banks specialize in non-marketable lending.

The supply of demandable debt by goldsmith-bankers as a substitute for specie produced a number of factors that complemented delegated lending with asymmetric information. To review, funds pooled by deposits and circulating debt allowed fractional reserve lending. The demandable nature of debt acting as a medium of exchange in lieu of specie encouraged relatively uninformed debt holders to monitor bankers through runs, which limited losses. The non-marketability of most seventeenth-century assets made the short leash of demandable debt an important confidence building check, in combination with the fact that goldsmiths were full liability bankers. The clearing of demandable debt between goldsmith-bankers originally facilitated the wider circulation of private currency as a means of payment, but the same institutions also permitted goldsmith- bankers to co-monitor and co-insure which might mitigate the runs and panics associated with asymmetric information. London's goldsmith-bankers navigated the difficulties of debt issue and

fractional reserves by developing a system to clear debt (Chapter 2), arranging for bank debts'

acceptance for settling taxes (Chapter 3), balancing delegated lending with strong reserves (Chapter

4) and managing the composition of their inventory of precious metal reserves (Chapter 5).

22 V. Conclusion

This introductory chapter has focused on the theoretical literature that addresses the challenges of intermediation overcome by London's individual goldsmith-bankers. These early bankers both provided transaction-cost-saving substitutes for commodity money and specialized in lending that created asymmetric information between depositor and banker. Unregulated goldsmith- bankers developed into financial intermediaries that jointly supplied both monetary substitutes and delegated lending. Moreover, goldsmith-bankers developed additional forms of intermediation, like inter-banker clearing, that capitalized on the complementarities between the two core banking services. The body of this thesis presents detailed examinations of how goldsmith-bankers added intermediary services that linked the major monetary and financial institutions of seventeenth century England to the customers of goldsmith-bankers.

Because primary evidence from the seventeenth century is limited, the experiences of individual goldsmiths are called on to speak to our overall understanding of these early bankers.

The accounts of Edward Backwell, a Restoration banker of the largest order, provide data on how goldsmith-bankers cleared debt between themselves and became enmeshed in the process of tax collection. The ledgers of Sir Francis Child, a smaller but growing banker from the West End during the Glorious Revolution, capture the trade-off goldsmith-bankers had to make between lending and maintaining capital reserves. Stephen Evance, also a Glorious Revolution era goldsmith-banker but situated in Lombard Street, recorded in his bullion accounts his sales on the international markets in which goldsmith-bankers provided a critical channel for bullion flows. The experiences of these bankers are recorded in Chapters 2 through 5, but are preceded by an overview of the contract law regarding the transferable debt in which the bankers specialized.

Transferable debt was a basic financial instrument goldsmith-banking used to reduce the movement of specie. Bank notes, checks and bills of exchange were debt contracts that allowed

23 users to avoid moving specie out of the goldsmith-bankers' shops. Transfer increased the liquidity of debt by freeing debt to circulate instead of returning for payment to the issuing banker. Chapter

1 examines how legal enforcement of transferability varied depending on whether the contract was transferred by endorsement or not. Signatories to a note or bill were liable in the case of a debtor's default, so endorsement allowed transacting parties to overcome differences in information through a credible, low transactions cost commitment. In contrast, transactions in which both parties had similar expectations of risk preferred transfer without an endorsement. Bank notes tended to pass by bearer (without endorsement) since most transacting parties had comparable expectations of the debtor defaulting while international bills of exchange were transferred by endorsement because merchants in foreign countries often did not know the debtor well.

Like transferability, the clearing of debt between goldsmith-bankers reduced the need to actually move specie. Moreover, mutual acceptance of fellow bankers' notes from the public increased the liquidity of notes issued by participating goldsmith-bankers. Chapter 2 describes inter- banker clearing as a decentralized process that arose endogenously between rival bankers. Rather than maintain myriad covering balances, bankers retained the uncleared debt of competitors as precautionary reserves against future clearing. The network was built around a few members of the

Goldsmiths' Company and their apprentices. Control of balances was enhanced through innovative accounting and responsiveness to unduly large negative balances.

Acceptance of goldsmith-banker notes by tax collectors also enhanced the liquidity of banker debt. Throughout the second half of the seventeenth century, goldsmith-bankers were regularly officers of and/or bankers to the major English public revenues. Chapter 3 discovers that by advancing deposits to the Exchequer, goldsmith-bankers could secure Treasury debt (called tallies) that would be repaid by the tax collectors themselves. By means of tallies, bankers and revenue officers could bypass the cash only procedures of the Exchequer. Tax collectors could accept

24 goldsmith-banker notes when settling duties and then swap the notes with the banker for tallies.

This arrangement let the public use private notes to pay taxes, allowed the tax collector to lessen the amount of specie he handled and expanded the circulation of a banker's debt.

Because goldsmith-bankers offered customers debt payable at demand, bankers were vulnerable to potential runs. Chapter 4 confronts the trade off between investment opportunities and maintaining capital reserves. Despite the high rates of return offered by the Treasury during times of war, goldsmith-bankers would not lower their reserves to assets ratios to accommodate unlimited lending. With the outbreak of the Nine Years War in 1689, the goldsmith-banker Francis Child reduced his reserves to assets ratio from sixty to fifty percent. Child's capital levels are an unique insight and suggest goldsmith-bankers retained substantial capital by modern standards against the likelihood of significant runs.

Such large stocks of gold and silver made a goldsmith-banker's capital a valuable inventory to manage. Chapter 5 explores how London's goldsmith-bankers participated in the international bullion markets. By playing the differences in exchange rates between gold, silver and bills of exchange (credit), traders could enjoy arbitrage returns. Bankers and their merchant partners exported undervalued English silver in trade for bills of exchange at favorable cross-channel rates.

Goldsmith-bankers acted as conduits for the flow of bullion into and out of England that led to a monetary crisis in 1695 and 1696. The recoinage and fixing of gold prices that resolved the crisis set England on the road to the gold standard.

Works by Henry Roseveare, Larry Neal and David Mitchell have filled in additional sketches of goldsmith-banking. In his unpublished Ph.D. thesis, Professor Roseveare revealed the importance of liquidity in London's bankers ability to attract deposits at the expense of direct public lending to the Treasury.(Roseveare 1962) Professor Larry Neal, advisor to this thesis, and David

Mitchell of the Centre for Metropolitan Studies, University of London, have more recently expanded

25 our knowledge of previously unstudied goldsmith-bankers. Professor Neal has explored the relationship between the goldsmith-banker George Middleton and the notorious Scot John Law.(Neal

1994) Middleton's activities during the South Sea Bubble in 1720 have documented the extensive international connections that goldsmith-bankers employed. From the 1670s and 1680s, David

Mitchell has flushed out the history of Sir Thomas Fowle, a goldsmith-banker in London's West

End.(MitchelI 1994) Fowle built a growing banking trade from a base as a retailer of the French style of plate then in fashion. Fowle's story suggests that many goldsmith-bankers emerged out of

large retailing and pawn brokering.

Taken as a whole, goldsmith-banking was an innovation that thrived well into the eighteenth century. The introduction of the Bank of England eventually limited private note issue, but checking, bills of exchange, brokerage and other services continued through the sweep of the

Financial Revolution. Goldsmith-banking survived the plague in 1665, the Great Fire in 1666, the

Stop of the Exchequer in 1672, the Glorious Revolution in 1688, the founding of the Bank of

England in 1694 and the South Sea Bubble in 1720. Seventeenth century firms like Child's and

Court's survived as independent banks into this century. As of this writing, Hoare & Company,

founded in 1672, is still privately owned and managed by its eleventh generation of bankers.

26 Chapter 1

Why Bills of Exchange were Endorsed and Bank Notes were Not;

Transferability of Debt Contracts in the Early Modern Era'

Pray do me the favour to pay this bird-man four guineas for apaire ofparekeets that I had of him. Pray dont let any body either my Lord or Lady know that you did it and be sure my selfe to pay you honestly again. Arthur Somerset

Chelsy, the 23 of September [16]962

The above note was addressed to the shop of London goldsmith-banker Sir Francis Child.

Checks, notes and bills of exchange were used extensively throughout the seventeenth century as substitute media of exchange for specie. In Arthur Somerset's case, his note to the goldsmith- banker Sir Francis Child was in lieu of paying the "bird-man" four guineas in exchange for two parakeets. Somerset used a debt contract to avoid the difficulties of arranging payment in specie.

The note is exceptional for the early modern era in its call for secrecy, in stating for what purpose it was written and in not mentioning how the debt could be transferred to others.

Francis Hughes provides a more common example of a seventeenth century order to pay.

1 In this chapter, the term "negotiability" has been avoided because negotiability is a legal concept. Given the incompleteness of our modern understanding of the actual level of legal recognition and enforcement of transferable debt in the seventeenth century, the author does not want to imply legal recognition or attendant legal implications where we lack a firm factual basis.

2 Unless noted otherwise, this and the following notes are from bundles of extant notes from the shop of London goldsmith-banker Sir Francis Child preserved by the Royal Bank of Scotland.

27 29 July 1687 Mr Francis Child desire to pay to Mr John Feche or bearer the sume of five hundred pounds on accompt of Your Servant Francis Hughes

ent Aug 6

Hughes wrote this order for the goldsmith-banker to pay Mr. John Feche five hundred pounds payable out of Francis Hughes' demand account with Child. The note does not specify what Feche gave Hughes but does expressly say Child is to pay "Feche or bearer." The note was entered on

Child's books on August 6, 1687, eight days after the notes creation. How many additional transactions the note facilitated over those eight days cannot be ascertained nor does the note record who finally presented the debt for payment. Clearly, however, Hughes thought it worth while to make the contract explicitly transferable.

Transferability was an important contractual consideration for debt used as media of exchange because transfer allowed the same debt instrument to facilitate multiple transactions.

Transfer created velocity. A transferable debt contract permitted the creditor to change without the debtor's explicit agreement or even knowledge. When Francis Hughes ordered Sir Francis Child to pay John Feche, Hughes passed ownership of Child's debt to Feche without Child's knowledge or explicit consent. The same would have held true if Feche had passed the note on to another party, say Thomas Brown. Sir Francis Child would owe Thomas Brown despite the fact that the goldsmith-banker never agreed to Brown as a creditor. The nature of transferability allowed the ownership of debt to be altered without renegotiation among all the parties to the contract.

Avoidance of renegotiation saved transaction costs, especially if the transacting parties were a distance away from the debtor.

Transferability also reduced clearing costs. Merchants wrote up notes and bills to avoid using specie; however, debt contracts still had to be paid in the future. Clearing of offsetting debt reduced the need to use specie to retire debt. With offsetting debts, only the net amount had to be

28 covered by money. Unfortunately, individual merchants rarely had a coincidence of each other's debt. Exchange banks on the Continent and clearing networks of goldsmith-bankers in London provided merchants with ways to coordinate systems of debt clearing and minimize the flow of specie. Transferability also diminished the need to clear debt by reducing the creation of new debt.

Instead of writing up new notes and bills, which would eventually have to be cleared, an existing debt contract could be used.

Transferability also complemented the use of debt as a medium of exchange. The original writing of a note or bill of exchange created an initial debt contract that permitted a transaction to avoid using specie. Transferability allowed the same contract to be re-used in a second, third or more transactions. A debt holder's ability to transfer debt could increase his valuation of the debt contract. John Feche might be content to present Francis Hughes' note to the goldsmith-banker

Francis Child; however, Thomas Brown might offer a more immediate and valuable outlet for the note. A debt contract often gained value from the liquidity that transfer could offer. For example, transferability increased the value of goldsmith-banker debt. Debt holders appreciated the liquidity that public circulation of bank notes fostered. Likewise, demand accounts were more useful to customers if the banker could be ordered by note to pay a new party out of the depositor's account.

While a tool for merchants and bankers, transferability was an aspect of a debt contract.

Upon breach of contract, debt transfer became a legal consideration. The details of legal recognition of debt transfer determined the forms that transferability took in practice. In seventeenth century

England, two types of transfer were common, transfer by bearer and transfer by order. The two forms had very different legal interpretations. The expectation of how a breach of a transferred debt contract would be enforced at law depended on the form of transferability used. Under a bearer clause, the debt holder could only take action against the original issuer of the debt. Under an order

29 clause, the debt holder could sue the original issuer and any other party later named on the note by endorsement.

After introducing the common contractual forms of debt used as media of exchange, this paper will show that the form of transferability a debt contract used was a function of the relative expectations of the parties involved in the transfer as to how the two common forms of transfer were enforced. If all parties in a debt transfer had the same expectations of the debtor defaulting and the ensuing enforcement, the "X or bearer" clause would be preferable. If the party transferring the debt had a lower expectation of debtor default than the party receiving the debt, then a contract with the "X or order" clause would be superior. Transfer by order enhanced transfer in situations of asymmetric expectations because endorsements enabled more confident transferring parties to commit their own credit as insurance in case of default.

Finally, this paper considers the legal history of transferability on the Continent and in

England over the early modern era. The origins of legal recognition of endorsements began in the

Low Countries and stemmed from problems of asymmetric information regarding debtors. Legal recognition of transferability in England changed with the ascendance of rival types of law and was eventually secured by legislative statute in 1704.

I. Common Forms of Debt Contracts Used as Media of Exchange

Debt contracts in which one party promised to pay another party were called promissory notes.3 The promissory note included several pieces of information: who owed whom, how much, in what currency, the terms of repayment and often where. Depending on usury laws, interest might also have been stipulated. In keeping with their informal, commercial nature, promissory notes

3 They were also referred to as "bills of debt", "transferable bonds", "bills obligatory".(van der Wee 1977, p. 301)

30 tended to be short run debts payable on demand or short notice. The loose nature of a promissory note (an informal, written, promise to pay on or near demand) allowed great flexibility as a medium of exchange. From antiquity through early Modern Europe, promissory notes were widely used by merchants. Below is an example of a promissory notes written by goldsmith-bankers. Such notes were the origin of bank notes and precursor to Bank of England notes which still carry the "promise to pay bearer" clause.

November 28, 1684 I promise to pay unto ye Right Honorable Lord North & Grey or bearer ninety pounds at demand. for Mr Francis Child & my self John Rogers £90:-

Chart 1.1 gives a schematic for the basic promissory agreement established between two parties. In step 1, the drawee gives money for the note. Assume Lord North paid Francis Child for the goldsmith-banker note above. The drawer (Child) received the drawee's value and gave the drawee (Lord North) a note promising payment under whatever conditions stipulated in the contract or assumed under local custom. In the note above, repayment was at demand. Step 2 in Chart 1.1 is at a time after step 1 when the note is presented back to Child for payment. The payee who collects the money could be Lord North, but if the note was transferred, payment would go to the bearer of the note. The payer (Child) collects and honors the contract by paying the bearer (payee) the promised amount.

The other common type of debt contract used in trade were orders to pay. Orders could be as simple as modern checks or as complicated as four party bills of exchange. This chapter's introduction provided two examples of orders to pay. Arthur Somerset's order to Francis Child to pay the bird-man and Francis Hughes order to Child to pay John Feche. Orders assumed that the debtor, in both cases Francis Child, would honor the note's command. A goldsmith-banker like

31 o 3. CO O (D 3" 3 o s 0) "0 S 3 (D 3 3 (D 3 (D +

Drawer Payer Sir Francis Child Sir Francis Child 32 »^- ^^ J 3 W 2 Q_ ro Z O

Drawee 0) 3 w m Payee

Lord North T "Bearer" z. 0 (0 Francis Child would accept orders written by demand account holders and frequently the banker

extended overdrafts at interest.(Chapter 4)

In the seventeenth century, both promises to pay and orders to pay were referred to as

"notes." It must be stressed that the records of London's goldsmith-bankers and other seventeenth century accounts do not generally distinguish between promises and orders. Throughout this thesis

the term "note" will be used in the inclusive seventeenth century sense of orders and promises

unless explicit reference is made to bank notes (promises), bills of exchange (orders) and drawn

notes (orders). After the founding of the Bank of England in 1694, the distinction between a bank note and a drawn note became increasingly common.

Bills of exchange were a seventeenth century exception to the broad title of "notes." Bills of exchange were regularly referred to as bills. The bill of exchange was a type of order to pay that was widely used for international and cross country payments. Bills of exchange probably originated in Italy, and by the fourteenth century the four party bill of exchange was spread across

Europe by Italian bankers.(van der Wee 1977) The bill of exchange falls into the general set of orders to pay with additional important caveats. The bill of exchange usually named three people because the drawer and payer were not the same person. Until the seventeenth century, the bill of exchange always involved two currencies.4 Bills of exchange were likely not to include repayment by demand; rather, bills of exchange were often only payable after a customary period called usance. Usance between London and Amsterdam, for example, was one month, so most bills written between London and Amsterdam would be payable one month after the date they were drawn.

4 In the seventeenth century, inland bills of exchange involving only one currency came into use in England.

33 The transfer of debt to a new party in a second currency created the commercial advantage of the bill of exchange for international payments. Two parties, for example in London, would create the bill with one of the parties, called the drawee, paying pound sterling to the other party, called the drawer, to gain possession of the bill of exchange. The drawer's bill promised the drawee his pound sterling's equivalent in foreign specie from the drawer's contact (the payer) in a foreign city. The bill of exchange was a complex agreement involving two currencies, a few parties, two locations and a period of time. The bill's exceptional quality was the inclusion of a fourth party

(the payee) who could present the foreign bill to the payer (the drawer's foreign agent) for payment.

The payee could be someone other than the drawee and receive the funds that the drawer ordered to be paid.

Chart 1.2 shows the schematic developed by DeRoover and Neal of the four party bill of exchange.(DeRoover 1974, Neal 1990) Step 1 occurred in London when the drawee gave the drawer pound sterling for the bill of exchange. The bill tells the payer, in Amsterdam, to pay the payee a fixed amount in Flemish currency. In step 2, the drawee traded the bill to the payee.

Unless the payee was stated on the bill of exchange, the drawee transferred the bill to the payee by

endorsement. In step 3, the payee presents the bill to the payer for the promised Flemish money.

The drawee gives out the currency to the payee in the drawer's name. In step 4, the payer settles

accounts with the drawer. The honored bill is returned to the drawer, and the payer holds a balance

against the drawer.

In commerce, the drawee would give the bill to the payee in exchange for goods or services.

The payee would collect the bill's value from the drawer's agent (the payer). The financial position

of two parties (drawer and payer) facilitated the commercial transaction of two other parties (drawee

and payee). The advantage accruing to the merchants purchasing and presenting the bills (drawee

and payee) was not having to carry coinage or worry about exchanging local for foreign specie.

34 O C §

f ^

k / CV)

03 3o" m 3 o oo a o 5:*f o (D (D a o 3" o —h 3 i 3 o o 03 3 o™T r-h =r. I I & 03 (0

3- 0) ro 4 o /* > 3" CO 0) k J 3 CQ (D 3 IB 3 I?

35 Bills of exchange worked both to and from a nation. If a London banker owed one hundred

pounds flemish to his Amsterdam counterpart, the Dutchman could draw a bill on his English debtor

in pound sterling and sell the bill for Flemish currency in Amsterdam.

The foreign bill of exchange took advantage of offsetting balances that merchants accumulated with each other in different ports, so that local currency could be used only for local payments, whereas bills drawn against balances held abroad would be used for foreign payments.(Neal 1990, p. 5)

The Dutch improved the means of clearing foreign bills of exchange by establishing an exchange

bank, called the Wisselbank, in 1609 which cleared debts within the bank.(van Dillen 1934) Clerks

within the Wisselbank would credit or debit accounts. The use of bank balances rather than coinage

reduced the cost of debt clearance.

II. The Contractual Forms of Transferability

Transferability of seventeenth century debt instruments was the alteration of who was creditor in a debt contract without the explicit agreement of all the original parties. Debt instruments (promissory notes, drawn orders and bills of exchange) achieved transferability through three primary contractual forms. The most limited type of transferability allowed the drawee to assign the drawer's debt to an agent of the drawee. The drawer promises to pay "X or his attorney" was an example. This was restrictive because the drawee's attorney had to establish his standing as

X's representative and had no better claim to the money than X did. If X died, his attorney could not expect the drawer to remit.

A more common form of transferability involved the drawer promising to pay "X or order".

With endorsements written on the back of the paper, a record of who had held the note was established. The system of note enforcement held everyone who endorsed the note liable for payment of the debt if the drawer failed to honor the agreement. This system at once created

36 increased safety for the party receiving the note and a new liability to the party passing the note.

Endorsement was common on bills of exchange beginning in late sixteenth century Amsterdam.

The oldest extant goldsmith note provides an example of a two party debt agreement with endorsement. Feild Whorwood, a London goldsmith, received sixty pounds from Sam Toffe.

Whorwood acted as drawer and Toffe as drawee of a note. Moreover, Whorwood promised Toffe a two pence per day rate of return (five percent per annum). Without any transfer, Toffe could later act as payee and collect his principal plus interest upon demand. As payer, Whorwood would settle.

The note read as follows:

Reed. 7th Dec. 1654 of Sam Toffe three score pounds wch 1 promise to repay upon demand witnesse my hand with 2d. a day interest Feild Whorwood

(Melton 1978, pp. 30-31)

Sam Toffe, however, did endorse Whorwood's note over to Humphry Molson. Molson presented the note to Whorwood for payment, sixty pounds plus interest. The endorsement read:

/ assigne the Contents of this bill

to Mr. Humphry Molson 7th June 1655

Sam Toffe

(Melton 1978, pp. 30-31)

The relationship between Toffe and Molson is unknown. Molson may have been Toffe's lawyer, business associate or creditor. The lack of any mention of transferability in the original note typifies the informal nature of promissory notes. The parties may have assumed transferability under local custom or by verbal agreement. Whorwood, however, had no known role in Molson's becoming the creditor to Whorwood's debt.

37 The other common form of transferability was the drawer promising to pay "X or bearer".

The presenter of the note had no formal connection to the original drawee, so the value of the note depended entirely on the drawer's honoring the agreement. "X or bearer" clauses were common on goldsmith-banker notes and eventually Bank of England notes. Bearer clauses were also used with orders, for example:

April 26, 1689 Pray pay on sight hereof to ye bearer ye sume of five and twenty pounds on my Acct. {£25 W. Foster

Foster's note was addressed to the goldsmith-banker Francis Child. The banker dated the note June

10 indicating the debt was in circulation for six weeks after Foster wrote it.

Transfer by bearer clause skirted the transaction cost of written endorsement, but left the person presenting the note, payee, with no options if the payer (drawer) refused to pay. Taking the time to endorse provided the payee with potential recourse to the transferor, yet the transferor was exposed to the parallel liability. With endorsement, Molson could turn to Toffe for satisfaction if

Whorwood failed to honor the note. If the reneging Whorwood had promised, "to repay Sam Toffe or bearer upon demand," Molson would be out of luck and Toffe off the hook.

A. Enforcement

The economic difference between types of transference (bearer or order) hinged on how enforcement worked in case of a debtor's breach of contract. Enforcement can be considered in a broad sense that includes reputation and group pressure, but courts of law were the primary means of forcing debtors to pay. The introduction of enforcement of contracts brought teeth to the legal distinctions between transfer by bearer and transfer by order. The potential of the law to force debtors and endorsers to pay changed the public's valuation of notes. Enforcement should be viewed as affecting both parties. The injured party's ability to sue mirrored the debtor's liability to

38 be sued. The probability of the drawee receiving compensation from the drawer, the amount of compensation the drawee could expect and the value of additional penalties the drawer might face from defaulting are all aspects of contract enforcement that alter the valuation of contract enforcement for parties on either side of a case. Enforcement influenced the method of transfer chosen for a debt contract. The difference between "X or order" and "X or bearer" was that transfer by endorsement ("X or order") allowed the holder of a note to take action against endorsers if the original debtor failed to honor the note. The value of endorsements depended on gaining compensation from the drawer and/or endorsers. A model of the debt contracts will demonstrate the importance of enforcement on the form of transferability chosen in contracts. First a model of the debt contract is developed without enforcement. Enforcement is then added to the model to show how the valuation of bearer or order clauses depended on expectations.

Assuming two initial parties, the drawer (payer) and the drawee (payee), the contractual relationship could be characterized by the drawer offering the drawee a return sufficient to match the drawee's opportunity cost.5 One common alternative lenders in London had was depositing specie with a goldsmith-banker. Goldsmiths offered up to six percent interest on short term deposits from 1660 to the 1690s.(Roseveare 1962) Feild Whorwood, from the example above, offered two pence a day on sixty pounds in 1654 or about five percent per annum. This example assumes the opportunity rate of return will be riskless.6

5 This approach assumes that the drawer can minimize the return offered to the payer. While bargaining surely occurred, the original contract has been simplified to expedite analysis.

6 Nothing in the seventeenth century was riskless; however, a bill of exchange drawn by an established merchant on the Wisselbank was close. A defaulter was liable under criminal as well as civil prosecution. "Anyone liable under a bill of exchange was subject to seizure of property and to personal imprisonment."(Assar 1987, p. 109) Also, losing credit at the Dutch bank was tantamount to being expelled from the largest international market in Europe. "All bills negotiated in or drawn on Amsterdam and amounting to 600 guilders (by decree of 11 December 1643 this amount was reduced to 300 guilders) or more had to be paid in the bank." (Assar 1987, p. 105)

39 Let r represent the opportunity rate of return the drawee could receive elsewhere for something of value Q. Recall the Feild Whorwood note:

Reed. 7th Dec. 1654 of Sam Toffe three score pounds wch I promise to repay upon demand witnesse my hand with 2d. a day interest Feild Whorwood

(Melton 1978, pp. 30-31)

Q would be sixty pounds and r would be Toffe's next best rate of return. The creation of a debt contract required the drawer (Whorwood) to offer the drawee (Toffe) enough interest and security to entice the drawee to part with Q (£60).

The drawer's offer had to be greater than or equal to Q(l+r).7 Let i represent the return that the drawer needed to offer the drawee to attract the drawee's money or product. Whorwood offered Toffe two pence a day. The interest rate, i, need not be explicit. For example, implicit interest rates could be garnered from the re-exchange of bills of exchange.

The value of the credit instrument to the drawee depended on the interest rate i and his expectation of default by the drawer. Let Pr(D) be the expected probability that the drawer will not honor the contract at time t+1. The expectation of default is a function of D, a measure of the drawer's reputation known at time t." Equation 1 expresses what the drawer will have to offer the drawee to gain Q.

0(1+r) <; [l-Pr(D)]*QiUi) i1)

7 Assuming maturity of the contract in one time period.

8 At this point, the analysis assumes both drawer and payer have equivalent views on the drawers reputation. Ddrawer=DdniWK.

40 Sam Toffe might expect Whorwood's chance of defaulting to be low, say Pr(D)=01, so Whorwood

offered Toffe [l-.01]*£60(l+.05) or £63.37 per annum.

For the debt to be created, however, a drawer like Whorwood had to satisfy Equation 1 and

still expect to make money himself. The drawer's expected return, at time t+1, from acquiring Q

should be greater than paying [l-Pr(D)]*Q(l+i) at time t+1. We assume goldsmith-bankers had a

rate of return, h>i, that they could earn by loaning out deposits like Q for one time period.9 As an

example, goldsmith-bankers often loaned deposits to the English Treasury for rates of return greater

than the usury rate of six percent.(Chapter 4)

Expectations of contract enforcement modified the relationship between drawer and drawee.

Defining enforcement from the drawee's (Toffe's) perspective, Pr(W) is the probability of winning

recovery should the drawer (Whorwood) default and assumes full recovery. Expected enforcement added benefit to the drawee, so enforcement is added to the right hand side of Equation 1. The enforcement clause would only enter the equation if the expected benefit to the payee exceeded the expected cost of enforcement, C. Except for punitive considerations, a drawee facing breach of contract would not take the drawer to court if the drawee expected court costs to be higher than the amount recovered. Equation 3 shows what the drawer will have to offer the drawee to create the credit instrument.

9 Equation 2 sets out what a drawer like Whorwood would require before entering the debt contract with Toffee. Equation 2 reduces to i < or = h. [l-Pr(D)]*

41

Confidence in enforcement of the contract in case of breach would improve the value of the credit instrument to the drawee. Enforcement would have increased Sam Toffe's valuation of

Whorwood's note by £0.58 from £62.37 to £62.95 assuming Toffe expected a sure, complete recovery at a cost of £5.10

However, enforcement created a liability for the drawer that mirrored the drawee's benefit.

Moreover, the drawer faced additional potential costs such as criminal penalties and lost reputation.

Lost reputation would have been ruinous to a merchant or banker. Within close circles of information, failure to honor a debt threatened destruction of a merchant's credit and ability to trade.

R will represent the discounted cost to the drawer of not honoring a debt above and beyond the court costs and any expected civil award to the drawee. Equation 4 adds the expenses of enforcement to the drawer's decision whether to create a credit instrument.

[l-Pr(D)]*

With enforcement the drawer could offer a lower return, i, since the drawee has some assurance of recovery; however, the drawer's liability is not only the drawee's asset (the expectation of a successful suit) but also additional expenses, R. The circumstances that would encourage note creation from the drawers perspective was a high rate of return investment opportunity, h, or little fear of loss of reputation, R.

As modeled, the marginal benefit of enforcement to the drawee is less than the marginal cost to the drawer.

MBdra=Pr(D)[Pr(W)*Q(Ui)-C] < MC.=Pr(D)[P>

The benefit to Toffe of enforcement was .Ol-[l-£6O(l+.O5)-£5] = £0.58.

42 Equation 4 minus Equation 3 reduces to Pr(D)(2C+R). The difference between the marginals are the expected court costs and the drawer's damaged reputation. If Whorwood defaulted, both he and

Sam Toffe would have to pay for court costs and the goldsmith would suffer a loss of reputation.

As constructed, Whorwood had an incentive to buy Toffe out of Toffe's ability to enforce the contract. However, this model does not consider how waiving enforcement might alter Whorwood's likelihood of default.

Enforcement, however, could become a net benefit by dropping previous assumptions of symmetric information. Whorwood (the drawer) often had a better idea of his probability of default than Toffe (the drawee) did. If the Pr(D)drawer is less than the Pr(D)drawe<. , enforcement becomes less burdensome to the drawer while remaining highly beneficial to the drawee. Such an informational asymmetry could create a reversal such that the marginal benefit of enforcement to the drawee becomes larger than the marginal cost to the drawer. Enforcement produces a net benefit. If enforcement was negotiable, the drawee could induce the drawer to agree to enforcement."

B. Transferability with Enforcement

Enforcement also changed the value of a debt when transferred. The difference between transfer by endorsement and transfer without endorsement (by bearer) was in how the two forms of debt contract were enforced. Returning to the example of a simple promise to pay between Feild

Whorwood and Sam Toffe, what would be the value of the notes to a transferee? How would

Humphry Molson value the note? Assuming Molson (the transferee) had the same redemption and

11 Should the Pr(D)drawer>Pr(D)drawC(. , the drawer would retain the motivation to avoid enforcement. If unavoidable, such drawers would be less likely to create a debt contract.

43 enforcement expectations as Toffe (the transferor), the note would have been equally valued except for the cost of passing the note.12

Imagine Humphry Molson held a shipment of pepper that he would sell for no less than sixty pounds. Moreover, Sam Toffe would be willing to pay no more than sixty-five pounds for

Molson's pepper. To purchase the spice, Toffe needed at least sixty pounds in coinage plus the transaction cost of acquiring and appraising the specie. Alternatively, Toffe could transfer

Whorwood's note to Molson in exchange for the pepper. Both parties valued Whorwood's note at sixty pounds plus interest discounted by their agreed expectations of the probability of Whorwood defaulting and of enforcement. As calculated above, both valued the note at £62.95. Whorwood's debt could have facilitated the transaction by relieving the other two of arranging the specie for payment. If the difference between the price Molson and Toffee agreed on for the pepper and

£62.95 was less than the cost of securing specie for the transaction, transferring the note was preferable to trading with specie. Use of the note meant Toffe did not have to find, transport and assess sixty pounds worth of scarce, dubious coinage. Moreover, Toffe did not have to write a new debt contract with Molson in lieu of the specie. Not creating a new debt meant Toffe would not have to find sixty pounds of specie later. With transfer, only the debt between Molson and

Whorwood remained to be cleared.

With an "X or bearer" clause and identical expectations of default and enforcement, Toffe

(the transferor) and Molson (the transferee) valued the note equally. Either expected to get £62.95 out of Whorwood (the drawer and eventual payer on demand). Under a "X or bearer" transfer, the transferee only has a claim against the drawer, so Molson could only have sought enforcement of payment from Whorwood. If the transfer occurred, the transferor lost all liability. Under transfer

12 Identical expectations would mean the transferee shares the transferrer's estimation of the probability of default and the probability of successful enforcement.

44 by endorsement, however, Toffe (the transferor) was added to the chain of liability. Toffe became

liable if Whorwood failed to honor the contract. Molson (the transferee) valued the note as before plus he enjoyed the extra comfort of being able to enforce payment against Toffe, who endorsed the

note, had Whorwood failed to honor the note. Equation 6 shows the marginal benefit accruing to the transferee from endorsement, viewed at t+1.13 The endorser's expected probability of

defaulting on the transferee is Pr(Dp) and the same expectation is shared by all parties.

Pr(D)[l-PKW0]*(Pr(Dp[Pr(W0C?(l+O-C] + [l-Pr(Dp](?(l+i)) = M2^„ (6)

On the far left of Equation 6, Pr(D)*[l-Pr(W)] is the probability of the original debtor defaulting times the probability that the note holder's suit will be unsuccessful. Pr(Dp)[Pr(W)*Q(l+i)-C] is the transferee's expectation of the transferor defaulting followed by a day in court. Once payment of the debt has fallen to the endorser, [l-Pr(Dp)]*Q(l+i) is the likelihood of the endorser not defaulting times the debt's principal and interest. The expected marginal benefit of an endorsement to the transferee will always be positive so long as some chance exists that the transferor will pay up without a court battle. Unless the cost of court, C, outweighs the expected amount of compensation, a possible court action will also be of positive value to the transferee.

While increasing the value to the transferee, endorsement decreased the value to the transferor since he was vulnerable to a potential liability. Equation 7 shows the marginal cost to the transferor being the same magnitude as the transferee's benefit except the sign on legal costs is switched and the additional hardship of lost reputation denoted by R.

Pr(D)[l-Pr(WO]*{Pr(Dp[Pr(WOO(l+0+C+/q + [1 -PriD^QO- +0> = MCtra^tw (?)

13 For simplicity, analysis assumes an injured party's law suits end with the first court victory.

45 Given the assumptions to this point, the marginal cost of transfer by endorsement outweighed the

marginal benefit by the difference of potential court costs and lost reputations. To repeat, Equations

6 minus 5 equals Pr(D)[l-Pr(W)]*Pr(Dp)(2C+R). Since the net result of endorsement was a cost to

the transferor, he would favor an "X or bearer" clause. The transferor could compensate a

transferee for the benefit that endorsement would have provided the transferee and still come out

ahead. Endorsements appeared to be a burden best avoided.

C. Transferability and Asymmetric Expectations

The verdict on transfer by order changes with the relaxation of assumptions on symmetric

expectations. Endorsement can become a net benefit under asymmetric expectations. If the

transferor has a lower expectation of debtor default than the transferee, endorsement facilitated the

use of debt as a medium of exchange. With Pr(D)^^>Pr(D)^^ , endorsement may prove

beneficial. When the transferor valued the liability of endorsing a note less than the transferee, the

endorsed note increased in value for the transferee more than the note devalued in the eyes of the

transferor. The difference in expectations worked two ways. First, the endorsee initially valued the

note less than the transferor did. Secondly, the endorsee values the transferor's endorsement more

than the absolute value of the liability the endorsement brought to the transferor.14

Returning to Molson and Toffe, if Toffe had a higher expectation than Molson (.99 and .60

respectively) that Whorwood would pay, Toffe valued the note more than Molson did. Assume also

that in all cases, plaintiffs only had a fifty percent chance of winning their case. Under the riskier

court situation, Toffe's valuation of Whorwood's note would fall from £62.95 to £62.64. The small

14 If the transferee has more confidence in the drawer than the transferrer, Pr(D)^^.

46 fall in Toffe's estimation of the note's worth stemmed from Toffe's low expectation of Whorwood's default. Molson's higher expectation of Whorwood's default lowered his valuation of the note to

£48.4.'* Molson's sour expectations cost Toffe £14.24 of potential purchasing power with Molson.

Transfer of Whorwood's note by bearer would not have changed the situation. Toffe's endorsement, however, would have. If Molson trusted Toffe and assigned an ninety-nine percent likelihood that

Toffe would pay if Whorwood did not (Pr(Dp)=.01) then Toffe's signature on the back of the note was worth £12.53 to Molson. To Toffe, the same endorsement only cost him £0.34 including a

£500 penalty for lost reputation in the (unlikely) case he defaulted.16 The endorsement closed a

£14.24 gulf to a £2.05 gap between Molson's and Toffe's valuation of Whorwood's note.

Under a regime of enforcement and asymmetric expectations, Toffe was able to provide insurance to Molson against Whorwood's possible default simply by signing the back of the note. If

Toffe knew Whorwood well while Molson did not know the goldsmith at all, Molson would have liked to know Toffe's true expectation of Whorwood's likelihood of default, but Toffe had an interest in playing up the goldsmith's credit worthiness. Endorsement legally bound Toffe to the contract. Endorsement could also change the transferee's entire expectation of drawer default and raise the transferee's valuation of the note. The value of the endorsement to the transferee might engender the confidence to adopt the transferor's favorable expectation of the drawer. Endorsement provided some guarantee of information because the transferor had committed a stake in the correctness of the information. Endorsement allowed Toffe to make a credible commitment to

Molson in favor of Whorwood's reliability and did so with very low transactions costs. Toffe's endorsement increased the viability of Whorwood's note as a medium of exchange, since people who would not otherwise accept Whorwood's note at a high valuation might reconsider if the

15 Molson's pre-endorsement valuation would be .6(63)+.4[(.5*63)-5].

16 The calculation of Toffe's cost run .01(.5){.01[.5*63+5+500]+.99(63)}.

47 endorser was trusted. Transfer by order allowed debt contracts to expedite transactions otherwise

crippled by distrust or ignorance.

However, transfer by order was not costless. Each signature exposed the endorser to a

potential law suit. If transacting parties had similar expectations of risk, the costs of endorsement

could outweigh the benefits. The appropriateness of debt transfer by order or by bearer depended

on how well transferring parties knew and trusted each other and the original debtor. International

bills of exchange and bank notes are each an excellent example of the economics of transferability.

Foreign bills of exchange involved debtors in other countries, so the extra security provided by

endorsement facilitated their use as a medium of exchange. With bank notes, the debtors (the

goldsmith-banker or later the Bank of England) were very well known. Within London, merchants

would have had similar information about the bankers and expectations regarding default. The more

prominent a banker became, the less benefit endorsement could provide. Bank notes were natural debt contracts better suited to bearer clauses because endorsements were as much or more of a

burden to the endorser than a benefit to the receiver of the endorsed note.

III. Transferability and the Law

In Northern Europe, Antwerp and then Amsterdam led the development of transferability in the sixteenth century. During the first half of the sixteenth century, trade in Antwerp, concentrated

in the Antwerp Exchange, turned increasingly international. "Unknown, unreliable newcomers kept on turning up."(van der Wee 1977, p. 325) In the sixteenth century, the Dutch began supporting negotiability of "bearer clauses" through courts (1507) and Imperial Edict (1537).(van der Wee

1977, p. 325)

48 The legal protection of the bearer contained a deceptive threat: the courts tended to regard all transfers from hand to hand as cessio that relieved the transferring creditor of all responsibility. The greater legal protection of the bearer consequently implied a decrease in his financial security; particularly if he was less acquainted with the debtor. To eliminate this disadvantage, the Antwerp commercial world applied the principle of assignment to the circulation of writings obligatory with a bearer clause.(van der Wee 1977, p. 326, emphasis added)

In Antwerp, the bearer-is-left-holding-the-bag type of enforcement was altered to incorporate the ability to sue a transferor. Why? Too many bearers were unfamiliar with the debtor. The rapidly growing international character of Antwerp contributed to a lack of information about debtors.

Through the rest of the sixteenth century, the Dutch had notes and bills of exchange with bearer clauses that allowed suit against transferors. In the early seventeenth century, the innovation of endorsement occurred." Instead of having to reconstruct transfers from receipts, etc., the transferors were written on the back of the bill. Moreover, the transferee could see who else, besides the transferor, had faith in the drawer. Endorsement was a perfect match for the bill of exchange and the practice spread across northern Europe.

A major difference between the English and Dutch stories of how debt transfer by order came about was that legal enforcement of transferability was outside of English common law over the seventeenth century. Within small merchant circles, legal enforcement was not essential. As commercial circles expanded in the seventeenth century, British traders sought security against default when unfamiliar with debtors. As a first resort, reputation was considered an important enforcer. Merchants were often long term players, so a solid reputation was important. Honoring a debt today meant easier acceptance of one's notes tomorrow. A reputation mechanism turned on expected future benefits of continued commercial contacts. Concern for reputation brought lower probabilities of drawer default.

17 An Antwerp example of a bill in 1611 had a clear assignment on the back.(van der Wee 1977, p. 329)

49 A reputation mechanism also required some means of differentiating between valid complaints and mischievous behavior. The body of rules or customs by which merchants judged agreements was called the "law merchant" or Lex Mercatoria. The law merchant was a body of custom that dictated appropriate commercial behavior, including standard contractual characteristics.

The law merchant was not a structured body of legal theory but a construct of experience and merchant customs. In turn, law merchant arbitration was sensitive to merchant needs. "To the honest merchant, whom the credit of the bourse bound as closely as any power of coercion, the rulings of arbitrators had all the weight of law."(Sutherland, p. 164) Mercantile arbitration did not carry the enforcement of remedies available to courts of law. "The fact that the instruments [notes and bills] could not themselves be enforced was a secondary consideration in a community which depended on good credit and honor."(Baker 1979, p. 303)

In the mid-seventeenth century, enforcement of mercantile instruments was largely outside the traditional English legal system. In 1655, a text written to advise merchants on the usage of bills of exchange did not discuss the legal system. The preface to Marius' Advice concerning Bills of Exchange remarked,

As for Attachments and some few other cases on Bils of Exchange, I have purposely omitted, lest I should give occasion to any turbulent spirit to make contest where none need. The right dealing Merchant doth not care how little he hath to do in the Common Law or things of that nature.18

Arbitration was low cost and familiar with merchant customs; common law was expensive and unfriendly to merchant customs. By 1651, however, English merchants had adopted endorsement into commercial custom. Marius' Advice Concerning Bils of Exchange gave sophisticated explanations of how bills payable "to assigns" and "to order" worked following continental rules.

Inland, three and two party bills were also explained. Malynes' Lex Mercatoria, three decades

18 Marius, Advice Concerning Bils of Exchange, 2nd edition 1655, in preface: "To the Reader", cited in Sutherland 1934, p. 167.

50 earlier, made no mention of these sophisticated advances in the bill of exchange. The rapid adoption of endorsable bills in the middle third of the seventeenth century parallels the expansion of international trade through the island.

In the sixteenth century, England's international commerce was in the hands of foreigners and only a few Englishmen. Legal enforcement was not paramount since transferable instruments in

Britain were mostly confined,

. . .within the limited group of Merchant Adventurers active in the wool and particularly cloth staple markets - men intimately acquainted with one another. Risks were therefore minimal. After 1600 the geographical horizon of commerce widened, the numbers of English merchants so enlarged that increasing unfamiliarity and uncertainty had to be counterbalanced by the introduction of the principle of negotiability in the circulation of bills of exchange.(van der Wee 1977, p. 349)

The events of the seventeenth century expanded the volume and diversity of trade along with the number of merchants. During the later half of the seventeenth century, London became the hub of a rapidly growing network of international trade. The volume of English international trade through

London increased by sixty percent from the 1660s to 1700.19 The birth and growth of re-exports over the period defined London's new role as entrepot between Britain's colonies and

Europe.(Zahediah 1994)

Beyond volume, a plethora of new commodities were passing through London. In addition to a growing trade in traditional woolens, sizeable markets arose in tobacco, sugar, calicoes and manufactures. While domestic consumption accounted for some of the demand for the new products, re-exports supported the diversity of transported commodities. Virginian tobacco and

19 Sixty percent is likely an understatement given overvaluation of key imports for tax purposes in the 1660s.(Davis 1954, pp. 164-5)

51 Barbados sugar flowed to the continent while linens, silks and other finished products left Germany and France for the Americas, all commerce via England.20

A gauge of the rising diversity of products being exchanged in London during the last half of the seventeenth century were the commodity price currents. From at latest 1609 to 1680, London was served by a single price current, and after 1680 a few competing currents. The price currents reflect the information demanded by merchants.

The earliest London commodity price currents printed prices for only about ninety commodities but that number increased to two hundred and fifty by mid- seventeenth] century, to around three hundred twenty-five years later, to four hundred and fifty by the 1690s and to nearly five hundred at the turn of the century.(McCusker and Gravesteijn 1991, p. 298)

With growth and diversification, more merchants traded in London and specialization grew among them. Paper credit became more costly to clear because any two merchants were less likely to have off-setting debts. Transferable debt and a network of inter-banker clearing alleviated the strain.(Chapter 2)

During the last half of the seventeenth century, bills of exchange continued to use endorsement, but another transferable instrument rose to prominence using "to bearer" clauses. The goldsmith-bankers of Restoration London produced a bank note as receipt of money deposited for safe keeping. On this note, the goldsmith promised to pay the depositor or bearer on demand. The goldsmith-banker notes, or running cash notes, became a common instrument for commercial transactions. Dudley North was away from England between 1661 and 1680; upon his return,

20 The Navigation Acts of 1651 and 1660 forbade trade to or from England's colonies except with England and on English ships.

52 He found divers usages in London very different from what had been practiced in his time there ... as first touching their running cash, which, by almost all sorts of merchants, was slid into goldsmiths' hands; and they themselves paid and received only bills; as if their dealings were in banco. He counted this a foolish lazy method, and obnoxious to great accidents; and he never could bring himself to wholly comply with it.2,(North 1826, p. 102-3)

Given England's lack of a clearing bank until 1694, the rise of private intermediaries to clear debts was natural.22 Chapter 2 will consider how London's goldsmith-bankers cleared debt used as a medium of exchange.

Londoners knew the reputation of the major goldsmith-bankers which would have aided the use of "by bearer" transferability. Alderman Backwell was banker to King Charles II, the Queen

Mother, the Duke of York.(Hilton Price 1890B, p. 275) Sir Thomas Vyner was Lord Mayor in

1654 and made a baron in 1660 and was followed by his son Sir Robert Vyner. James Hoare was

Comptroller of the Mint in 1661 and from 1679-1682 he was Warden of the Mint. The Martins, an established goldsmith family on Lombard Street, claimed direct ancestry to Sir Thomas Gresham, the renowned goldsmith of a century before. Sir Charles Duncombe inherited much of BackweU's business after 1672 and became Secretary of the Treasury.23 The goldsmith-bankers were prominent men whose reputations were well known and shaken only by dramatic events that affected all of the members of the mercantile community they served.

Rather than a chain of endorsements leading to transferees farther and farther removed from the debtor, the goldsmith-banker notes of London circulated in a community familiar with the

21 In banco refers to the clearing of debt set-offs in clearing banks such as the Wisselbank in Amsterdam.

22 The Dutch had the Exchange for most of the fifteenth century and the Wisselbank as of 1609.

23 All these goldsmith backgrounds are from F.G. Hilton Price (1890A).

53 drawers. How geographically distant goldsmith notes travelled has been unknown; however, such notes should not have ventured out of a circle of people familiar with the major London bankers.

TV. Legal Recognition

The use of endorsement on bills of exchange and "to bearer" on bank notes occurred in a legal vacuum. The lack of any clear common law recognition of the transferability of notes, or in certain periods, bills of exchange created a potential constraint on the expanding use of transferable credit instruments. Merchants agreed that courts of law were more expensive than resolutions under merchant customs, but the legal system could bring enforcement to those undeterred by loss of reputation. As the diversity of merchants increased, the long arm of the law became more important.

Prior to the English Civil War, the prerogative courts of the Crown were sufficiently flexible to support merchant customs. King's Council, ". . . gave legal force to commercial decisions by granting orders for arbitration by merchants or for ad hoc commissions."(Sutherland 1934, p. 164)

Of course the demand for such legal support came with the de facto laissez-faire trade policies of

the Interregnum.24 Yet, the same constitutional shock that broke the Crown's trading monopolies

also broke the Crown's prerogative courts. The Civil War that expanded the demand for legal

enforcement of transferability also had destroyed the best legal supplier.

Aside from special dispensations from King's Council, another potential court of choice for

merchants was Admiralty. The Lord High Admiral had received jurisdiction over commercial cases

involving foreign merchants or causes of action that had arisen outside of England in the fourteenth

century.(Ogilvie 1958, p. 123) Under the Tudors, Admiralty flourished, and merchants preferred

24 "The immediate effect of the civil wars was a temporary collapse of the system of monopolistic companies and the throwing open of the foreign trade of the country to all comers."(Ashley 1935, p. 159)

54 Admiralty, a civil law court, to common law because foreigners understood and were familiar with civil law. England's principal trading partners operated under civil law, especially the Dutch.25

Civil law had been adopting elements of the law merchant and so was closer to agreeing with merchant custom than English common law.

While Admiralty recorded transferable instruments, a transferee's ability to sue remained unclear. The Medieval transferability clauses of "X or brynger hereof and "X or assigns" were found, but the Dutch innovation of endorsement was not found.(Holden 1955, p. 26) Admiralty also recognized bills of exchange. Around 1600, England's commercial circles were small compared to

Holland's, and Dutch financiers met most British financing needs. Not surprisingly, complex cases would be tried abroad, and development of merchant customs, "scarcely touched the moribund merchant courts of England. "(Sutherland 1934, p. 157)

If Admiralty Court was not sophisticated by Continental standards, it was doing a brisk business by English standards. Common law coveted Admiralty's jurisdiction over commercial cases. In 1602, common law took a first attempt towards capturing some of Admiralty's business by recognizing bills of exchange.26 The transfer of debt ownership between drawee and payee common in the bill of exchange proved a legal hurdle. A basic point of common law was that, "a third person can not become entitled by the contract itself to demand the performance of any duty under the contract."(Holden 1955, p. 17) Common law accepted bills of exchange under a different approach in 1612. The bill of exchange was justified via merchant custom, the Law Merchant.27

25 See Appendix A.

26 Martin v. Burke, 1602, a drawer successfully sued a negligent acceptor via financial assumpsit.(Holden 1955, p. 31)

27 Oates v. Taylor, 1612, merchant custom was stated at considerable length and found sufficient to support the contract. As legal theory, the argument that custom prevailing between merchants could originate a legal duty focused on the meaning of custom. In common law, custom had to be very old, "time wherof the memory of man runneth not to the contrary" and a specialty because general custom

55 Common law began accepting the establishment of merchant custom as sufficient grounds to accept

bills of exchange but only for merchants.28 Common law made the effort to compete for

commercial business, yet common law was not very competitive since a large cost differential still

existed between it and civil law.

While common law competed by trying to offer what merchants needed, Chief Justice Coke

set out to limit Admiralty's jurisdiction. In 1606, Coke, Chief Justice of King's Bench (common

law's highest court), declared the law merchant a part of common law implying that all commerce

fell into common law jurisdiction. Although common law judges, "recognized that the Law

Merchant was part of the Law of the Land, they did not profess to take judicial notice of merchant

customs."(Holden 1955, p. 35) Common law sought to eliminate a competing institution through

political competition rather than by providing the transferability merchants desired. The dispute

between common law and Admiralty met some resolution in a statute under Charles I, 1632, that

recognized a few maritime questions belonged exclusively to Admiralty. Up to the civil wars, "the

administration of commercial law suffered inevitably as the result of the conflict of jurisdictions."(Ogilvie, p. 124)

While courts fought for jurisdiction, transferability remained without explicit common law

support before the English civil wars. In his 1622 guide, Lex Mercatoria, Malynes admitted that transferable instruments were not possible under common law. After the civil wars, the reinstated

prerogative courts of the Crown conformed to the legal precedents of common law, and Admiralty

Court was a shadow of its former self. Interregnum ordinances in 1648, 1649 and 1653 protected

Admiralty jurisdictions set out in 1632, but all were invalidated in 1660. Later attempts to secure

was common law. Bills of exchange were old and only merchants who were using bills of exchange specialized in international commerce.(Holdsworth 1926, p. 160, and Holden 1955, p. 32)

28 In Eaglechild's Case, 1661, a bill of exchange between two non-merchants was not sustainable by reference to the Law Merchant.(HoIden 1955, p. 33)

56 jurisdiction for Admiralty in 1661 and 1669 also failed despite the support of Admiralty by London merchants.(HoIden 1955, p. 65)

While winning the political competition, common law still did not recognize transferability.

In 1681, Chancery had tried to support "X or bearer" clauses in bills of exchange, but common law dissented.29 Not until 1692, in Hodges v. Steward, did common law finally support endorsement of bills of exchange with an "X or order" clause by extension of the merchant customs argument.(HoIdsworth 1926, p. 163) The invocation of merchant custom had become a brief affair, and

under cover of these convenient phrases about the custom of the merchants, it was easy to introduce into common law both the legal principles familiar to the continental lawyers, and the commercial practices familiar both to English and to foreign merchants.(Holdsworth 1926, p. 161)

In common law, the custom underlying the negotiability of bills of exchange began losing any relationship to location or class of user. Bills of exchange and transferability could be used by anyone. Justification by the special merchant custom was a brief, legal facade.

Since the legal support for transferable bills of exchange was merchant custom and merchant custom saw no fundamental difference between bills and notes, merchants expected common law to support transferability of promissory notes. In 1704, Chief Justice Holt noted that merchants had been using notes for decades and accepted note transferability as equally legitimate as bill transferability.30 Four years earlier, Chief Justice Holt drew a legal "bright line" stopping the merchant custom argument from spreading negotiability to notes payable to X or bearer. In Clerke v. Martin, Holt said,

Sheldon v. Hentley, 1681.(Holdsworth 1926, p. 165)

Buller v. Crips, 1704.(Holdsworth 1926, p. 172)

57 . . .that this note [payable to bearer] could not be a bill of exchange, that the maintaining of these actions upon such notes were innovations upon the rules of the common law; and that it amounted to the setting up of a new sort of specialty, unknown to common law, and invented in Lombard Street, which attempted in these matters of bills of exchange to give laws to Westminster Hall.3'

What infuriated Holt was that bills of exchange provided negotiability and could be tailored to three or even two parties, so duplicating everything a note could do. Moreover, merchant custom being ancient, as understood in common law, could not change to incorporate transferable promissory notes.

Holt's stand to defend the purity of legal theory revealed that common law would only go so far to accommodate merchant "opinionativeness." The attempts to gain negotiability for notes represented merchant desires for legal enforcement of the transferee's rights. While the Chief

Justice felt a bill of exchange could do everything a promissory note could, merchants disagreed.

Holt's decision harmed private credit and the running cash notes of the Bank of England.

Merchants pushed for a statute to correct the judge. In February of 1704 Parliament passed a statute establishing the transferability of notes in either form: X or order, or X or bearer.32 Chief Justice

Holt's fear came to pass; Lombard Street would give laws to Westminster Hall.33

Common law recognition of full transferability of informal debt contracts was slow in arriving. While England struggled to recognize transferability, the Dutch had clear, efficient negotiability. Early in the seventeenth century, competition from the Court of Admiralty did spur common law to recognize bills of exchange, but political competition stripped the power of

31 Clerke v. Martin, 1700, quoted in Holdsworth (1926), p. 172-3.

32 The statute treated notes as if they were inland bills of exchange. Up to the statute, inland bills were not negotiable under X or bearer, but since notes and bills were equated and X or bearer notes were explicitly allowed, inland bills to bearer also became negotiable.(Holdsworth 1926, p. 173)

33 Chief Justice Holt, King's Bench, in Clerke v. Martin, 1702, quoted in Holdsworth (1926), p. 175.

58 Admiralty. Without rival domestic institutions, common law resisted adopting the negotiability of credit instruments, for common law had a philosophical bias against transferability.

V. Conclusion

In 1704, the demand for legal enforcement of transferred credit instruments circumvented the courts and relied on statute. Full negotiability had arrived fifty years after Sam Toffe endorsed

Feild Whorwood's promissory note to Humphry Molson, one hundred years after the Dutch adopted endorsement. By statute, Parliament forced common law to face the facts. The boom in English trade during the 1660s, 1670s and 1680s had coincided with the beginnings of a financial revolution in England. Transferability expanded mercantile credit while limiting the amount of debt needing to be cleared. Transferability brought monetary style velocity to debt instruments. By adding endorsements, differences in information were overcome. Endorsement let the potential benefits of transferable instruments reach exchanges far removed from the debtor.

The transference of debt as a medium of exchange took two primary forms. Transfer by bearer eliminated the liability of the transferor and expedited the passing of notes whose debtor was well known and trusted. Transfer by order allowed endorsers to commit their own potential liability to their expectation that the debtor was worthy. In transactions where the new creditor was less familiar with or less trusting of the debtor than the signatory. Both forms, however, had to be settled eventually. The transferable debt issued by goldsmith-bankers increase in value by also being easily redeemed. London's goldsmith-bankers facilitated the spread of transferable debt by developing a system of mutual debt acceptance. Chapter 2 examines how par acceptance by rival bankers of each other's fiduciary promises and orders arose endogenously.

59 Chapter 2

Uncleared Balances and Goldsmith-Bankers:

The Endogenicity of Decentralized Inter-Banker Debt Clearing

In the seventeenth century, notes, orders and bills (collectively called demandable debt) acted as a medium of exchange that spared the costs of moving, protecting and assaying coin or bullion. As previously presented in Chapter 1, the transaction costs associated with repeatedly creating paper debts were lessened by the transfer of debt. Likewise, the clearing of off-setting debts reduced the costs of redeeming demandable debt for specie. By two bankers comparing their holdings of each others debts, only unmatched claims needed to be paid in specie. Clearing limited the movement of specie when settling debts, and London's goldsmith-bankers endogenously developed a system of clearing that captured these savings of transactions costs.

London's goldsmith-bankers maintained a system of clearing without any centralization through clearinghouses. Rather, the goldsmith-bankers' clearing system produced many of the same savings generated by a clearinghouse without pooling members' bullion in one location.

Clearinghouses avoided moving specie because members kept covering balances in one location and changed ownership by ledger entry. Bullion only left clearinghouses when a participating banker had insufficient or excess balances. A decentralized clearing system, however, required each banker to settle his net position separately with each other banker. London's goldsmith-bankers managed the costs of clearing on a direct, individual basis by regularly not demanding immediate repayment.

Waiting to clear allowed off-setting debts to accumulate and spread the transaction costs of settling between two bankers over more mutual acceptances. Moreover, by settling net clearing with debt,

60 goldsmith-bankers could exchange information without the cost of actually moving specie. This chapter argues that the goldsmith-banker's system, which delayed clearing, arose endogenously.

Each goldsmith-banker had an incentive to clear debt with their competitors, so inter-banker activities developed without external regulations.(Selgin 1988, Dowd 1993, Glasner 1989) This chapter contributes the view that once created, uncleared balances acted as insurance against potential adverse clearings in the future. Self-interest was sufficient reason for member banks to both accept and retain each other's debt. Moreover, the goldsmith-bankers reduced their exposure to default and abuse that followed from holding uncleared balances by filtering members through an apprenticeship process.

The goldsmith-banker system of clearing was a fundamental break from the seventeenth century system of exchange banking well established in the Low Countries, Italy and elsewhere.

London's system of debt clearing was not centered in a clearinghouse or central bank. The goldsmith banker himself was owed for each overdraft until the note was cleared. In contrast, the

Amsterdam Exchange Bank (Wisselbank) only accepted a note if the offsetting credit was already covered by a deposit. London's goldsmith-bankers and merchants knew how exchange banks worked, and many pamphleteers called for the establishment of a government sponsored bank. Yet, the goldsmith-bankers chose not to create a private clearinghouse. This chapter submits that decentralized clearing worked quite well to minimize transactions costs and monitor members.

Works by Gorton (1987) and Goodhart (1988), however, suggest bankers might freely place themselves under regulatory control within a clearinghouse framework. These authors describe regulation and central banks as potentially evolutionary outcomes that assuage the public by bankers monitoring and disciplining each other. Dowd (1994), however, "disputes claims that regulation and central banking were a natural, spontaneous response to inherent market failures, and in doing so to suggest they are not economically justified as improvement over a free market."(Dowd 1994, p. 290,

61 emphasis in the original) London's goldsmith-bankers support Dowd's contention. The goldsmith- bankers did not organize a private clearinghouse or formal self-regulation. However, establishing a clearinghouse would have been difficult, for securing an acceptable legal structure for a private exchange bank in London was problematic. The private ownership structures available in the seventeenth century inhibited the concentration of capital. Normally, firms were limited to ownership by individuals and partnerships. With no limited liability, an extended partnership in which all members had to put specie in up front would have entailed tremendous risk. To trust any

individual with the money could have been riskier still. The creation of a joint stock clearinghouse was also possible but would require royal support and parliamentary action. Such political charters came only with hefty price tags. The Bank of England, founded in 1694, had to loan the Treasury

£1,200,000. Moreover, Parliament viewed the goldsmith-bankers with distrust. The bankers were

regularly accused of usury, but Parliament was also concerned that the great bankers provided the

king with a source of independent financing. Finally, one cannot separate the lack of state

sponsored centralization from the fear of expropriation hanging over Restoration London. England's

kings had forced loans and confiscated property during the first half of the seventeenth

century.(Ashton 1960) Moreover, Charles II defaulted (the "Stop of the Exchequer") on most

government debt in 1672. Until the Glorious Revolution in 1688 and the placing of the king within

the Common Law, the crown could not be trusted to respect property rights.(North and Weingast

1989)

This chapter examines the decentralized debt clearing system of London's goldsmith-bankers

in quantitative terms through the accounts of Edward Backwell. Backwell's ledgers run from March

of 1663 until March of 1672 but miss twelve months of records from September through December

of 1663 and March through December of 1665. While one banker's clearing accounts with his

fellow bankers cannot portray the entire system, Edward Backwell's accounts do detail his inter-

62 banker balances. Earlier writers have interpreted Backwell's extensive relations with other goldsmith-bankers as evidence that the goldsmith-banker was a "banker's banker." R.D. Richards concluded that Edward Backwell,

Was undoubtedly both the central or reserve bank and the clearing house of the post- Restoration period. [Backwell's bank] was the indispensable precursor of the Bank of England, a precursor which was of paramount importance in this outstanding era of English economic expansion.(Richards 1929, p. 30)

Although the network of goldsmith-bankers was an "indispensable precursor of the Bank of

England," Edward Backwell was not a clearing house. At least nineteen goldsmith-bankers held accounts with Backwell over the decade form 1663 to 1672, but they did not clear with each other through Backwell.1 Rather, each banker cleared his own and Backwell's notes with Backwell.

Occasionally Backwell's clearing accounts would involve a third banker. To argue these infrequent entries constituted a clearing house would suggest that all other goldsmith-bankers had no great volume of clearing business except with Backwell. The author knows of no reason to assume that Backwell was special in this regard. Moreover, no argument has been presented by earlier writers as to why other bankers would accept Backwell's notes but not those of other prominent bankers. Such a system would constrict the circle of note acceptance to the detriment of everyone except Backwell. Instead, evidence points towards a system of dispersed clearing. For example, Sir

Francis Child was a small but growing West End banker rather than the Lombard Street giant that

Backwell was, yet the ledgers of Sir Francis Child from the 1670s record regular clearing activities with a number of goldsmith-bankers.(Child's Ledgers, Mitchell 1994)

This chapter puts forward a new interpretation of Edward Backwell as an important member in a web of clearing arrangements between bankers. Each goldsmith-banker had an incentive to join

1 Backwell also had regular clearing transactions with the money-scrivener firm of Morris and Clayton. The money-scriveners were substitute financial intermediaries specializing in mortgages and formal deeds. (Melton 1986) Their accounts with Backwell demonstrate that the scriveners and goldsmiths both benefited from participating in a system of mutual debt acceptance.

63 the network in order to satisfy liquidity conscious customers and expand his own note circulation.

The endogenous growth of the informal acceptance and retention of debts meant each banker had to

balance the value of uncleared debt as precautionary reserves against the risk of default. Although the holding of uncleared reserves created some room for abusive over-issue of notes, the decentralized system's monitoring members was incentive compatible because each member had the

incentive to redeem uncleared debt beyond his precautionary demand. Moreover, through apprenticeship, proximity and social ties, the network established reputations and engendered trust.

The system of goldsmith-banker debt clearing spanned the second half of the seventeenth century. While this paper only considers quantitative evidence from 1663 to 1672, the story of the system runs much longer. The system's roots began somewhere in the years of the Protectorate

(1648-1660).2 By 1663 Edward Backwell was running a large banking business and was clearing demandable debt with a large number of fellow bankers. With the Stop of the Exchequer in 1672, a half dozen goldsmith-bankers who lent large sums to Charles II were forced to suspend payments

(Richards 1930, Horsefield 1982). Edward Backwell was greatly harmed by the Stop of the

Exchequer and suspended inter-banker clearing activities in 1672.

Many goldsmith-bankers, however, were not major lenders to the Crown and did not suffer ruin in 1672. Five years after the Stop, forty-four goldsmith-bankers were listed as keeping running cashes in the Little London Directory (1677). Charles Duncombe, Backwell's apprentice turned heir to the business, in partnership with revenue agent Richard Kent, was thriving. In the 1670s and

1680s, the ledgers of Sir Francis Child show that he was clearing debt with prominent West End goldsmith-bankers like Thomas Fowles and Lombard Street bankers such as Charles Duncombe and

Stephen Evance.(Mitchell 1994) Despite the plague in 1665, the Great Fire in 1666, the Dutch fleet

2 For example, the first extant Backwell ledger is labeled letter "I" and begins in January of 1663. If each of the earlier ledgers covered a year, then Backwell's first bank ledger began in 1655.

64 marauding the Thames in 1667, the Stop of the Exchequer in 1672 and a severe panic in 1682, the system of goldsmith-bankers continued functioning through the foundation of the Bank of England in 1694 and the turn of the eighteenth century.

I. Backwell's Inter-Banker Balances

Edward Backwell was an early modern banker of the largest order.(Hilton-Price 1890,

Richards 1930, Clark 1938) Having served as an apprentice to the important Cromwellian goldsmith-banker, Sir Thomas Vyner, Alderman Backwell became a banker to members of the restored royal family, the customs revenue, numerous crown agents including Samuel Pepys, the

East India Company and hundreds of merchants, gentry, professionals and others.(Clark 1939,

Roseveare 1963) From his shop, the Unicorn on Lombard Street, Backwell ran a diverse business.

Interest was paid on deposits; loans were supplied; bills of exchange, tallies and various types of Treasury-Exchequer payment orders were discounted; promissory notes, which circulated freely, were issued; cheques were used; bullion was bought and sold; foreign coins were changed; systematic accounts were kept in special ledgers. (Richards 1928)

The scope and scale of Backwell's bank make him a central player in the goldsmith-banking system of the 1660s, the first decade with surviving quantitative evidence of inter-banker clearing activity.

Backwell's transactions with his fellow goldsmith-bankers must be interpreted from the perspective of Backwell's prominence and breadth of activities.

Chart 2.1 presents the aggregate daily balance of nineteen goldsmith-bankers with Edward

Backwell.3 Each goldsmith-banker's account with Backwell was encoded and calculated separately

3 Sundays have been excluded from Chart 2.1, and the rare Sunday transaction has been moved to the prior Saturday. The sample includes the earliest generations of goldsmith-bankers: John Colvill, John Lindsay, John Mawson, Francis Meynell, John Portman, Jeremiah Snow, George Snell, Thomas Vyner (Chart 2.2) and Robert Welstead. Thomas Cook, John Hinde, Benjamin and Edmund Hinton, Joseph Horneby, Thomas Kirwood, Thomas Pardo, Thomas Row, Bernard Turner and Thomas Williams all began accounts with Backwell after March 1663. These goldsmith-bankers individual balances are charted in Appendix B. Finally, only those that have been identified by other sources (Heal 1935,

65 Pound Sterling o (Thousands) 00 c 5L Q> 3 O CD CO

rjo 0) 3 :*• 8 O 3"

00 0) ro o $ 5? O (D

3 CD

Hilton-Price 1890A) as goldsmith-bankers have been listed. As yet unknown bankers are not included, so the sample is biased towards exclusion.

66 and then summed. At various times Backwell held both positive and negative balances relative to his fellow goldsmith-bankers with a daily mean of-£1,445 over the 2,511 day sample.4 The series had a standard deviation of £4,700. As Table 2.1 reports, the median total of banker balances held by Edward Backwell was -£813 and was considerably higher than the mean. At six percent annually, Backwell's average daily balance of-£1,445 cost the goldsmith-banker nearly £700 in forgone interest over the eight years sampled.5 However, the few hundred dollars in opportunity costs accounted for less than one percent of Backwell's £77,279 net lending income (interest received less interest paid) over the same period.

Table 2.1

Distribution of Daily Balances

for All Bankers on Edward Backwell from January 1663 through March of 16726

By Number of By Value of Observations Observations (Standard Normal Distribution) Minimum =-£17,487 10% -£7,731 -£12,349 33% -£2,038 -£3,340 50% Median = -£813 Mean =-£1,445 66% -£16 £449 90% £3,450 £9,459 Maximum = £17,476

4 Although graphed, balances from January 1, 1663/4 to March 24, 1664 and from March 25, 1665 to December 31, 1665 are not included in the summary statistics.

5 Six percent was the maximum legal interest rate under the usury laws at this time.

6 Distribution excludes missing observations for January through March 1663/4 and March through December 1665.

67 Backwell's accounts with his fellow goldsmith-bankers mingled clearing with other types of transactions. The accounts include loans and bullion sales between Backwell and his colleagues.

These effects are separated out for a sample of three bankers in Section III (below), but the values in Chart 2.1 and Table 2.1 remain largely a mixture of direct loans, uncleared acceptances and other business transactions. However, a few extreme inter-banker relationships have been removed from

Chart 2.1.7 Chart 2.2 presents Backwell's complete, unaltered balances. The removal of the few largest transactions allows a much clearer examination of the everyday workings of Backwell's clearing activities.

Chart 2.1 and Table 2.1 exclude the account of Sir Thomas and Sir Robert Vyner from

Backwell's total balances with all bankers. Most of Backwell's very largest daily balances were a result of his unique relationship with the shop of his former master Sir Thomas Vyner and Sir

Thomas' nephew and heir, Sir Robert Vyner. Sir Robert Vyner assumed full proprietorship of the

Vine after 1665. Both Vyners were men of great importance - Sir Thomas received Oliver

Cromwell's first grant of knighthood and Sir Robert was a personal friend and financier to Charles

II.(Clark 1941) Chart 2.3 plots the daily balances of the Vyners with Backwell and reveals that the

Vyners occasionally ran balances of over twenty and even thirty thousand pounds. In the extreme, the summer of 1668 saw Vyner's debt reaching £35,000 by late June. In 1667, after the disastrous

Second Dutch War, depositors ran on Vyner. At that time, Vyner was owed at least £1,500,000 by the Crown.(CIark 1941, p. 35) In 1668, depositors remained uneasy and the Crown pressed for

7 Adjusting Chart 2.1 (as described below) lowered the daily mean from an unadjusted -£1,337 to - £1,445. The variance of balances, however, fell substantially from Backwell's unadjusted position of £11,008 to £4,700. Besides Vyner, Chart 2.1 also excludes extraordinary spikes in the balances of Sir Francis Meynell and Thomas Row. On April 18, 1663, Meynell was credited by Backwell for £34,666:13:4 on behalf of money received by the sale of Dunkirk. In August of 1670, Row deposited over £30,000 in foreign coin with Backwell. In January and February of 1671, Row was repaid by the East India Company for whom Backwell was their banker. Both outliers have been removed from Chart 2.1 to highlight the remaining transactions.

68 Pound Sterling CO o c (Thousands)

00 SL 0) 3 O CD C/J O O (D 3" 3 3 CD

00 O

CD

69 Pound Sterling CO O c (Thousands) (D

00

0) 3 O CD CO O > o UJ 0) 3 3- CD ro CO • s GO «-+ 00 0) o *r CD

70 more funds with threats of withholding interest payments.(Roseveare 1962) During that difficult summer, Backwell appeared to have provided Vyner with substantial credit.8 Excluding the Vyners reveals the Backwell's more common inter-banker relationships.

II. Endogenous, Decentralized Clearing

This section considers Backwell's daily inter-banker clearing behavior (as described in aggregate above) as having been an ongoing process of controlling precautionary debt reserves. In contrast to stories that involved the management of positive balances by firms or the public

(Baumol, Miller and Orr), the reciprocative acceptance and retention of debt by goldsmith-bankers required banks to manage the uncleared debts of other banks. The inventory problem was balancing one's holdings of another banker's debt against the expectations of a rival banker accumulating off­ setting debts. Moreover, this analytical approach develops a framework for examining decentralized clearing systems that is endogenous. Both mutual debt acceptance and debt retention occur without regulation or system wide planning because bankers' individual incentives are sufficient to promote clearing. Like a market, the decentralized clearing network spontaneously orders itself out of bankers' acting in their own interest.

Other authors have shown that banks have a natural incentive to exchange their own debt for that of their rivals (Selgin 1988, White 1989). A competing bank would accept a rival's note at par in exchange for his own bank debt because, "If the notes acquired are redeemed sooner than the notes issued, interest-earning assets can be purchased and held in the interim."(White 1989, p. 227)

This chapter adds the possibility that a goldsmith-banker accepted a rival's note from the public in

8 An undated letter from Sir Robert Vyner to Arlington reads, "I have beene by many accidents much postpon'd soe yt ye money due to mee is soe farre off that I can not possible make it useful to mee. All Credit in London is much Shortened of late. I am attempting a way to enlarge my owne and doubt not to effect it to his Maties. advantage as well as my owne, if I am (like ye lame dogg) but helpt over this Style." Quoted in Roseveare (1962).(PRO.SP.29/225, Cal. State Paper, DOM, 1667-8, p. 113)

71 anticipation of a clearing. Possession of a rival's debt reduced the likelihood of a negative net

clearing position with that rival. Whether a goldsmith-banker expected to retain a competitor's note

long enough to invest or only momentarily to clear, he did so without his rival's knowledge or

permission. A goldsmith-banker could not proactively force competitors to provide covering

balances or reactively charge fees or interest because the banker was willing to accept rival's note

without these provisions. Moreover, each banker's competitors knew this because they behaved in

the same way. A goldsmith-banker could not credibly threaten to cease accepting other banker's

notes.

Mutual par acceptance expanded the circulation of goldsmith-banker notes and checks

beyond what a single banker's debt would have enjoyed. The positive externality stemming from

widespread par acceptance naturally derived from each banker's desire to swap his notes or demand

accounts for his competitors' debt. High transactions costs of actual redemption stemming from

distance could create non-par acceptance, but because most of London's goldsmith-bankers were

situated close together on Lombard Street, one alleyway south of the Royal Exchange, any

difference in time and effort between redeeming a note with one goldsmith or another would have

been negligible. Such proximity promoted mutual par acceptance and helped goldsmith-bankers

become major suppliers of notes and checkable demand accounts in late seventeenth century

London. With regular acceptance of competitors' debts, goldsmith-bankers could lower their costs

of settling claims by clearing first and then settling net balances. The accumulation of off-setting

debts, however, took time, for the simultaneous creation of off-setting debts would be rare. If

bankers demanded specie redemption the very moment after accepting a rival's note from the public,

Lombard Street would have been filled with goldsmiths, or their apprentices, scurrying between each other's shops. Instead, goldsmith-bankers allowed uncleared balances to build.

72 As with mutual debt acceptance, tolerance of uncleared debts and the consequent savings from clearing off-setting debts grew endogenously out of banker self-interest. Each goldsmith- banker had an incentive to not clear immediately for a combination of reasons. Each debt issuing banker could expect the competition would shortly acquire one of his own notes from the public and demand repayment. Moreover, the cost of paying your rival outweighed the benefits of collecting on his note, assuming both notes are of equal value. If our banker invested the money from a rival's note in an interest bearing asset but the competitor arrives moments later demanding the money back, our banker will face costs liquidating the asset. If our banker does not invest the specie, he has wasted time and money to transport and protect the precious metal just to have it taken right back. The low cost option was to hold the other banker's note until he showed up with one of your notes.

Fear of a negative net position from clearing could motivate a goldsmith-banker to refrain from demanding repayment. How long a banker waited to clear off-setting debts depended on his conditional expectations, assuming risk neutrality. A goldsmith-banker would know the amount of competitors' debt he had accepted from the public but would not know how much off-setting debt a competing banker had accumulated. However, with regular clearing contacts, a banker would be aware of the historical average and variance of off-setting debts. A banker would have an expectation of net clearing balances conditional on the amount of the a rival's notes he had accepted to date. If our banker's conditional expectation was that clearing would bring him an adverse net position, he would have little reason to seek out clearing; rather, he would begin preparing a means of paying his anticipated shortfall. If, however, our banker had a conditional expectation that clearing would work in his favor, he still had a reason to hesitate to initiate clearing. Because the potential cost of facing a negative balance still existed, our banker might retain the uncleared debts as insurance. Uncleared balances insulate a banker's monetary reserves from the potential shock of

73 adverse clearing. As noted above, the cost of redeeming a rival's note would not be recompensed if the specie was quickly returned as repayment on another note. Bankers had an incentive to acquire precautionary reserves comprised of rivals' debts to protect the stability of precious metal stocks.

As described so far, clearing would occur infrequently. Each pair of goldsmith-bankers hesitated to initiate clearing except when faced with a large expectation of a favorable net position,

However, the more time periods two bankers delayed clearing, the greater the variance of their expectations became. Waiting too long created more risk than it controlled. Assume two bankers have learned that they both accept each other's notes from the public on a daily basis with identical, independently distributed normal distributions, so the expected daily net clearing position is zero.

As each day without clearing passes, the expected net clearing remains zero plus or minus modification because of our banker's actual acceptances of his rival's notes. The variance of the possible net clearing position, however, is growing with each passing day.9 The uncertainty of the actual net clearing eroded the benefits of delayed clearing. An expanding variance of the conditionally expected net balance between two bankers opened up the risk of a large drain of reserves in the extreme negative case and exposure to substantial default in the extreme positive case. As discussed below and in Chapter 4, goldsmith-bankers were unlimited liability bankers that appear to have kept substantial reserves suggesting a conservative disposition. Drains of pure base money rapidly reduced a banker's reserves to assets ratio and threatened a banker's liquidity. Fears of reserve drains especially applied to bankers who conditionally expected a negative net clearing position. As the variance rose with time, clearing could well be initiated by a banker who expected to pay but wanted to limit the potential for a devastating one day shock. As clearing was delayed, fear of the unknown grew and eventually prompted clearing.

9 If X and Y are a pair of statistically independent random variables then V AR(X+Y)=VAR(X)+VAR(Y).(Newbold 1990)

74 Clearing, however, need not require the actual movement of specie. Recall, both banker's might initiate clearing in the face of growing uncertainty, but once the clearing has occurred, the actual net position would likely be modest relative to the extreme potentials. The owing banker would gladly offer the owed banker a new note or positive balance rather than specie. The owed banker might well accept a note or credit to an account so as to limit his chance of facing a negative net position the next time. Moreover, both bankers might regularly retain some of each other's notes or ledger credits as off-setting insurance against adverse net positions. Unless one of the bankers continuously faced negative net clearing positions, both bankers favored passing some level of each other's notes back and forth and avoiding the movement of specie. The records of Edward

Backwell, examined throughout this chapter, suggest that the inter-banker movement of specie to settle net clearings was uncommon. Moreover, the cost of retaining uncleared balances was small.

In an environment where unregulated debt issuers regularly interact, a system of mutual debt acceptance, delayed clearing and retention of precautionary debt reserves developed endogenously.

Bankers had a natural incentive to exchange their own debt with the public for the debt of competitors. The same bankers had an incentive to then build and retain some uncleared balances as insurance against the expensive possibility of specie drains. The clearing of off-setting balances would then occur more to exchange information and limit risk than to move bullion between bankers. An expectation that a negative net position could be settled by debt would encourage more frequent clearing because information could be garnered with little risk of having to move specie reserves. London's goldsmith-bankers naturally coalesced into decentralized network of debt clearing that minimized the movement of specie but promoted the exchange of information.

The story becomes enriched by considering some of the many possible interactions between numerous goldsmith-bankers participating in the system. The goldsmith-banker system produced similar results to centralized clearing but with different methods. As the web of bankers grew with

75 more members, a third banker's debt could pass in lieu of specie payments. The flow of many bankers' notes among numerous bankers mimics the multi-party clearing potential of clearinghouses.

In the classic case, Banker A owes Banker B $10, Banker B owes Banker C $15 and Banker C owes Banker A $10. As a resolution, a clearinghouse would transfer $5 from Banker B to Banker

C's balance. In a decentralized system, Banker A might give Banker B a $10 note. Banker B could then supplement Banker A's $10 note with his own $5 note to settle with Banker C. Finally,

Banker C could return Banker A's note to Banker A and retain the $5 claim on Banker B. Only

Banker B's $5 debt to Banker C remained.

Moreover, the decentralized system of clearing would monitor and discipline members. As

Selgin (1988), Glasner (1989) and Dowd (1993) have pointed out, clearing not only saved transactions costs but controlled the over-issuance of debt by member bankers. Goldsmith-bankers overissuing debt faced adverse clearing and consequent specie outflows. The decentralized clearing system would monitor and discipline members because the system was incentive compatible. Each banker knew the amount of a rival's debt he held and had an incentive to demand specie when uncleared balances persisted beyond amounts justified for precautionary purposes. Bankers and the public could expect the system to correct any banker's overissue of debt. Because of the slack created by precautionary debt reserves, however, an overissuing banker could float debt among his colleagues longer than the immediacy of a clearinghouse would permit. A sample of uncleared debts in Section III, below, finds that over ninety percent of rival's debt that Backwell accepted were cleared within five days, so a potential float was only a few days at best.

The endogenicity of the goldsmith-banker clearing brought systemic resiliency. Bankers only risked their net clearing position if a single banker failed, and the demise of a competitor meant more business for the remaining bankers. When Edward Backwell and the few other largest lenders to the Treasury were ruined by the Stop of the Exchequer in 1672, the remaining bankers

76 continued clearing between each other.(Mitchell 1994) Likewise, Mitchell suggests the Great Fire

of London in 1666 was a boon to those West End bankers fortunate enough not to have been razed

while Backwell's ledgers record that inter-banker activity was maintained although substantially

diminished.(Mitchell 1994, Backwell's Ledgers) The clearing network between goldsmith-bankers

also overcame the deaths of the most senior bankers. Sir Thomas Vyner, Sir Francis Meynell and

John Colvill died in 1665, 1666 and 1670 respectively. Sir Vyner's shop was passed into the hands

of his nephew Robert Vyner, and Sir Francis Meynell's business was carried on by his brother Isaac

Meynell. Both transfers provided a continuation of inter-banker clearing. John Colvill was survived

by his widow Dorothea Colvill who settled many affairs in her account with Backwell after her

husband's death. Without centralization, the goldsmith-banker system of clearing withstood failures,

fires, the plague and deaths of key members because no single member or group of bankers were

essential for creating or sustaining clearing activities.

III. Accounting and the Nature of Banking in London

The nature of the clearing accounts in Backwell's ledgers was markedly different from

Continental exchange banks of the seventeenth century. Goldsmith-banking was dominated by

transactions between customer and banker and not transactions between customer and counter party facilitated by a banker. Unlike exchange banks, London's goldsmith-bankers were constantly acting

as credit creating counter parties. Credit creation allowed for note issuance and fractional reserve

banking. The introduction of bank credit by London's goldsmith-bankers marked a fundamental

break from traditional exchange banking.

The difference in accounting reveals the essential differences in the actual process of debt

clearing between London and Amsterdam. Entries in the ledgers of Amsterdam's Wisselbank

(Exchange Bank) always listed a corresponding counter party other than the bank. This was

77 because the Wisselbank only honored orders to pay backed by positive balances.(van der Wee 1977, van Dillen 1934, Assar 1987) In turn, the Wisselbank's system of accounting listed the off-setting counter party in the description of all transactions. (Amsterdam Municipal Archives) Although fully aware of how the Wisselbank handled its affairs, London's goldsmith-bankers cleared debts without regular cross referencing (referral to an off-setting counter party account). Instead of clearing between accounts, Backwell himself became a principal. The goldsmith-banker assumed the off­ setting debits and then later cleared the transaction. By stepping into the clearing process as a counter party, Backwell created credit.

The differences recur when dealing with how the intermediaries accounted for themselves.

To keep its system of accounting consistent, the Wisselbank listed itself as a counter party. When a customer deposited (withdrew) money at the Wisselbank, the Exchange Bank's strong room was debited (credited).(Amsterdam Municipal Archives) Backwell did create personal accounts for his household expenses and bullion transactions, but no explicit account was kept by Backwell as a counter party relative to his customers. Instead, Backwell's position is the omnipresent but unstated mirror image of his ledgers.

London's bankers did not maintain an explicit account for themselves as counter parties because they performed a different type of banking than found at the Wisselbank. The foremost examples of this change were bank notes. The Wisselbank did not operate with bank notes

(promises to pay), only orders. The creation of short run debt accompanying the widespread use of bank notes in London meant a much higher level of banker to customer transactions than

Amsterdam's predominantly customer to customer system. An explicit account for Backwell himself would have meant a veritable duplication of paperwork. The cost in time and material would have been substantial. The relative number of banker to customer transactions in Amsterdam were much smaller, so the logging of the Wisselbank's own account was manageable.

78 Unlike an exchange bank, Edward Backwell had to balance the benefit of delaying clearing and retaining precautionary debt reserves against the risk of large net positions and abuse by other bankers. By waiting to clear, Backwell risked facing a large specie outflow or a substantial exposure to a colleague's default. Moreover, rivals might attempt kiting schemes that paid one banker's debt with credit from a new banker. Backwell and the other goldsmith-bankers, if we accept Backwell as representative, curtailed their exposure to abuse by monitoring balances and by clearing accounts when risk and expectations warranted.

This section presents two categories of evidence from Edward Backwell's ledgers to support the view that the goldsmith-bankers were engaged in decentralized clearing that was responsive to information and conditional expectations. First, a change in Backwell's accounting system that improved the banker's ability to monitor his own accounts allowed Backwell to become more tolerant of uncleared balances. An improved knowledge of the uncleared debts he held (his precautionary debt reserves) lowered the variance of Backwell's net clearing positions with other goldsmith-bankers. Second, a micro-level examination of three goldsmith-bankers' accounts with

Backwell over the year 1670 reveals that Backwell initiated clearing when he faced a growing exposure to uncleared debts beyond his desired precautionary reserves.

A. Accounting

Monitoring balances through careful, timely accounting aided Backwell's control over

uncleared debts.10 Backwell had a system of accounting as an institutional framework that

informed the banker of members' balances with him. Information was not cost free. Many models

of inventory control assume that acquiring knowledge of account balances to be effortless (Miller

and Orr); however, in the seventeenth century, information costs were a real constraint. The more

10 Appendix C provides additional explanations of Backwell's accounting system.

79 quickly, accurately and cheaply bankers could stay informed of balances, the better they could control their running debts. The variance of balances would fall as a banker could quickly address

uncomfortably large negative balances. Very large exposures could be attended to quickly, so average balances could be reduced. A diminished risk of dangerously large overdrafts lessened a goldsmith-banker's demand for precautionary debt reserves.

During the years covered by the ledgers of Edward Backwell, the goldsmith-banker altered his system of accounting towards improved monitoring. In his ledgers, Backwell switched from a system of independent entries to a system of connected double entry accounting. Throughout his

ledgers, Backwell added each folio's credits and debits separately to balance each account. The

innovation meant Backwell began pairing each transaction with a specific off-setting entry. Before

1666, Backwell did not record entries as balancing a particular pre-existing entry. The practical effect of knowing on the spot which entries had been cleared was a quicker knowledge of balances still outstanding. Without connected entries, the entirety of the debits and credits since the last totaling had to be added and the difference found to know that account's balance at that moment.

With precise write offs, only those debits or credits not yet cleared needed to be summed, so transaction costs were lessened and balances made easier to ascertain.

An example from the accounts of Thomas Row will assist. Table 2.2 presents the goldsmith-banker Thomas Row's clearing account with Backwell for March of 1664/5 and January of 1665/6. The ledger (Ledger N) covering the nine months in between is not extant. In March of

1665, the descriptions of debit and credit entries make no mention any off-setting transaction.

Debits (on the left) are paid out by Backwell to Row ("him" entries) or others such as Robert

Blanchard, a goldsmith-banker, and Morris & Clayton, a partnership of money-scriveners. On the right side, Row's account is credited for money and notes deposited by Row and other individuals

80 such as fellow goldsmith-bankers Sir Thomas Vyner and Sir Francis Meynell. The credit and debit transactions bear no immediate relationship with each other.

By January of 1666, the system of accounting had changed." Virtually ever entry now had a precise off-setting transaction. On both the debit and credit sides, many entries were described solely in reference to earlier transactions. On January 5, Row was debited for £450 with the description "the 4th." The description refers to a note for £450 that Row had credited to his account one day earlier. Row was withdrawing funds to balance a specific credit. The same occurred on the credit side. On January 5, the ledger described Row's credit of £300 as "the 4th." Row was clearing his account with Backwell with regards to the specific debit of £300 that Backwell paid to

Mr. Boone from Row's account. Even the credits with named descriptions exactly match debits such as with Conguard on the 15th, DiBusty on the 19th, Conyers on the 23rd.

" The diffusion of the accounting innovation was complete across all of Backwell's accounts in Ledger "O" (1666). Thomas Row, the goldsmith-banker in Table 2.2, had been Backwell's apprentice, so the new accounting was not functioning within Backwell's closest circle before the general changeover. Over the life of the missing Ledger "N" (March 25, 1665 to December 31, 1665), the switch occurred. Where innovation came from and why the innovation was adopted when it was remain unclear. The timing of the innovation coincided with a very hectic period for Backwell. While out of the kingdom on Charles II 's business in 1665, Backwell suffered a run that required the Crown to pronounce full backing of the tallies and treasury orders held by Backwell.(Clark 1939) The switch also paralleled the introduction of a new hand to the ledgers, likely Charles Duncombe. The beginning of the new handwriting corresponds with the year of Duncombe's entering his apprenticeship with Backwell and matches the handwriting on the back of a note written by Backwell to Duncombe held today by the Royal Bank of Scotland. Duncombe's arrival came soon after the loss of Backwell's cashier Robert Shaw, who was killed in the plague that swept London in 1665.(Clark 1941) The shocks of 1665 may have created a break in the natural order of business that allowed Backwell an opportunity to adjust his banking practices. The events of 1665, especially the untimely loss of his cashier, may also have jarred Backwell into instituting an accounting regime allowing for easier monitoring of his own staff.

81 Table 2.2

Account of Thomas Row at Edward Backwell's Shop (a) March 1664/5 Day Debits £ s d Day Credits £ s d 1 Wm Home 1000 6 Johna Dawes, which 100 was reed at Dan Edwards 3 Ball 200 7 mony 707 8 Hugh Mason 86 9 Wynne 30 ordered the 13th day of Feb 8 Morris & Claiton 96 12 6 9 Mild gold at 15d, 60 63 15 9 him old gold 71 6 8 10 R Holder 76 6 4 16 so much Joseph pd 100 11 Hen Ballow 100 his man yesterday 17 Blanchard 50 15 mild gold 50, 7 60 11 3 17 him p contant 500 16 mild gold 250 at 15d 265 12 6 18 Dunkin 160 16 several! 666 10 4 20 him p Contant 600 17 milled gold at 15d, 21 5 20 20 him p Contant 300 18 mild gold, at 15d,15 15 18 9 17 Nic Warren 29 7 11 20 three notes rec of 499 18 him 16 one salver & cup 21 11 8 20 T Vyner 273 15 74oz at 5s lOd 22 lent old gold 350 350 21 mild gold, 191 at 15d 20 3 9 23 him p contant 500 21 Ball ordered 10th Feb 370 10 23 wm Prickman 130 21 Finch ordered 23 Feb 100 24 content p Robt 500 24 Meynell 400 Porter 24 Savage 400 Table 2.2 (cont.) (b) January, 1666 Day Debit £ s d Day Credits £ s d 1 20 milled gold at 21 13 4 1 note b110 16 1 20d p Tho Temple 4 his man "100 1 note MOO 4 his man b110 16 1 4 note '450 4 Boone "300 5 the 4th c300 5 the 4th -45O 5 note 118 15 9 Corcellis '60 10 money 500 10 DiBusty f254 14 10 money 300 10 Sr John Fredericke g200 13 Adrian & Manrois 400 200, Bathurst 200 11 Snow 400 15 the 15th 275 14 7 12 Dulivier 223 8 9 15 the 10 »«4 14 54

12 Freeman 160 15 the 12 160 12 Frohocke 44 15 the 12 165 3 9 12 Dulivier 165 3 9 15 the 12 223 8 9 15 Dunkin 275 14 7 15 the 4 e60 15 the 5th 118 15 15 the 12 44 15 the £800 400 15 Conguard 50 15 Adrian 200, 400 17 the 10 100 Bathurst 200 16 Wm Batteshay 100 19 Lawr DiBusty 588 4 5 18 Coguards 50 20 mild gold 300 at 19d 323 15 18 Dibusty 588 4 5 23 Conyers 100 22 Conyers 100 26 Wade 100 24 mild gold 323 15 24 small cup 18 6 26 Wade 100 29 the 27th 251 3 4 27 Dibusty 251 3 4 The change in accounting produced very precise folio balances. Under the earlier regime,

the difference between credits and debits at the end of a folio would be transferred to the top of the

next folio for the continued life of the account in question. With the change to connected entries,

these residuals all but disappear. After the switch, most folios end in perfect balance: not one pence

difference between debits and credits. The major benefit to Backwell and his customers of the new

connected entries system was not a cleaner ledger. The ledger was a final record rather than work

in progress. The change in the final ledgers revealed that a significant adjustment in the point of

transaction accounting did occur between March of 1665 and January of 1666. While we do not

know how the exact process of recording transactions worked, Backwell did begin pairing off­

setting payments at the front desk.

The change in accounting procedures reduced the average and the variance of the aggregate

balances held by Backwell. From March 1663 to March 1665, Backwell shouldered an average overdraft £1,037 larger than after 1666. The standard deviation also fell from £5,792 to £4,311.

Importantly, the total pound volume of transactions increased in the later period. Measuring debit transactions only, the average daily debit rose from £1,208 to £1,642, so the fall in daily average and variance of the banker's balances was not driven by less business.

The character of Backwell's clearing activities became relatively more tolerant of initial overdrafts but more expeditious in reducing ongoing negative balances with the innovation in accounting. The improved flow of information allowed Backwell to reduce overdrafts quickly which lowered the average balance and the overall variance despite a larger flow of clearing.

However, the new accounting regime allowed for more variance relative to average balances. The coefficient of variation, the standard deviation divided by the absolute value of the mean, rose from

2.57 before 1666 to 3.55 afterwards. The rising coefficient of variation indicates that the mean fell

faster than the variance. With his improved information, Backwell was not restricting initial debt

84 acceptances. Rather, the goldsmith-banker was tolerating even larger precautionary debt balances because he could now better control his exposure to very large net balances in his favor. Backwell still did not know the off-setting debts held by rivals until the time of clearing, so the precautionary benefits of holding competitors' debts remained while the risk of retaining uncleared debt fell.

B. Clearing

Beyond monitoring his exposure to fellow bankers, Backwell actively cleared debts on a daily basis. The author proposes that the debts between bankers were cleared when one of the two bankers felt either 1) that the expected value of net clearing, conditional on his own holding of the other banker's debt, was greater than the precautionary reserves that banker desired or 2) that the risk of a large net position, with the passage of time, was becoming too great. This section looks for the manifestation of these implications through a micro-level examination of a sampling of

Backwell's clearing accounts. Unfortunately, the ledgers of Backwell's colleagues are not known to still exist, so Backwell's relationship with other goldsmith-bankers can only known from Backwell's perspective. Lacking a view of the full clearing pattern, this chapter examines behavior for consistency with the analytical predictions of a decentralized clearing process.

However, certain features of the decentralized clearing system that this chapter presents are testable given only Backwell's side of the events. Detailed consideration of sampled accounts do prove consistent with the view that the goldsmith-banker did not allow debt to go uncleared for more than a few days. This supports the proposition that the growth of risk as clearing was delayed encouraged the initiation of clearing. This chapter's thesis also suggests that Backwell would initiate clearing when a goldsmith's negative balance exceeded Backwell's desired precautionary reserves. The data do not state who initiated clearing. However, until clearing occurred, the other goldsmith-bankers did not know the level of their accumulated uncleared debts with Backwell.

85 Therefore, the timing of clearing should have been correlated with the inflow of other bankers' debt

into Backwell's shop (/"Backwell was initiating clearing to prune excess reserves. Moreover, we would only expect Backwell to be initiating clearing in response to debt inflows when the account

in question was negative. In contrast, if another banker's account with Backwell was positive, we would not expect clearing to correlate with the inflow of debits into Backwell's shop. In this case,

Backwell would want to retain the other banker's debts to diminish a potential outflow of specie.

The statistical analysis below finds that the frequency of clearing increased when two of the sampled goldsmith-banker's uncleared debts accumulated. Both goldsmith's usually ran negative balances with Backwell, so Backwell carried precautionary reserves against both and had an incentive to avoid excess uncleared debts. Moreover, the third sampled banker held an overall positive balance with Backwell, so we know this banker held precautionary reserves on Backwell.

Unlike the other two accounts, the timing of clearing of the third banker was independent of the flow of uncleared debts into Backwell's shop. These findings are consistent with the analytical framework that Backwell initiated clearing when the negative balances he held grew particularly large. The third case of a positive balance is consistent with the proposal that the other banker was initiating clearing with Backwell. Lacking the third banker's ledgers, this possibility cannot be further addressed.

The post-1666 accounting system allows us to calculate when clearing debits with Backwell were created and when they were cleared. Again, this information exists because Backwell went through the trouble to collect and maintain these records. By his own efforts, the goldsmith-banker revealed that he valued timely knowledge of when debits were created and retired. The regression analysis below considers the clearing accounts of three goldsmith-bankers on a detailed level for the year 1670. The goldsmith-bankers, Thomas Row, John Lindsay and Benjamin Hinton have been selected because all three operated on Lombard Street with Backwell and all three had clearing

86 activities of similar sizes. Charts 2.4.1-3 report the daily pound value of debits on Hinton, Lindsay and Row respectively that Backwell recorded as needing to be cleared in 1670. Table 2.3, below, provides the summary statistics for each banker's distribution of individual debits to be cleared with

Backwell in 1670. Although Benjamin Hinton had fifty percent more debits requiring clearing than

Thomas Row did, all three bankers had similar flows of notes by the public into Backwell. All had identical medians and modes of £100 except for Lindsay's median of £116.

Table 2.3

Characteristics of Debits with Backwell to be Cleared,

By Banker in 1670

Benjamin Hinton John Lindsay Thomas Row Mean £161 £229 £173 Median £100 £116 £100 Mode £100 £100 £100 (No. of Modal (64, 14%) (48, 13%) (33, 10%) Observations, % of Total Observations) Standard Deviation £217 £317 £201 Minimum £5.40 £9.95 £6.75 Maximum £2,000 £2,911 £1,447 Number of Observations 457 380 324

The three bankers, however, had different backgrounds. Thomas Row had been Backwell's apprentice until gaining his freedom in 1664.(see Table 2.9) Row provides a proxy for the relationship between close bankers. Row and Backwell had close business dealings including the loan of £30,000 from Row to Backwell in August of 1670. John Lindsay was free of the City of

London as a member of the Goldsmiths' Company in 1658 and had regular dealings with Backwell over the life of Backwell's extant ledgers. Lindsay represents an older, proven, long-run business

87 Pound Sterling

3" <-+ O 3 co" C 3 o CD 0) (D Q. O O 3" CD D) ft > ro O • O 4^ CD T3 CD £L

00 0) O I

88 Pound Sterling

a.3

CO

CO C 3 O, 3CD o Q. O 0) CD ST ro >o o 4^ CD ro "SD " Q.

00 0) Io

89 CO Pound Sterling O ro ro c OI o 01 o 01 O o o o o (D O o o o o l _i 1_ i J3 O

CO C 3 O. CD 0) —* CD Q. O o CD 3" W 0)

>o ro o CD 4^ s GO

Q.

00 (oD

90 colleague. Benjamin Hinton ended his apprenticeship in 1666, but he did not begin clearing activity with Backwell until late in 1667. Given the fire in 1666 and the runs on Lombard Street in 1667,

Hinton's delay seems warranted. Hinton represents the status of a newcomer to the network.

The three goldsmith-bankers also differed in their overall balances with Backwell. Charts

2.5.1-3 present Hinton's, Lindsay's and Row's total daily balances with Backwell over 1670. Both

Lindsay and Hinton usually had negative balances while Row's balance was usually positive.

Hinton's and Lindsay's daily balances also had higher variances than Row's account had. Chart 2.6 adds a graph of the daily balance of all goldsmith-bankers with Backwell over the year 1670.

Backwell's total exposure to his fellow goldsmiths fell considerably after the summer of 1670. The dramatic flip in goldsmith-bankers' overall balance was caused by two notes presented by George

Snell to Backwell on June 13, 1670, for £6,000 payable on July 13 and £5,500 payable on August

13 of that year.12 Table 2.4 reports the summary statistics for the goldsmith-bankers' balances.

Table 2.4

Characteristics of Banker Balances with Backwell in 1670

Benjamin John Thomas All Bankers Hinton Lindsay Row Mean -£543 -£370 £90 -£2,452 Median -£354 -£190 £152 -£2,126 Standard £607 £591 £407 £6,048 Deviation Minimum -£3,026 -£3,611 -£2,243 -£17,480 Maximum £238 £784 £2,402 £17,476

12 The £6,000 was duly paid by Backwell on July 13, 1670. The £5,500 was paid off by installments of £1,000, £1,801:15:9, £1,000 and £1,098:4:3 on August 13, 15, 17 and 19 respectively. On July 13, Snell was credited for an additional £5,000 that was cleared over the following week.

91 Pound Sterling CO O &) rb rb c Okl oi o OI o Ol Ol o o o o o o o (D o o o o o o o o o o o _l_ 00 I 9) hi 00 «r -nl 2_ Ell 0) 3 (Q fl)i o CD O w > O —fi 00 -° §i CD 3. *&' ro 3. Ol 3' I 3' I! 3 §1 D 81

92 Pound Sterling oCO c 6 cb o Ol o o (D o o

00 2. 3 O o CD O —*% c_ 3. o ro in

CO

93 Pound Sterling CO O rb rb ± ro ro c O Ol Ol O Ol o 01 oOl O Ol o o o o o (D o o o o o o o o o 8 8 o o J L J I

00 0) 0) 3 O o CD =r O 0) -4% —i !-+• 3H" O ro 3 Ol 0) CO CO ZD O

94 Pound Sterling CO o c (Thousands)

00 2L 0) 3 O CD CO O

o 3- CO ro 3 b)

0) 3 CD

95 Given all three bankers had very similar flow of debits into Backwell, the considerable difference between Hinton's and Lindsay's negative balances on the one hand and Row's positive balances on the other must stem from the credit side of their accounts with Backwell. Thomas Row presented to Backwell many more notes for credits than did the other two goldsmith-bankers sampled. The initiating credits summarized in Table 2.5, below, are entries that Backwell recorded as himself owing the other bankers. These credits were then later cleared by an explicit debit. The calculations in Table 2.5 exclude nearly £42,000 in bullion that Row was credited with from August

20, 1670 until finally being completely repaid by the East India Company in January 1671.

Table 2.5

Characteristics of Credits with Backwell to be Cleared,

By Banker in 1670

Benjamin Hinton John Lindsay Thomas Row Mean £137 £257 £301 Median £100 £140 £163 Mode £50, £100 £100 £200 (No. of Modal (5, 11% each) (4, 8%) (3, 4%) Observations, % of Total Observations) Standard Deviation £146 £305 £440 Minimum £10.7 £31 £20 Maximum £730 £1,300 £3,011 Number of Observations 45 48 71

The larger inflow of Backwell's notes from Row relative to Hinton or Lindsay suggests Row retained more of Backwell's notes (net of clearing) than did the other two goldsmith-bankers. When time came to clear, if Backwell did not have an offsetting note of Row's, Thomas Row was likely credited by Backwell for the surplus notes Row held. Row's consistently positive balances would

96 be explained if Row was routinely presenting Backwell with more notes than Backwell was confronting Row with. Lindsay's and Hinton's consistently negative balances might be explained

by the reverse situation, Backwell was receiving more of their notes (Table 2.3) from the public then they were receiving of his notes. Unfortunately, we lack the other bankers' ledgers, so the hypothesis cannot be verified directly.

However, the implications of who was initiating clearing can be examined by calculating the time each debit Backwell accepted took to clear. The same accounting system that permitted the separate examination of debits (goldsmith-banker notes and checks) presented to Backwell by the public and credits (Backwell's debts) presented to Backwell by account holding bankers also allows the calculation of how long entries took to be cleared. The days until clearing for each of the three bankers are presented in Chart 2.4.1-3 and reveals prompt clearing occurred. Table 2.6 sorts the observations by the days each debit took to clear. The days until clearing include Sundays, so these figures overstate clearing times for days greater than zero (the same day). Moreover, the bias increases with the span of time until clearing. A majority of notes are cleared after one or two days and over ninety percent are cleared within one week. Inter-banker debt clearing was a recurrent, short run affair. Thomas Row and John Lindsay are striking in Table 2.6 for their similarity.

Moreover, both were routinely clearing faster than Benjamin Hinton. In two days, Lindsay and

Row each averaged the clearing of three quarters of their respective notes. Hinton managed only two-thirds over the same period. The former pair of bankers both broke ninety percent clearing in less than five days while Hinton took two additional days. Overall, however, debts were cleared in a timely manner consistent with the theory that the rising variance of net clearing balances, which accompanied delay, promoted settlement.

97 Table 2.6

Days to Clear Banker Debits

0 = Same Day

Note: Days to Clearing includes Sundays

Benj amin Hinton John Lindsay Thomas Row Days to Debits per Cum % Debits per Cum % Debits per Cum % Clear cent cent cent 0to

The regression analysis tests whether Backwell was likely to initiate clearing with goldsmith-bankers that ran large deficits. The model takes the correlation between the average time to clear and the inflow of debts into Backwell's shop as a measurement of Backwell's initiating clearing. Since the amount of Backwell's acceptance of rival's debts from the public is not known to the other bankers until clearing, settlement initiated by other bankers should be independent of inflows. The hypothesis is that size of Lindsay's and Hinton's debts and deficits will be correlated to the time their uncleared debits took to clear. The debts and surplus of Thomas Row, however, are expected not to be correlated with the timing of debt clearing between Row and Backwell.

98 The following econometric model has been constructed with the number of days a debit took to clear serves as the dependent variable.

t+6 f-1 r+6 liuT,) = «+£YA + £ bfl, * £ (,r, + AC?, + r\W, * QM, * e, i=t i=t-6 i=f-6

where: Y, = Days to Clear a Banker's Debit originating on Day t D, = Value of that Banker's Debits on Day t B, = Value of that Banker's Balances on Day t T, = Value of Backwell's Balances with All Bankers on Day t Q, = Time Trend on Day t W, = Dummies for the Days of the Week on Day t, Wednesday always=0

Mt = Dummies for the Month on Day t, January always=0 e, = error on Day t.

The pound value of a goldsmith-banker's balances and debits are explanatory variables that capture whether Backwell is responding to the flow of notes into his shop from the public. The balances of a given banker's account before the day of the debits in question measures that account's recent history. The pound value of the actual day's debit (day=t) captures the possibility that Backwell might clear larger notes faster. The stream of debits from future days (t+1 through t+6) reflect that Backwell might speed up clearing as more notes arrive. The hypothesis is that the time until clearing for Lindsay and Hinton will be negatively related to the value of each banker's debits received by Backwell. The dependent variable, however, should be positively related to each bankers' balance because negative balances promote shorter clearing times. For Thomas Row, however, both the inflow of debits and the size of balances should be unrelated to the time until debts were cleared.

99 The dummy variables for each day of the week allow for the possibility of recurrent clearing on any particular day. For example, if Fridays developed into a day of regular weekly clearing, the

Thursday dummy variable should consistently indicate that deposits on Thursday clear more quickly than deposits on Wednesday. The dummy variables for each month control for monthly seasonal behavior.

A number of parameters are beyond the measure of this model. Foremost, if Backwell decided not to accept a certain banker's note, the entry would never have been recorded. Also, notes in the seventeenth century often had conditional terms of repayment such as "payable on three days after sight." Backwell's ledgers lack the information to control for such possibilities or to even suggest how common conditional terms of repayment were within the goldsmith-banker clearing system.

The dependent variable, Days to Clearing, has been logged to report percentage changes.

The model has remained semi-log because many of the dependent variables have observations of value zero. A model for each of Hinton, Lindsay and Row was selected from the range of forty-two variables described above and is reported in the Appendix.,3 AIC and SBC criteria were used to select the particular model for each banker.M However, to promote clarity of comparison, a second series of models of identical structure were regressed. Each variable with a ninety-five percent significance or better from each banker's first model were pooled into a single model which was imposed on all three bankers. The regression results from this second series of models are also reported in the Appendix D. The imposition of a homogeneous model caused the second series of

13 The dependent variable, Days to Clearing, has been aggregated to the daily level with same day observations weighted by the debit's pound value.

14 Both the AIC and the SBC criteria balance the fit of the residuals with the number of parameters. Model selection favored the model for each banker that minimized both criteria. The AIC equals N*ln(SSE/N)+2P and the SBC equals N*ln(SSE/N)+P*ln(N) where SSE is the sum of squared errors, N is the number of Observations and P is the number of parameters.

100 estimates to lose explanatory power relative to the heterogeneous first series; however, both series provide similar conclusions. The results from the homogeneous second series are presented in

Tables 2.7 and 2.8 as the percentage change in Days to Clearing for each independent variable.

Table 2.7

Percentage Change in Days to Clearing

for each Independent Variable's Mean Value

(Bold indicates coefficient significant at the ninety-five percent level)

Hinton Lindsay Row

Debit 1.8 -9.3 -2.9

Debit t+1 -7.1 -5.3 -0.3

Debit t+6 4.0 -4.4 -2.6

Balance t-3 0.7 5.6 2.3 Balance t-4 4.3 -10.6 0.2 Balance t-6 -10.0 -0.0 1.8 Monday -19.6 -21.1 -1.3 Saturday 23.7 14.0 24.4 Time Trend -40.1 -32.7 -44.2 February -64.0 1.2 12.1 March -38.8 -19.5 -22.9 May 26.1 -21.3 31.0 August 2.2 36.2 3.3 November 10.9 -49.2 -25.0 Total t-3 -0.4 -3.6 0.5 Total t+1 3.6 7.5 -0.5 Total t+3 -0.2 0.9 -2.7

101 Table 2.8

Percentage Change in Days to Clearing

for each Independent Variable assuming

a One Hundred Pound Observation or One Day for the Time Trend

(Bold indicates coefficient significant at the ninety-five percent level)

Hinton Lindsay Row Debit 0.57 -2.11 -0.97 Debit t+1 -3.36 -1.87 -0.18 Debit t+6 1.90 -1.06 -1.68 Balance t-3 0.11 1.28 2.04 Balance t-4 -0.73 2.47 0.17 Balance t-6 1.68 0.00 2.18 Monday -19.56 -21.07 -1.32 Saturday 23.68 13.95 24.33 Time Trend -0.22 -0.18 -0.24 February -64.04 1.22 12.12 March -38.84 -19.47 -22.89 May 26.11 -21.26 30.99 August 2.27 36.19 3.28 November 10.95 -49.17 -25.01 Total t-3 0.02 0.13 -0.02 Total t+1 -0.14 -0.21 0.02 Total t+3 0.01 -0.04 0.11

The blocked areas of Table 2.7 highlight the distinctions between Benjamin Hinton and John

Lindsay on the one hand and Thomas Row on the other. Under identical models, Row's Debits at time t and t+1 have poor significance and weak coefficients relative to the estimates for the other two bankers. Lindsay's variables for originating Debits and next day Debits are significant with strong negative values relative to Row. Likewise, Hinton's t+1 Debits are significant and negative.

Table 2.8 confirms these findings by reporting percentage change in the dependent variable for a one hundred pound change in the monetary dependent variables instead of the mean value used in

Table 2.7. The standard units reveal each one hundred pounds of debits packed much more of a

102 quickening effect for Hinton and Lindsay than for Row. The lagged balances of each banker's tell a similar but weaker story.

We can conclude from the regression analysis that the debits of John Lindsay and Benjamin

Hinton, both of whose accounts had negative balances, were being cleared faster as Backwell accepted more debts. This relationship agrees with the proposal that Backwell initiated clearing when facing uncleared debts beyond his desired precautionary level. Moreover, the debits of

Thomas Row, whose balances were generally positive, demonstrate no strong relationship with the time debts took to clear. This finding agrees with the proposal that Backwell was not initiating clearing with Thomas Row in response to new debts. Recall Backwell accepted similar streams of

Lindsay's, Hinton's and Row's debts from the public. The differences in their relationships with

Backwell stem from the their overall account balances. Finally, the possibility that clearing occurred consistently on some day or days of the week for individual bankers or as a group is hard to support with the regressions results. Debits on Mondays tended to clear more quickly while

Saturdays definitely cleared more slowly, but these results could easily be explained by the lack of business on Sundays.

TV. The "Gold Boys" Network

Although responsive to large uncleared balances, the decentralized system of goldsmith- banker clearing exposed participants to a risk of default. London's goldsmith-bankers reduced their exposure to fraudulent colleagues through the screening process of apprenticeship. Most new bankers to the clearing network were filtered by apprenticeship to established goldsmith-bankers.

The author is not saying the "old boy" clique of goldsmiths created an absolute barrier to entry.

The endogenous nature of the system as described above meant existing players had an incentive to transact with newcomers. The author has no evidence that the network had formal rules or acted

103 jointly to punish or exclude. Rather, the connections created by apprenticeships and years of

business dealings created a relative advantage for network members. Trust allowed for tolerance of

debt and less recourse to specie settlement.

Trust between goldsmith-bankers was a function of information and commitment. Bankers

had to know that their colleagues had more to gain from future business than from fleeing London

with all the bullion they could carry. Unlike the Maghribi traders story developed by Greif (1989,

1993), London's goldsmith-bankers could not rely on internal punitive measures. A banker that had

absconded would not be punishable by the group. The lack of repeated play for abuse meant

colleagues had to monitor each other's business health, non-transportable assets and family ties.

Apprenticeship, proximity and social interaction reduced the risk members faced when holding

uncleared balances.

The informal system, however, did have a formal backbone - the Goldsmiths' Company. As

one of the oldest and wealthiest of London's great livery companies, the Goldsmiths' Company was

entrusted with guaranteeing the purity of worked precious metals. The goldsmith-bankers formed a

powerful yet informal sub-set of the larger company. The goldsmith-bankers used the Goldsmiths'

Company's formal system of apprenticeship to train new bankers. Throughout London's trades,

"seven years' 'genteel servitude' remained the commonest introduction to the world of

business. "(Earle 1989, p. 86) In exchange for seven years of non-wage skilled labor and often an

initial fee, the master taught the apprentice the necessary banking skills, introduced him to

established bankers and developed the ground work for a long professional relationship. '5

Generations of such apprentices turned banker under the same master produced networks of

goldsmith-bankers related by apprenticeship. Family trees of who was apprenticed to whom emerged.

15 Many apprentices appear never to have made it to become freemen of the City much less bankers. (Goldsmiths' Co.)

104 Chart 2.7 demonstrates how the earliest bankers like Sir Thomas Vyner and Robert Welstead produced numerous banking off-spring by apprenticeship. However, the majority of apprentices appear not to have become bankers, so the process did act to filter potential goldsmith-bankers.

Moreover, successful apprentices like Edward Backwell in turn produced more goldsmith-bankers.

For example, a chain of great financiers to the English government ran from Sir Thomas Vyner under Cromwell through Alderman Edward Backwell under Charles II to Sir Charles Duncombe under William III. The extended families of goldsmith-bankers continue into the eighteenth century.

Chart 2.8 presents additional apprentice relationships between generations of goldsmith-bankers.

Not all goldsmith-bankers might have come up through the London apprentice system, but the author knows of no Restoration era examples. Richard Kent was in partnership with the goldsmith

Charles Duncombe, but Kent worked as a revenue officer and had no record of a goldsmith's training.(Goldsmiths' Company) Rather Kent's association with Duncombe was typical of the marriage between bankers and tax collectors examined in Chapter 3.

During the years covered by Backwell's ledgers, a number of apprentices gained their freedom and set up independent shops. Backwell began clearing accounts with them soon after and sometimes even before their date of freedom. Table 2.9 lists goldsmith-bankers who gained their freedom of the City of London and began clearing with Backwell during the years covered by his ledgers. Included in the table is Thomas Cooke who ended his apprenticeship under Thomas Row in 1671 and Row himself who gained his own freedom in 1664 having been apprenticed to Edward

Backwell.

105 >

CD 3 o" (D (0

C 3" 3 ® o °CD- 5O oo 3 ro u>

CD c° I! o Tl —n CD CD Q. O 3

106 o° c 3 o O —n o CD 8- I. > T3 5 •D -ni CD 3 3 r-H o I CD H b" CO 3" -H CD o 3 C —t O 3 Q. (D o Q. CD 00 O U) CO 3- 3 Z ro CD CD —x i—t- CO Z bo O -< -n CD 7T 0)

Tl -t CD CD Q. O 3

107 Table 2.9

Apprentices Turned Bankers Clearing with Edward Backwell

Apprentice Master Date of Freedom Date of First Clearing Entry with Backwell

Thomas Cooke Thomas Row 19 January 1671 11 May 1671

Benjamin Hinton George Day 29 June 1666 15 November 1667

Joseph Horneby Edward Backwell 6 July 1666 20 August 1666

Thomas Kirwood John Lindsay 4 October 1668 7 January 1669

Thomas Pardo Robert Welstead 17 May 1671 8 April 1671

Thomas Row Edward Backwell 15 July 1664 18 April 1664

Robert Ryves Robert Welstead 16 August 1671 1 September 1671

Bernard Turner Robert Welstead 8 October 1664 15 February 1665

Thomas Williams Edmund Hinton 29 September 1669 3 February 1670

Sources: Goldsmiths' Company, Backwell's Ledgers

Apprenticeship allowed the master and the other bankers time and proximity to judge a newcomer. The relationships built during apprenticeships were important because no explicit custom or legal statement is known that delineated who was in or not in the system. Being a system of inter-banker clearing and not a clearing house, the arrangements between each banker were different and personal relationships became important. Each banker faced no known restrictions in tailoring their exposure to other bankers.

Moreover, the network of clearing may have been systems within a system. For example,

Backwell cleared extensively with his fellow apprentices of Sir Thomas Vyner (George Snell and

108 Sir Robert Vyner) and the second generation of Vyner apprentices (Jeremiah Snow, Thomas Row and Joseph Horneby) but much less so with those of Robert Welstead and his apprentices Bernard

Turner, Robert Ryves and Thomas Pardo. Undoubtedly, apprenticeship generated information and loyalties that supported greater ties and trust. The quantitative sources from only one banker's records, however, will not separate the effects of apprenticeship relations from volume of business in determining the extent of clearing.

V. Conclusion

Edward Backwell conducted extensive clearing operations as a member of London's seventeenth century network of goldsmith-bankers. The quantitative evidence presented in this paper demonstrates that Backwell was managing substantial levels of uncleared debt payable by nineteen of his fellow bankers. With the introduction of an innovative accounting process Backwell was better able to monitor his exposure to negative balances. The risk of extreme net clearing positions was also mitigated by clearing usually within a few days of a debt's acceptance. Evidence also suggests Backwell hastened the initiation of clearing with rivals facing negative balances as

Backwell's holding of their uncleared debts grew.

The evidence presented above supports this chapter's analytical description of London's decentralized clearing process. Backwell did not clear all rivals' debts immediately, nor did uncleared debts accumulate for weeks or months. The overall promptness of clearing supports the contention that the accumulation of risk created by delayed clearing spurred settlement. Moreover,

Backwell's responsiveness to negative balances bolsters the view that the goldsmith-banker retained uncleared balances as precautionary reserves against possible future adverse clearing positions.

When negative balances grew uncomfortably large, Backwell initiated clearing.

109 The emergence of endogenous, decentralized clearing among bankers began with the mutual acceptance of debt by rival bankers as previously spelled out by other authors.(Selgin 1988, White,

Dowd 1993, Glasner 1989) This chapter contributes the follow-up insight that a delay of clearing arose endogenously from competing bankers' desire to hold precautionary reserves. Such delay was essential for off-setting debts to accumulate and the saving of transaction costs to be realized in the absence of centralized clearinghouses, exchange banks or central banks. Moreover, the system was self regulating because each banker's incentives increasingly promoted clearing to occur as time passed and/or uncleared balances became undesirably large.

The network of clearing between goldsmith-bankers was critical to the spreading use of bank notes and checks as means of payment. Goldsmith-banking created a positive externality in that each banker's notes circulated more widely than without mutual acceptance and clearing. The network allowed the public to use bank debt to facilitate a wider circle of transactions. The next chapter considers another avenue, taxes, that goldsmith-bankers pursued in expanding the range of bank note acceptance. As with clearing, the system by which goldsmith-bankers arranged the public's ability to pay taxes with bank debt developed out of bankers' self-interest. Unlike debt clearing, opening the tax system to notes and checks required goldsmith-banker interaction with the government, specifically the medieval labyrinth of procedures known as the Exchequer.

110 Chapter 3

How a New Fashioned Goldsmith Avoided the Ancient Course of the Exchequer:

A Story of Private Bank Notes Paying Public Taxes

Both the transferability of goldsmith-banker debt and the clearing of that debt between bankers improved the success of notes and checks as substitutes for specie in the settlement of transactions.(Chapter 1, Chapter 2) Taxes, however, were obligations for the merchants of late seventeenth century London that might appear beyond the reach of goldsmith-banker debt. The

Tellers of the Exchequer, the cashiers of the English government, accepted only specie. Tax collectors, however, could accept notes and checks written on goldsmith-bankers from the public, but those same tax collectors would then have to redeem the bank debt for coins before presenting receipts to the Exchequer. Because the collectors of the major English taxes like the Customs or the

Excise generated monthly revenues of tens of thousands of pounds, holding such sums of goldsmith- banker debt was a considerable risk, even if the chance of default was small. Not using goldsmith- bankers, however, left tax officials with the costs of protecting, assaying and transporting all that specie. Tax collectors needed banking services without the risks of holding potentially unbacked bank debt.

By acquiring government debt, goldsmith-bankers encouraged tax receivers to trust bank

debt. Goldsmith-bankers regularly lent to the Treasury and received debts called tallies.' Most tallies were advances with a designated revenue stream committed to repaying the loan. When the

1 Chapter 4, for example, describes how the goldsmith-banker Sir Francis Child began investing heavily in tallies after the Glorious Revolution in 1688.

Ill tally fell due, the tax receiver of the designated branch of the public revenue paid for and retired the tally. This chapter contributes the insight that how the tax collector paid the goldsmith-banker for the value of a tally was a private matter between the two parties which bypassed the cash-only

Tellers of the Exchequer. A tax collector and a goldsmith-banker could swap bank debt for tallies.

When a goldsmith-banker held tallies payable by a tax collector's branch of the revenue, the risk of accepting bank notes and checks were greatly reduced because the tax collector knew the goldsmith had reliable assets (from the tax collector's perspective) to cover bank debt.

This institutional arrangement between tax officials and government debt holders allowed goldsmith-bankers' investments in tallies to be both interest earning assets and a means of expanding the circulation of bank debt into the important sphere of taxes. For a goldsmith-banker, the ability of his debt to pay taxes expanded the circulation and increased the value of all his debt.

Moreover, the tax collector was able to spare the transactions costs of dealing with only specie by accepting bank debt and taking advantage of other bank services. For the Treasury, cooperation between lenders and tax collectors encouraged both the advancing of funds into the Exchequer and facilitated the payment of taxes by the public. Throughout the second half of the seventeenth century, the coupling of goldsmith-bankers and revenue officers was a mutually beneficial conjunction of England's public revenue and London's private bankers. Revenue cashiers could bless a banker's note with tax paying power. Tally holding bankers could guarantee the tax collector that those notes would be exchanged for tallies.

For example, this chapter considers the relationship between the goldsmith-banker Edward

Backwell, whose inter-banker clearing accounts were examined in Chapter 2, and Richard

Mounteney, the Receiver and Cashier of the Customs revenue. The banker and the tax collector

each had something to offer the other. Richard Mounteney, Cashier of the Customs for the Port of

London, accepted Backwell's bank notes, checks and bills of exchange from the public for the

112 payment of taxes, and taxes created a regular, at par use for Backwell's demandable debt.2 The

increased value applied to all the goldsmith-banker's circulating demandable debt, not just that

actually used for tax payments. Knowing that any Backwell note could be used to cover the

Customs duties solidified public acceptance of those notes and expanded their circulation. At par

acceptance of notes for tax purposes also sent a signal of credibility from a branch of the King's

revenue. The very definition of money was a royal prerogative. By extension, government

acceptance of non-interest bearing private debt legitimated Backwell's notes. The Customs Receiver

and not the Exchequer accepted Backwell's paper. For practical purposes, however, few people

ever dealt directly with the Exchequer. The tax collector was the gate keeper that mattered. If

Richard Mounteney accepted Edward Backwell's note, the goldsmith-banker's debt was as good as

silver.

In fact, goldsmith-banker's debt was better than silver. Backwell's customers, Richard

Mounteney and all the others, saved the transaction costs of collecting, assaying and protecting

specie.3 As Cashier of the Customs, Mounteney processed tens of thousands of pounds per

month.(Chart 3.1) The entire English gold and silver monetary supply hovered somewhere around

twenty million pounds, so a single large month's customs revenue could account for a one or two

percent of the entire English monetary stock.4(Horsefield 1960, p. 258-9) Instead of moving coin,

merchants could minimize transactions costs by writing an order to Mounteney for payment of

customs duties. A merchant's remaining balances could then be used for other purposes. Richard

2 The author reasons Mounteney accepted Backwell's notes at par because the cashier kept a busy account with Backwell in which third parties regularly credited Mounteney's account. Backwell would have had to accept these notes on himself at par. Likewise, if Mounteney discounted Backwell's notes, third parties could cash the notes first with Backwell and then bring the money to Mounteney.

3 Over the year March 25, 1670 through March 24, 1671, Backwell paid Mr John Brattle £32:16:3 for assaying and Mr Keemar five shillings a week for manning the watch.

4 The figure was calculated by Sir Isaac Newton for 1688.

113 CO o Pound Sterling S (Thousands) CD i i i i i o oo o> 4^. ro O 00 0) 4* rb 7T CD O O O O O O CD /ell' s lllillllllBI I i i i o o i i i < 1 1 A p r Ma y Led g r o ^ (D = j (c0 7n C- = i—h w c | f-r 3 5 o | 3

o h 16 7 ^

NAu g : ;q f IS a. co S Of ^1 S) o = 1 -1 00

3" ZIMIIIIIIIIII I l o v D 0) 167 1 ' ""* L 0) 3 O (D i O | CD

Dc_) =i 3I •nl (D = S°"lI 51

114 Mounteney, in turn, handed the paper over to the goldsmith-banker for the tallies the tax collector

needed to settle his own accounts with the Treasury. The tax officer needed paid tallies as proof to the Treasury of revenues collected. In contrast to everyone else, tallies were a credit for revenue

cashiers rather than a loan. Richard Mounteney owed the Treasury and could settle with tallies.

Backwell already had the government debt, and the goldsmith-banker had to accept his own notes at

par, so Mounteney could accept Backwell's notes from the public with little risk of not getting the

tallies.

The tally for bank note swap that both Richard Mounteney and Edward Backwell enjoyed

formed the basis of a wide ranging financial relationship. Backwell acted as Mounteney's banker

and handled large portions of the Customs revenue. Over the year 1670, Mounteney's daily balance

of £16,730 was worth one thousand pounds per annum to Backwell at six percent simple interest.5

The connection between the rise of bank notes, the collection of the public revenue and the

funding of Treasury debt can be seen as early as the Restoration era relationship between Edward

Backwell and Richard Mounteney. Subsequent generations of goldsmith-bankers and tax collectors

continued the partnership. The team of Charles Duncombe and Richard Kent was the zenith of the

late seventeenth century alliances between goldsmith-bankers and revenue officers.6 Charles

Duncombe was following in the footsteps of his mentor, Edward Backwell, under whom Duncombe

was apprenticed until 1671. What Duncombe learned first hand as an apprentice, and what this

paper develops from Backwell's existing records, was Edward Backwell's purchasing of tallies

secured on the Customs coupled with the goldsmith-banker's extensive connections with the then

5 Six percent was the limit under the usury laws of Restoration England. Loans to the Exchequer, however, were exempt and could range up to ten percent in this period. Given Backwell's known lending to the Crown, a six percent return is a conservative assumption.

6 "Charles Duncombe and Richard Kent, for example, advanced the king between £150,000 and £200,000 a year in their capacities as cashiers of the Customs and Excises."(Brewer 1989, p. 94)

115 Receiver and Cashier of the Customs, Richard Mounteney. Backwell and Mounteney were an early example of the late seventeenth century interplay between goldsmith-bankers and tax collectors that enabled private bank notes to settle tax obligations.

Because the citizenry had to pay taxes, the relationship between banker and tax collector captured a type of seigniorage. If a tax collector limited whose debt he accepted, then the honored debt gained added value as a privileged means of meeting tax obligations. A goldsmith like Edward

Backwell enjoyed both the direct and indirect returns from lending to the Treasury. Besides earning interest, holding tallies allowed Backwell to secure a portion of the state's capacity to collect seigniorage. Edward Backwell and Richard Mounteney merged both tallies and taxes with private bank notes to facilitate tax collection and increase the demand for bank debt. Such partnerships were a critical early step in the development of both tax collection and government debt issue and foreshadow the relationship that would develop between the Bank of England and the Treasury.

I. Advances on the Customs

Numerous goldsmith-bankers like Edward Backwell specialized in investing in government debt as a form of delegated lending. By pooling deposits, goldsmith-bankers could maximize bargaining power with the Treasury and spread the fixed transactions and information costs of dealing with the Exchequer over larger amounts.(Roseveare 1962, Chapter 4) Moreover, bankers could offer depositors liquidity. Merchants were averse to tying up large amounts of their funds in direct purchases of taIlies.(Jones 1972) Backwell offered less interest (or none on demand accounts) than the Exchequer, but depositors could invest smaller sums with easier terms of withdrawal in a far more accessible location.(Roseveare 1962) The bankers pocketed the difference between the lower interest rates offered depositors (limited to six percent by usury laws) and the higher Treasury

116 rates of return, for the King exempted himself from the usury laws and offered rates ranging up to ten percent.

In 1672, London goldsmith-banker Isaac Meynell, described his role as an intermediary

between depositors and the Treasury as:

The money [was] paid in at [Meynell's Shop] and purporting a promise or obligation for payment of the Sume with damages at six pounds p Cent upon demand or within few days after as the usual! course was, yet it was well knowne to all persons that soe left money with [Meynell]... that the money were soe left with [Meynell] with reference to the said course of Lending to the King and payment in the Exchequer and [Depositors] did know that [Meynell] could not take in these great sums of money which they did and agreed to pay such interest for the same on demand as [Meynell] usually did (How be it the note or bond did mention it should be paid at some certain day or within a few days after notice) upon any other account then relative to the said loane to his Majestic.

And as Moneyes were left with [Meynell] as aforesaid [he] did dispose of them though not by the expresse direction of the said [Depositor] and of every other individual person yet by the general! allowance and approbation of them all.(Chancery Court, C9 110/27)

In 1670, Edward Backwell lent deposited money to a number of revenue branches, but, as

Table 3.1 shows, Backwell's lending to the Customs revenue dwarfed his other lending. Backwell was a member of the Customs Farm, a syndicate that advanced money to the Treasury and received dividends for the Customs collected beyond a contracted amount. However, Richard Mounteney, the Receiver of the Customs, actually orchestrated the collection of the Customs duties. Far more important than being a member of the farm, Backwell was the banker for the Customs farm. In the name of the Customs Farm, Backwell loaned tremendous amounts of money. The farm was only an accounting label for Backwell by 1670. Chart 3.1 sketches the Customs Farm's daily balance with

Backwell, with negative balances ranging over £160,000 in that year.

117 Table 3.1

Backwell's Purchases of Debt on Different Revenue Branches

From March 1670 to March 1671

Revenue Branch Approximate Amount Loaned

Customs Farm £228,000 Excise Farm £27,000 Exchequer, Sir George Downing £50,000 Loans on Wines Act £40,000 Crown Lands £20,000

Source: Backwell's Ledgers

The Customs Farm account was debited when Backwell advanced money on the Customs and credited when tallies were repaid. As explained in the next section, Backwell's advances on designated Customs' revenues were settled by Richard Mounteney. Interest, however, was paid by the Exchequer, so the Customs Farm account in Backwell's ledgers makes no mention of interest.

Mounteney settled only principal with Backwell. Payments to Backwell, however, came to an end just one year after the transactions sampled in this paper.

In September of 1671, Backwell's farming syndicate lost control of the Customs revenue and a commission was installed to oversee the revenue's collection. Backwell and other bankers curtailed their advances but did buy up existing orders (debts not secured on a particular period's revenue).

With financial pressure building, the English Treasury effectively suspended most payments out of the Exchequer on December 18, 1671.(Roseveare 1962) The Stop of the Exchequer was made official on January 5, 1672. Under the Stop of the Exchequer, Backwell held forty-seven tallies of

Pro on the Customs worth nearly £30,000, eighty-nine orders on the 1671-1673 Customs worth over

£125,000 and five orders on the Customs in general worth just over £26,000.(Roseveare 1962,

Appendix V, p. ii)

118 Although Backwell's debts were never repaid, the system of advancing money was soon reestablished. A new generation of bankers, including Backwell's own apprentice Charles

Duncombe, soon were working with tax officers. In 1674, Richard Kent became the Receiver-

General and Cashier of the Excise revenue by Letters Patent bearing the Great Seal of England.

Kent's appointment was arranged by his mentor Sir Stephen Fox, Paymaster of the King's Forces.

Fox had committed to lend up to £210,000 per annum secured on Excise revenues and wanted his man, Kent, to take leadership of the Excise (a tax on domestic products, especially beer).(Clay, p.

98) Kent, in turn, had formed a banking partnership with Charles Duncombe, a goldsmith-banker of

Lombard Street. Kent and Duncombe would expand their operations by also becoming Receivers

General of the Customs in 1677. "By 1680 they were advancing the equivalent of 40 per cent of the government's revenue."(Roseveare 1991, p. 25) The Stop of the Exchequer may have broken a generation of lenders to the Crown, but the advantages of pairing bankers with revenue agents were not dissipated because the system of advancing money to the Treasury was retained. After the Stop, tax receivers were still directly retiring tallies and avoiding the Tellers of the Exchequer.

II. Avoiding The Ancient Course of the Exchequer

In compiling his ground breaking work, The English Public Revenue 1660-1688, CD.

Chandaman corrected a major error in calculating the English public revenue. Many transactions recorded by the Exchequer were not cash issuances. Rather, tallies (government debt) were being issued in anticipation of future revenue. For Chandaman's purposes, understanding how the

Exchequer accounted for tallies was essential for correctly calculating the actual English public revenue. For our purposes, the same Exchequer procedure is significant because it alters our understanding of the relationship between tax receivers and the Exchequer and does so on a large scale.

119 If the use of the Sol tally for payments had been only an occasional expedient, the effects of this distortion might not be too serious, but it is perfectly clear, both from the abundance of matching entries in the accounts and from the magnitude of the government debt by anticipations recorded at various times, that this tally was employed on a very large scale during the greater part of the Restoration period.(Chandaman 1975, p. 294)

The end result of the Exchequer's particular system of processing anticipations (loans secured by future revenue streams) was that tax receivers could settle their obligations to the

Treasury with tallies rather than cash. The tax receivers acquired the tallies because the lenders, who advanced their money to the Exchequer in exchange for tallies, collected their repayment directly from the tax receivers. The cash collected from tax payers did not go through the

Exchequer to get to the tally holding lender. Because the tax receipts did not pass through the

Exchequer, the receiver and the lender could settle payment for the tallies in any manner they found expeditious.

Tax receivers could settle directly with tally holders and avoid the Tellers of the Exchequer because the major revenue streams were not audited by the "ancient" course of the Exchequer.

Rather, tax collectors were answerable to the Auditors of the Imprest in the Upper

Exchequer.7(Aylmer 1974) The accounting of the Customs revenue by the Auditors of the Imprests

(often shortened to "Prests") was part of a larger trend. By the 1630s, two thirds of all incoming revenue audited by the Exchequer were under the view of the two Auditors of the Prests.(AyImer

7 The nature of the Auditors of the Prests as a simpler, more direct system of accounting than the "ancient course" stems from wartime origins. Henry VII withdrew various revenue accounts from the Exchequer as early as 1513. "It is clear that the summary methods introduced by Henry VII and continued by Henry VIII left no room for such a [the normal Exchequer] process."(George 1916, p. 48) These extra-Exchequer auditors were eventually merged into the Court of Augmentations and Revenues with the explicit naming of "the Audytors of the Prestes".(George 1916, p. 49) The business of the Auditors of the Prests was moved into the Exchequer with the dissolution of the Court of Augmentations and Revenues by Queen Mary. The explicit offices, however, were first created six years later by Queen Elizabeth in January of 1559/60.(George 1916)

120 1974, p. 35) As the excise and hearth taxes were added during the Interregnum, these two auditors dominated the accounting of all the great revenues.

The accounts taken by the Auditors of the Prests, moreover, were gaining in importance with the expansion of revenue and expenditure, while those taken by the Auditors of the Exchequer [the ancient course] were both absolutely and relatively declining. After the Restoration these latter ... found themselves mainly controlling accounts which were becoming formal or obsolete.(George 1916 p. 54)

The literature presents two compatible reasons why the Auditor of the Prests were in ascent. The

Auditors of the Prests charged lower fees and benefited from clearer accounting than the Auditor of the Receipts.(AyImer 1974, p. 209; George 1916) The Auditors of the Imprest were also under direct supervision of the Treasury, which improved oversight.

The actual procedures of the Exchequer allowed the critical detachment of the Tellers of the

Exchequer from much of the actual money collected from taxation. The tellers were the cashiers of the Exchequer. They collected, held and dispensed all specie in or out of the Exchequer. When advancing funds on a revenue, the lenders like Backwell would present the money to the Tellers of the Exchequer and receive a tally. Between the depositing of funds with the Tellers and the cutting of the tally, both the Upper Exchequer and the Treasury were notified. Chart 3.2 shows that the first response to money entering the Tellers was for the Auditor of the Receipt to send a prest certificate to the Auditors of the Imprest (Steps 1 and 2). The Auditors of the Imprest oversaw the great revenue accounts and answered directly to the Treasury. This initial diversion from the direct production of a tally ensured the Treasury's awareness of the inflow of advances.

The Treasury Lords would then issue a warrant to the Auditor of the Receipt for the eventual creation of a tally. The Lower Exchequer, however, did not immediately convert the

Treasury's warrant into a tally. Most tallies were a receipt called tallies of 'Sol'. To have taken the farmer's money and given him a receipt would have left the Lower Exchequer retaining a receipt by the amount of the farmer's advance. To generate an offsetting issuance, the Lower Exchequer

121 i o I CO CO o c <- (D 3 30 O O —h 0) 30 (D O (D g "SO £ 3D cr O 0) 3 o § 3 2. O CD O CO 3

CD -Q 3D c a a Qm. O cr zr -\ Z H <3 OJ 0) o" g S a—« X J3 o_ II Ol 3D 00 0) H (D —i O O c

122 would have had to collect the money from the tax receivers and then issue the money to the tally holding farmer.

To avoid the tax receiver passing his collections through the Exchequer, a circuitous route was taken. The Treasury warrant directed the Auditor of the Receipt to issue an order to the farmer

(Step 3). The Clerk of the Pells recorded this order as an issuing of funds. The farmer then presented the order to a Teller (Step 4). The Teller treated the order as incoming funds which started the process of creating a tally.8 Before giving the tally to the tax farmer, the same Clerk of the Pells who counted the earlier order as an issuance now counted the tally as a receipt. As a result, the Lower Exchequer had balancing Issue and Receipt Rolls.

With tally in hand, the remaining players in the processing of the public revenue (the farmer, the tax receiver, the Auditors of the Imprest and the Treasury) were free of the medieval procedures of the Lower Exchequer. The farmer held the tally and drew interest until the revenue stream (specified on the tally) was collected by the tax receiver. The tax receiver and the farmer then swapped the taxes for the tally (Step 5). Finally, the tax receiver held the tally for his accounting with the Auditors of the Imprest and eventual declaration of accounts to the Treasury

(Step 6).

Once liberated from the Lower Exchequer, the farmer and receiver could settle accounts by banker debt. Section III uses the example of Backwell and Mounteney to reveal how their freedom of action greatly expanded the use of private notes and bills. By having to settle with Backwell instead of the Tellers of the Exchequer, Richard Mounteney (Receiver of the Customs) could pay

8 Upon receipt of the order from the payee, the Teller sent a bill down the "Pipe" to the Tally Court for the striking of a tally. With the Teller's bill, the Tally Cutter notched the four foot wooden sticks with the proper amount. The Auditor of the Receipt then added the information detailing the revenue stream and the marking "Sol" for receipt. The Clerk of the Pells then copied the tally's information into the Receipt Book before the Deputy Chamberlains split the tally and checked the tally to the book entries. The Exchequer kept the "foil" half while the original payer left with the "stock" half. (Chandaman 1975, Appendix 1)

123 off Backwell's tallies with Backwell's own notes. Using paper reduced transactions costs in collecting taxes which benefited Mounteney. For Backwell, the public's ability to pay taxes with his notes increased their value. The magnitude of Backwell's and Mounteney's cooperation is detailed below.

III. The Banker and the Tax Collector

During the last years of the 1660s, Edward Backwell, a prominent Restoration era goldsmith-banker, acted as banker for both the Customs Farmers and the Receiver of the Customs revenue. Moreover, Edward Backwell was a member of the Customs Farm from September 1667 to

September 1671.(CTB VIII, p ii, p. 1414) Richard Mounteney was Receiver of the Customs,

Cashier to the Customs Farm and Comptroller of the Customs for the Port of London over the same period.(CTB) On behalf of the Customs farm, Backwell advanced money to the Exchequer in return for tallies. In time, Backwell was repaid for his tallies by Richard Mounteney out of the proceeds of the Customs revenue.

The exchange of tallies for Customs revenue formed a basis for a mutually advantageous arrangement between Backwell and Mounteney. Mounteney accepted Backwell's notes and bills in payment of customs duties. The use of bills and notes offered Mounteney savings of transactions cost relative to weighing, transporting, protecting and assaying specie. Except for very rare occasions, the Exchequer only accepted money. If the arrangements in Section I, above, were not the case, Mounteney would have had to present specie to the Exchequer.

Mounteney's acceptance of Backwell's notes for tax purposes also benefited the goldsmith- banker. The public's ability to pay taxes with his notes would have increased their valuation and circulation. Taxes were unlike other possible uses of a goldsmith-banker's note. The requirement to pay taxes was backed by the coercive power of the English crown. The threat of force coupled

124 with the ubiquitous nature of taxes, at least among the merchant classes, guaranteed that holders had

an outlet for goldsmith-banker notes. Finally, at par acceptance of notes by tax officials sent a

message of credibility. Through Mounteney, the state signalled to the public that Backwell's notes

were as good as money for retiring tax obligations.

The importance of Backwell's and Mounteney's relationship lies not just in the

private/public connection but also in the magnitude of the operation. Over the twelve months from

March 1670 to March 1671, a total of just over £700,000 was credited to Richard Mounteney's

accounts with Backwell. Even as a flow measure, the size of Mounteney's account with Backwell

was staggering by the scale of the day. Over the same period, the customs revenue was only

£242,704 and the entire public revenue was £1,100,929. (Chandaman 1975, pp. 352-3) Backwell's

relationship with Richard Mounteney, the Cashier of the Customs, involved credit on a very large

scale.

Mounteney had two demand accounts with Backwell. One was labeled "Notes," the other

"Bills." Neither recorded any interest payments on balances. In the seventeenth century, the term

"note" often referred to both promises to pay (bank notes) and orders to pay (checks). "Bills,"

however, was a term usually reserved for bills of exchange which were always orders to pay. The distinction between promises and orders mattered in determining how bank credit originated. An order required an offsetting balance within Backwell's ledgers. For example, on October 31, 1670,

Richard Mounteney's Bill Account was credited for £343:16 per John Lindsay. On the same day,

John Lindsay's account with Backwell was debited the same amount per Richard Mounteney. A

bank note was a promise by Edward Backwell to pay the note's bearer. A bank note credit by

Backwell's note in Mounteney's "Note Account" would not require an offsetting debit in another

account because Backwell himself was charged on the note. Many entries in Mounteney's note

account lack simultaneous debits and indicate Backwell's bank notes were commonly used. Caution

125 must be taken in separating promises and orders because both were regularly passed through

Mounteney's Note Account.9

Chart 3.3 plots the combined balance of Mounteney's two accounts (Note and Bill) over the accounting year March 25, 1670 to March 25, 1671. Especially from April through August, credits built until a mid-month clearing occurred. The intra-month timing of transactions corresponds to the fact that monthly tax revenues were due on the fifteenth of each month. The monthly clearing of

Mounteney's account is largely the inverse of Mounteney's paying off tallies on the Customs revenue held by Backwell in his capacity as banker for the Customs Farm. Chart 3.1 shows the daily balance of the Farmers of the Customs with Backwell. Debits on the Customs Farm account occurred when Backwell advanced money to the Exchequer in the Farm's name and received a tally

(see Sections I and II above). These entries specify by which month's Custom revenue each tally would be paid and the fees paid to the Lower Exchequer for creating the tallies. On the other side, the Farmers of the Customs are credited for paid tallies. For most months, Richard Mounteney is named explicitly as paying the Customs Farmers for tallies.

The author submits that Richard Mounteney received tallies in exchange for credits he accumulated on Backwell. The debit side of Mounteney's Note Account is silent regarding the basis under which debits were created. The simultaneity of the farmers' credits and Mounteney's debits, however, leaves no doubt about the symbiotic nature of Backwell's role as banker to both the

Customs Farm and the Receiver of the Customs. Chart 3.4 combines Mounteney's daily balances

(in gray) with the Customs Farmer's debits (in solid black). The mid-month reduction of

Mounteney's account corresponds on a same day basis with the crediting of the Customs Farm by

Richard Mounteney in payment of tallies due on that month's Customs revenue.

9 For example, on November 16, 1670 Richard Mounteney's Note Account with Backwell was credited for £6,900 by the East India Company. On the same day, the East India Company's account was debited £6,900 in the favor of Richard Mounteney.

126 CO O Pound Sterling c 3 (Thousands)

3 O* 3" 0) —t is w" GO

CD 9L 0) 3 O CD

127 CO o Pound Sterling c (Thousands) O 3 c (D (0 & 3 CO 0) —t 3 O CD S" O -3 —^

CD ^ 3 CD CO* DO 2. 3 o CD

128 Chart 3.5 illustrates the clearing procedure between Mounteney and the Customs Farm within Backwell's bank by means of related "T" accounts. On July 28, 1670, Richard Mounteney was debited by £38,960 with no mention as to where the funds were going. On the same day,

Backwell credited the Customs Farmers' account per £23,733 from Richard Mounteney. The credit was explicitly in repayment of seven tallies on the Customs the Backwell had advanced on behalf of the Customs Farm between August 1669 and July 1670.'"

Mounteney's July 28 debits were larger than the tallies held by Backwell because the debits were settled by specific credits in Mounteney's Note Account. The debits in the Note Account on

July 28 explicitly offset two earlier credits. On July 14, Backwell transferred the £10,358:14:7.5 balance of the Bill Account to the Note Account. This transfer covered forty-two entries that had arrived as credits in the Bill Account since June 20, 1670. The remainder of the July 28 debit balanced a £23,601:17:3 credit "by Note" on July 15, 1670. Most months had both a transfer from

Mounteney's Bill Account and a large "by Note" credit around mid-month. Table 3.2 presents both types of credits to the Note Account. The left side reports each month's transfer of the Bill Account

Balance. The right side reports a large note that was cleared (debited) on the same day that the

Custom's Farm was credited.

10 Of the creation of the seven tallies was debited to the Farmers as £5,000 to Pepys on August 21, 1669; £11,614 to the Treasurers of the Navy on October 30, 1669; £2,500 to the Earl of Bath on February 8, 1670. The additional, unnamed tallies were for £3,000, £467, £402:13:4 and £750 on June 22, July 14, July 15 and July 22, 1670 respectively.(BackweH's Ledgers)

129 > a

CD O CD a. o 0) 3 3 O Q. CD CD 0) o 3 ^ o o c o" 3 CD - - Ig 5 CO 10 3" w o1 0) O 3 8 # 8 a —i 3 Q. —+> -n H 3 a> O O c CD' 3 3 CD CD C 3 O CD CD 0) ro CD 00 o 3 3. (0 r-t- CO o> 3" 3" CD CD' bi o —* O ^oi CcO 3" o CD 3 o CO Tl CcO 0) I-* —t o 3 1IS*I 3 CD CoD o—o (/> % > 8

130 Table 3.2

Credits into Mounteney's Note Account to be Cleared with Backwell and Customs Farm

Transfers from Bill Account Transfers by "Note"

Date Value Date Value 13 April, 1670 £5,556 21 April, 1670 £27,448 17 May £11,912 17 June £11,677 17 June £27,590 14 July £10,359 15 July £23,602 18 August £9,845 11 August £23,714 23 September £11,618 19 September £47,901 21 October £6,866 24 November £13,988 7 December £8,348 27 December £8,560 16 January, 1671 £6,133 14 January, 1671 £4,408 13 February £3,763 13 February £15,311

Total £86,077 Total £165,074

Source: Backwell's Ledgers

Except for May 1670, each month's payments to the Customs Farm were from the credits transferred from the Bill Account or gained by the note entry also reported in Table 3.2. The balance of the Bill Account was an accumulation of orders (likely bills of exchange) written or endorsed to Mounteney's credit, payable by Backwell. The origins of the note entries are a mystery. Their use in retiring the Customs tallies held by Backwell suggest they were part of the revenue collection process. That the note entries in Table 3.2 are Backwell's notes that Mounteney has collected and in turn presented to the banker is the simplest story. However, the author cannot claim certainty. The tallies, however, were paid by these two particular credits and not out of

Mounteney's Note Account in general.

May 1670 was an exception. May's tallies were paid from the credit of a number of small entries with a wide variety of counterparties. The note entries in Table 3.2 never listed any counter

131 party. The May outlier suggests that the use of the two entries (the Bill Account transfer and the single "by Note" credit) was an arrangement of convenience and not necessary for the retirement of a month's tallies.

Where the funds went that remained after paying the tallies is equally unresolved. For example, Richard Mounteney could have taken the £10,226 credited to "Other" in Chart 3.5 in the form of money, a note, a bill of exchange, credit on another account in Backwell's ledger or any combination thereof. The accounting within the ledger was silent on the form credits took. Recall, we only know about the payment of tallies from the matching Customs Farm accounts because

Mounteney was explicitly named. We do know, however, that Richard Mounteney did not keep the surplus (Bill Account + Note Entry - Tallies > 0) as a credit on his own Note Account. He moved the funds somewhere, implying that they were not Mounteney's to keep with his banker, Edward

Backwell.

The monthly flow of funds through Mounteney's Note Account encompassed more business than the retirement of Customs Tallies presented above. Mounteney carried on substantial business through the goldsmith-banker after swapping the mid-month credits recorded in Table 3.1 in exchange for the Customs tallies held by Backwell. Chart 3.6 pairs daily credits to the Custom

Farmers with Mounteney's daily debits. Observations off of either axis means both accounts had transactions on the same day. While not every Customs Farm credit is matched with a Mounteney debit, the great majority are. A large percent of Mounteney's debits, however, are not accounted for by payments to the Customs Farm. A considerable portion of Mounteney's account, then, is additional business built on the tally retirement relationship.

Mounteney needed a banker. As the Receiver and Cashier of the Customs, Richard

Mounteney commanded a large flow of funds. For example, numerous deposits of money over the year March 1670 to March 1671 totaled £79,552:11:10 indicates that Mounteney used Backwell to

132 £ Credits on the Customs Farm CO o c (Thousands)

CO 30) CD o £ o

o ar t zr 3 redi t O C Z3 co oo (/) 0) • 3 O) Q. O CD U" (O

133 process other aspects of the Customs revenue beyond the payment of taxes." Mounteney also used his Note Account with Backwell to facilitate a wide range of activities beyond the Customs.

For example, Backwell purchased £50,200 worth of tallies on the Wine Act on behalf of

Mounteney.12 As part of his position as a revenue officer, Mounteney was likely expected to lend to the Exchequer.

A third Richard Mounteney account in Backwell's ledger hints at the breadth of Mounteney and Backwell's financial alliance. In the ledger covering March 1670 to March 1671, an account simply titled "Richard Mounteney, Esqr.," records a group of loans made by Richard Mounteney to

Edward Backwell totaling £13,225:16 at six percent per annum.13 In total, the banker paid

Mounteney £807:-:4 in interest for these loans over the course of the accounting year March 1670 to

March 1671. The next highest sum that Backwell paid anyone in interest over the same period was

Sir Edmund Turner at £472, so Mounteney's loan to Backwell was considerably larger than even the largest of Backwell's other deposit customers. The author cannot say if these funds were

Mounteney's own, a diversion of tax revenues or part of some quid pro quo with Backwell.

Backwell does not list any interest payments to Mounteney in the surrounding years, so the loans

11 The "Money p note" entries were March 31, 1670, for £4,000; May 19, £7,000; June 11, £10,000; June 17, £6,100; June 24, £5,700; July 14, £9,000; July 30, £10,000; August 19, £10,000; September 17, £6,000; December 19, £3,000; February 28, 1671, £2,000; March 21, £4,265:7:10; March 22, £1353:19:-; and March 23, £1,133:15:-. (Backwell's Ledgers)

12 Mounteney's Note Account was debited for payments, "Into the Exchequer on the Wine Act," on November 17, December 1, 10, 20 and 30, 1670, and January 7, 1671, for £2,000 respectively; January 10, 1671, for £2,200; February 8 and 10 for £8,000 each; February 23 for £5,000 and February 27 for £8,000. (Backwell's Ledgers)

13 The principal was broken into five separate loans for £5,000; £2,500; £2,500; £2,000 and £1,225:16. All had half year interest payments that fell due on June 24, 1670 and Christmas Day, 1670. Although the principals for both £2,000 and the £1,225:16 loans were paid by Backwell on December 25, 1670, new three month loans for the same amounts were issued on the same day. The change in term structure would have spread interest payments out more evenly over the year. An additional loan for £450:4 was made on June 24 and retired on December 25, 1670 at six percent per annum returning £13:10:- interest. (Backwell's Ledgers)

134 appear to be a one time occurrence. The loans indicate personal funds of large amounts were part of the relationship between the tax collector and the banker.

A large financial partnership between Mounteney and Backwell was built on the foundation of Backwell's role as a bridge of credit between Richard Mounteney and the debt holding Customs

Farm. On a monthly basis, bill and notes were presented by Mounteney to Backwell, tallies were retired and the surplus was moved on. In addition, Richard Mounteney used his Note Account with

Backwell to process cash tax receipts, loan funds to the Treasury and engage in a variety of other activities. Backwell's long run debt to Mounteney added a further layer to the symbiotic relationship between the banker and the tax collector.

IV. Banker-Debt Holders as an Optimal Arrangement

The ability of Backwell's paper debt to replace specie as a means of retiring tallies was not necessarily contingent on the banker's status as a member of the Customs Farm or as a holder of tallies. However, the coupling of banking and debt holding internalized a potential bottleneck. The tax collector accepted a means of payment from the public before he settled his own accounts with either the tally holders or the Exchequer, so a revenue cashier needed to have a credible expectation of exchanging notes for specie or tallies. By holding the tallies himself, a banker like Backwell could assure the revenue agent that notes accepted for tax settlements would be exchangeable for tallies.

The importance of Backwell's dual role as banker and tally holder can be illustrated by developing the alternative. For example, the tax collector could accept Backwell's notes for payment of taxes and then present the banker with the notes demanding money which, in turn, the tax collector could pass on to the holder of government debt. Such a scheme required a customs cashier like Richard Mounteney to handle cash and hinged on the revenue officer's expectation that

135 a banker like Backwell would honor the notes with specie. Given the size of revenue transactions, the amounts of money would have been potentially tens of thousands of pounds. Setting aside the transaction costs of dealing with such large sums, the size of the transactions alone meant the result of even the most unlikely default would have been crippling.

To reduce the movement of specie between a tax collector and a note issuing banker, a cashier like Richard Mounteney could have presented Backwell's notes directly to the tally holders.

To arrange use of his notes for tax payments, Backwell need only have been in a position whereby the tax collector expected the holder of Treasury debt would accept Backwell's paper debt for retirement of that tally. The tally/note exchange, however, only transferred the risk of default to the debt holder (assuming par acceptance throughout), but does not inherently address the underlying risk. Chart 3.7 stylizes the relationship between the players in the revenue system in which both the tax collector passes bank notes onto the debt holder and the note issuing banker does not hold

Treasury debt. The tally is created by the advancement of money to government shown as Step 1.

The money could go through the Exchequer as described in Section I or directly to government purchasing agents such as Samuel Pepys or Sir Stephen Fox. The lender would receive a tally in receipt for the funds advanced. Separately in Step 2, the public would deposit money with the banker for bank notes or demand accounts. Goldsmith-banker notes (promises or orders) offered the public a low transactions cost, near money medium of exchange.

The public and the tally holder begin to connect when the public pays their taxes with the banker note in Step 3. The tax collector accepts the banker's note with the expectation that the debt holder will swap the banker's note for retirement of the tally. The banker's note is not legal tender, so the tally holder has no initial legal obligation to accept the paper debt in Step 4. After the tally- note exchange, the current note holder (former tally holder) could present the banker with the note for payment in cash. Step 5 returns the note to the banker and the initial principal to the lender.

136 (oD "?< -H PO "D § £ O 3 3o" H W 0) s ,-+ LU 2: 3 o 0) I I ?r 3 CD 77 o o CD Q. Q. 1 —% CD CD 3 o —% -i 3

137 The schematic is simplified because the lender could have passed the goldsmith-banker note on to other parties or, after presentation to the banker, could have accepted forms of payment other than money.

Again, the tax collector was under no legal obligation to accept the goldsmith-banker's note from the public, so the tax collector had to believe that the tally holder would take the goldsmith- banker's note in retirement of the tally (Step 4). The benefit of heightened credibility and expanded circulation that the goldsmith-banker's notes enjoyed when the tax collector accepted them hinged on the expectation that the goldsmith-banker could deliver the cash. Every involved party benefited from avoiding the movement of specie, but the expectation that the goldsmith-banker's debt could be cashed cast a shadow of risk over the process.

Backwell minimized the uncertainty of final note payment by becoming the original debt holder. By holding the tallies himself, Backwell guaranteed Richard Mounteney that the tallies for note exchange would succeed. The tax collector knew that Backwell, unlike third parties, had to accept his own notes or default. The likelihood of default was low because Mounteney knew

Backwell held a form of assets, tallies, that the Customs Cashier required. Backwell needed to hold the tallies himself to create this certainty. In fact, the Customs Farm did owe a large deficit to

Backwell over the year from March 1670 to March 1671 as pictured earlier in Chart 3.1 indicating the goldsmith-banker retained the tallies until paid off by Mounteney. Backwell was a banker loaning large sums to the Treasury in the name of the Customs Farm, but Backwell held the debt and collected the interest.

Chart 3.8 shows a second schematic in which the banker and the debt holder from Chart 3.7 above are merged. The public's deposits with the banker and the banker's advances to the government are retained from the earlier story. When the goldsmith-banker's note passes from the public to the tax collector, a cashier like Richard Mounteney knows he can present Backwell with

138 3J o 3- H 0) •^ sX Q. O <: o o c CD 1 £ 3 -T CD *<

A CD D CD 5L *<" o O o 3- CD z o 3. oCD 00 ff T3 0) 0 CO o z 3 a o CO W Qni. ?r < ;+ <0- $ CD o *< cCD CD X -O -o CL oCD CD "0 o O O" 9) 1 3 ?r+r rn «< a O CD (D -^ —t 00 W —% 0) 3 CD

139 the banker's own note and that Backwell already has the tallies that Mounteney needs. Section II detailed a one year example of this relationship.

The fact that Backwell actually held the Treasury debt increased the surety for all parties concerned that the banker's paper debt would retire the tallies when the hour came. The expectation that Backwell's notes would swap for the tallies he held was the linchpin of the systematic avoidance of specie. Confidence in the eventual swap permitted the realization of the transaction cost savings of paper over specie for both the tax collector and the public. Without both public and tax official confidence, the banker would not have enjoyed the expanded valuation and circulation of his paper debt that recognition of his notes as a means of tax settlement could provide.

For Backwell, the flow of Customs revenue through his bank also provided a superb vantage with which to monitor the tax. Mounteney's account may have made the goldsmith-banker the most informed potential delegated monitor regarding the Customs which would complement Backwell's large holding of tallies secured on the Customs. Intimate knowledge of the Customs revenue would have improved Backwell's ability to act as a delegated monitor for the depositors. Backwell's roles as a note issuer and as an intermediary between the investing public and the Treasury were mutually reinforcing. The goldsmith-banker's notes could settle tax liabilities because Backwell held tallies.

Backwell could purchase more tallies because the depositors valued liquidity and credibility, both of which the banker's notes gained from the tax collector's acceptance of those notes. To the banker, to the public (sensitive to the liquidity of the note they were passing), and to the tax collector, the banker's role as a government debt holder was desirable.

V. Conclusion

The persistent presence of important Lombard Street goldsmith-bankers in or around

England's major revenues suggests goldsmith-banker notes may have been accepted for tax

140 payments for most of the second half of the seventeenth century. Routine revenue acceptance would have encouraged public use of goldsmith-banker notes across the spectrum of commercial activity. Moreover, inter-banker clearing meant those goldsmiths that did not hold tallies could still benefit from tax acceptance. For example, in 1670, Richard Mounteney presented Backwell with the notes of most of the goldsmith-bankers that Backwell cleared with. The network of goldsmith- banker clearing interlocked with those bankers tied to particular revenue streams. The combination of these two institutions (revenue acceptance of private notes and inter-banker clearing) fostered an environment in which private script became widely used, at least within London.

This and the preceding chapters have revealed Edward Backwell to be a prominent goldsmith-banker but not a unique figure. Backwell participated in but was not the center of or necessary for the network of debt clearing between goldsmith-bankers. Likewise, Backwell was neither the largest lender to the king nor the only goldsmith-banker to ally himself with agents of the public revenue. Sir Robert Vyner faced a greater loss than did Backwell under the Stop of the

Exchequer in 1672. After the Stop, forty-four goldsmith-banker shops were keeping demand deposits and clearing debt in 1677. In the 1670s and 1680s Sir Charles Duncombe joined with

Richard Kent to secure control over the collection of both the Excise and the Customs. Backwell's reputation as a proto-central banker stems from his ledgers being the only extant records of a

Restoration goldsmith-banker of his size. However, a truer story places Edward Backwell as part of a group of innovative goldsmith-bankers that firmly established debt issuing, fractional reserve banking in London.

The goldsmith-banker and time period examined in the next chapter change from Edward

Backwell in the years following the Restoration of Charles II to Sir Francis Child in the years surrounding the Glorious Revolution of 1688 in Chapter 4. Sir Francis Child was a junior partner in a small goldsmith-banking house in the West End when the Stop of the Exchequer broke Edward

141 Backwell, Sir Robert Vyner and a few of the other largest Lombard Street bankers. The next chapter considers how Child balanced the risk of government default against the extraordinary returns available from lending to the Treasury of King William III. The Nine Years War (1689 to

1697) brought an unprecedented supply of tallies into London's financial markets, and Sir Francis

Child responded by becoming a lender to the Crown.

142 Chapter 4

Tallies or Reserves? Sir Francis Child's Balance Between

Reserves and Extending Credit to the Crown, 1685-1695

As Chapters 2 and 3 have shown, a resilient private banking system existed for decades before the Glorious Revolution of 1688. Since Cromwell's Protectorate, money-scriveners, great merchants and especially goldsmith-bankers were providing a sophisticated variety of financial services to both the merchants of the City and the gentry of the West End. The goldsmiths, in particular, had a thriving system of banking that cleared checks and bank notes on a daily basis from as early as the Restoration in 1660.(Chapter 2) During the second half of the seventeenth century, the customers of London's goldsmith-bankers could deposit money at interest, write checks, negotiate overdrafts, discount bills of exchange or government debt (known as tallies), pass bank notes, exchange foreign specie and borrow funds secured by pawns, property, bonds or reputation.

William of Orange arrived in England in November of 1688 and by May 1689, England was at war with France. The new king had need of London's financial intermediaries to raise funds for the costly Continental war. The Nine Years War (1689-1697) between William III and his

Habsburg allies on one side and Louis XIV of France on the other was in scale, expense and debt much larger than the English conflicts of preceding decades. Under the new constitutional arrangements that accompanied William's ascension to the throne, Parliament provided the Crown with a stream of revenues conditional upon the legislature's regular re-authorization. Although

Parliament came to provide extraordinary funds over the course of the Nine Years War, expenditures grew faster than receipts.

143 With extraordinary revenues controlled by Parliament and expenditures necessitated by conflict, the English Treasury's immediate war-time budget deficit had to be covered by other means. One revenue raising alternative, seigniorage, was untenable. England had enjoyed a stable, sterling based silver standard since Henry VIII. Calling down the value of silver would have antagonized creditors and the landed class, constituents upon whom the new Monarch depended.(Li

1963)' Instead, borrowing was the traditional source of short run funds for the British Crown, and therein lurked a problem. Successful large scale borrowing had depended on the sale of debt to intermediaries such as London's goldsmith-bankers; however, earlier regimes had not treated their financiers well. Leading up to the English Civil War, Charles I had coerced loans and lost all credit.(Ashton 1960) Charles II had made great use of London's goldsmith-bankers in financing the

Second Dutch War (1665-67), but defaulted on the eve of the Third Dutch War in 1672. Charles

IPs default, called the Stop of the Exchequer, left London's most prominent goldsmith-bankers broken and without recourse. The Crown's credit remained weak until constitutional changes of the

Glorious Revolution.

In 1689, William HI turned to London's financial community for loans, and the records of

London goldsmith-banker Sir Francis Child establish how this banker responded. Francis Child was a junior partner during the Stop of the Exchequer in 1672 and had become a senior partner when

William III came to power in 1688. Although Child's banking shop did not hold government debt in either 1672 or 1688, the goldsmith-banker began purchasing tallies in 1689 and thereafter became an important player in England's financial war machine. The story of Child's lending behavior before, during and after the Glorious Revolution reveals a goldsmith-banker's shift from eschewing government debt to the active accumulation of Treasury securities. The goldsmith-banker was

1 When the coinage was reminted in 1696, the value, weight and purity of the silver coinage was maintained. The cost of recoinage was met with new taxes, and no seigniorage accrued to the Crown.(Li 1963)

144 responding to a flood of war debt offering high interest rates. To enable his purchases of government debt, the goldsmith-banker both shifted funds away from private sector lending and lowered his reserves to assets ratio.

Despite the substantial change in his lending behavior, Child could have extended even more loans to the Treasury, yet he chose not to. Child did not forsake lending outside the circle of

Treasury securities, and the goldsmith-banker's capital reserves were kept at substantial levels. This chapter submits the explanation that goldsmith-bankers like Sir Francis Child would not unreservedly purchase tallies because they faced potential ruin if they did. Child ran a self owned shop with unlimited liabilities at short run demand. The goldsmith-banker maintained reserves sufficient to handle severe runs rather than further extend his exposure to government debt. To the degree London's other bankers faced similar trade-offs, Francis Child represents the restraint private intermediaries chose in acquiring government debt.

The limits of London's private financial sector coupled with a very expensive war created a financial difficulty for the Treasury in the years following the Glorious Revolution of 1688. The expense of the Nine Years War has received considerable analysis, especially in Jones' War and

Economy in the Age of William III and Marlborough, yet this analysis has not addressed why private bankers would not lend more to the state in support of the war. The reluctance of private bankers to decrease their reserve ratios created a gap between what William III wanted to borrow and what private bankers like Child were willing to fund. This gap spurred famous innovations in public finance at the heart of the Financial Revolution. For example, the Bank of England was founded in 1694 explicitly to expand the funding of government debt. The unwillingness of financiers like Child to loan more to the Treasury is an unstated underpinning of such defining works on England's Financial Revolution as Dickson's The Financial Revolution in England,

Clapham's The Bank of England, Horsefield's British Monetary Experiments, Richards' Early

145 History of Banking in England and Roseveare's The Financial Revolution. The restraint of

London's private financial intermediaries was a driving precondition for the public sector

experimentation at the heart of the Financial Revolution.

I. The Shift to Tallies

With the death of Robert Blanchard in 1681, Francis Child took over the senior partnership

of the goldsmith-banking shop the Marygold, within Temple Bar, which straddled the far west end

of the City of London. In the decades to follow, Child built a successful banking house serving the growing West End. Unlike the goldsmith-bankers of Lombard Street, Child did not specialize in commercial credit for overseas trade, foreign bills of exchange or coins from abroad. He did,

however, cater to the needs of his West End clientele by retailing in jewelry and plate and by providing various financial services. Child accepted deposits, both time and demand, which he in turn loaned. Francis Child also maintained fractional reserves ranging from fifty to sixty percent of his total assets.

Over the years stretching from the early 1680s to 1695, Francis Child had an outstanding

loan account totaling thirty to fifty thousand pounds on any given day. In the mid-1680s, Child lent mostly to his own customers. Often the loan agreement was in the form of a pawn with the borrower securing the loan with plate, jewelry or a penal bond. Child even referred to his loan portfolio as the "Pawn Account." This goes back to the traditional aspect of goldsmiths acting as

pawn brokers. However, Child did regularly loan funds with no mention of collateral. Child also

lent to customers in the form of overdrafts. Unfortunately, Child did not tabulate the daily progress of his clients' overdrafts in any centralized way, so analysis of overdrafts is constrained to only the broadest of conclusions. Overdrafts are not considered further except to note that they ranged from twenty-five to thirty-five thousand pounds at any time (see Table 4.1, below). By 1695, England

146 had been at war with France for six years, and now Sir Francis Child had become a major participant in financing the government debt the war had created. Child's stock rose quickly after the Glorious Revolution of 1688. Child was elected Alderman and knighted in 1689, became

Sheriff of London in 1690 and the Prime Warden of the Goldsmiths' Company in 1691. His rise in station paralleled his move into the public debt market.

Child's reasons for shifting a large portion of his lending business to the Crown were twofold. The Crown offered access to offices such as Child's appointment as the King's jeweler.

On the purely financial side, the Crown exempted itself from the usury laws by offering up to a fourteen percent return. All others were legally constrained at no more than six percent per annum.

The royal premium was not a new arrangement. Goldsmith-bankers had been mediating between depositors, limited to a ceiling return of six percent, and the Treasury for decades. The Nine Years

War vastly increased the supply of tallies at high rates of return. "The immediate result of

England's entry into the war against France in 1689 was to make public expenditure increase between two and three times."(Dickson 1967, p. 46) Annual government borrowing rose tenfold from the years under William Ill's predecessor, James II, to the early part of the Nine Years War.

Table 4.1 presents the annual borrowing figures for the English government. Except for a slackening in 1692, the Nine Years War necessitated yearly borrowing that was many times larger than Charles II's most profligate year (Michaelmas 1670-71) of £597,000.(Chandaman 1975, p. 353)

The Treasury had to find a market for this debt and established bankers were one traditional outlet.

Attempts by the Treasury to sidestep financial intermediaries and sell debt directly to the public on a large scale had failed in the 1660s, so the government had few alternatives to the goldsmith-bankers for pooling funds seeking investment.(Roseveare, 1962) The vast new quantities of debt expanded the opportunities for new players like Francis Child to begin participation in the government debt market.

147 Table 4.1 English Government Borrowing (£,000)

1685-6 (Ms.-E.) 1686-7 (Ms.-E.) 1687-8 (Ms.-E.) 1688-91 (p.a.) 1692 254 153 181 1,703 344

1693 1694 1695 1696 1697 2,269 1,798 2,378 3,325 4,767

Sources: Chandaman 1975, p. 361; Jones 1988, p. 70.

The increased quantity of tallies (English government debt) was accompanied by a diminished risk of default. The re-arrangement of political power underlying the arrival of William in 1688 permanently shifted the regular authorization of revenue to the Parliament. Heightened

Parliamentary control over revenue and expenditure reduced the likelihood that the Treasury would fail to meet debt payments. Also, the placing of the Crown under the umbrella of the Common Law also constrained the king's ability to abuse royal prerogatives.(North and Weingast 1989)

The significance of the increased security afforded to the Crown's creditors arose from earlier government defaults. Charles II had stopped most debt payments out of the Exchequer in

1672 (a result of his heavy borrowing) and ruined the generation of goldsmith-bankers then lending to the Crown.(HorsfieId 1960, Richards 1928) In 1672, Child was a junior partner to Robert

Blanchard and their shop did not hold government securities. Child likely witnessed the fall of Sir

Robert Vyner, Edward Backwell, Jeremiah Snow and other goldsmith-bankers first hand, so the concern over the effects of another default in the 1690s was not negligible.

Greatly expanded debt issuance by the Treasury coupled with less risk of default encouraged goldsmith-bankers to expand their loan portfolios in favor of government debt. For Francis Child, this shift was pronounced. Chart 4.1 plots tallies as a percentage share of Child's monthly loan

148 Percent CO 2 -»-roco4^ano)-vicx>(DO oooooooooo _i 1 1 i i i i i 1—

2L

0) O

o 3, o3 4^ I o 0) 3 CO

149 creation. The figures have been encoded and the account reconstructed from Child's extant ledgers.

The records are mostly complete except for the nine months from April to December 1690 from

which only the records of a few transactions survive. Although presented throughout this paper, the

last three quarters of 1690 must be viewed as greatly understated.

Before the summer of 1689, the goldsmith-banker had purchased no tallies. In the years that

followed, government debt became the major portion of the goldsmith-banker's investment portfolio.

In some months, tallies reached eighty, ninety and even one hundred percent of Child's lending

activities. The portion of total monthly loans taken by government debt, however, varied greatly

from month to month. Receipts from repaid loans followed a pattern similar to lending. Returns

from government debt become a large portion of Child's total flow of returns starting in 1689.

The size of Child's lending activities did not suffer with the switch to government debt. In

Table 4.2, total loans picked up in 1688 and remained high in 1689. The results from 1690 are

limited by missing data. Lending in the years 1691 through 1694 are high compared to the pre-war

years of 1685 through 1687. Child's lending fell off in 1695, but the year was notable for inflation

and the collapse in the value of clipped silver coins.(Jones 1988) Until the monetary crisis was

resolved in 1696 and 1697, uncertainty would have diminished incentives to lend.(Horsefield 1960,

Li 1963) Throughout the ten year period, the variance in amount Child loaned per month remained

substantial. Child lent in spurts, both in total and to the Treasury. Receipts also arrived in highly varying amounts.

150 Table 4.2

Total Amount Loaned by Francis Child

Year No. of Loans Monthly Monthly (Jan- Total Average Standard Dec) Months Deviation 1685 7 £32,708 £4,673 £4,058 1686 12 57,088 4,757 3,639 1687 12 52,345 4,362 3,568 1688 12 120,275 10,023 7,641 1689 12 116,877 9,740 6,185 1690 3 3,392 1,131 844 1691 12 81,377 6,781 3,808 1692 12 97,912 8,159 8,487 1693 12 143,381 11,948 8,999 1694 12 77,388 6,449 6,681 1695 12 52,864 4,405 3,316

Source: Child's Ledgers

The year 1688 was an exceptional one for Child. In the midst of a deepening political crisis that would end in James ll's leaving England, Francis Child loaned large amounts to a number of well connected people. Child lent heavily to both Whig and Tory merchants during the uncertain months of 1688. Child lent Sir John and Francis Eyles, brother merchants who later became financiers for William III, over five thousand pounds. Sir John was a West Indies merchant while

Sir Francis would go on to direct the Bank of England and the old, the new and then the Unified

East India Companies.(DeKrey 1985) The goldsmith-banker also lent thousands to Sir Samuel

Dashwood, an important Tory Levant Merchant. In a rare entry, Child even lent three thousand

pounds directly to the East India Company in February of 1688. Child was elected alderman and

151 sheriff in 1689 as a Whig.(DeKrey 1985) Only later, in response to Whig support for both the New

East India Company and the Bank of England, did Child became a Tory.

During the height of the political crisis in late 1688, the future was less certain. In October of 1688, Child changed tack. Instead of the extensive lending of earlier months, Child collected

£37,735 and loaned no new funds. In December of 1688, Child's total estimated loan portfolio was

less than £10,000, at most one-third his regular lending level. With no offsetting increase in reserves or fall in liabilities, Child's estimated net worth plunged to nearly £25,000 in the red.

Child did not record his December 1688 level of lending. The author's estimate (see Table 4.4) cannot explain where the funds collected from the completed loans in late 1688 went.

With the establishment of the new regime in 1689, Child loaned considerable amounts to established customers and dabbled in his first tally purchases. The new Lord Chamberlain, the Earl of Dorset, was one of Child's established account holders. With such contacts, Child was knighted, elected Alderman and held the post of Sheriff of London, all in the year 1689. Child lent £116,877 over the course of the year. Unfortunately, the folios containing March through November of the

next year, 1690, have mostly been lost. Judging from the goldsmith-banker's outstanding loans in

December of 1690, around £28,000 worth of new lending went out without the arrival of off-setting

receipts during the last nine months of 1690. In 1691, the goldsmith-banker resumed lending liberal

quantities with government debt as a large percentage of total lending.

Party loyalties aside, Child's response to the increased supply of government debt was a

reaction to the high rates of return offered on tallies. The comparative examples of the same

principal amounts in Table 4.3 highlight the extra returns the goldsmith-banker could acquire.

152 Table 4.3

Internal Rates of Return for Sample Loans

Loan Receipt Internal Rate of Return 27 July 1688 21 June 1689 5.5% William Brownlowe £2000 principal £2000 £97:18:4 interest 27 June 1689 12 August 1689 6.9% Exchequer £2000 principal £2000 £16:19:6 interest 12 August 1691 3 February 1692 8.3% Tallies on Land Tax £2000 principal £2000 £78 interest 6 November 1691 16 March 1692 4.5% Mr. Maynard £250 principal £250 £4 interest 27 June 1691 15 October 1691 6.7% Exchequer for Tally £250 principal £250 £4:18:6 26 September 1693 2 December 1693 9.6% Tally on Customs £250 principal £250 £5:16:3 interest £1:1:9 discount

Source: Child's Ledgers

During the first year of his reign, William III supplemented extra interest premiums with his personal attention to encourage lending by Child and other goldsmith-bankers. The King was present with his Lords of the Treasury when securing loans from such prominent goldsmith-bankers as Francis Child, Stephen Evance, Charles Duncombe and Joseph Hornby.(CTB IX, pp. 35-44) The royal premium was also supplemented by offices. As mentioned above, in 1689, Child was made

William Ill's "Jeweller in Ordinary" for the duration of the war. The King's jewelry account was worth over £20,000 in business over its first four years.(Samuel 1977, p. 49) Similarly, goldsmith- banker Stephen Evance received an officership in the Excise over the course of the war.

153 The Treasury needed to offer premiums and the other inducements because lending to the

Crown, despite the constitutional reforms of the Glorious Revolution, was still risky. William Ill's government could collapse and/or the war could be lost. The goldsmith-bankers caught holding tallies in the government's default of 1672 (the Stop of the Exchequer) never regained but a fraction of their principaI.(Horsefield 1982, Richards 1928) After years of petitioning, the Treasury established a permanent stream of interest payments but never fully compensated the jilted debt holders. Moreover, the goldsmith-bankers involved were hounded by litigation for years and never regained their prosperous former positions. Child's conservative accumulation of tallies in 1689 suggests a testing approach on the goldsmith-banker's part. Only with successful maintenance of the debt would the new regime show they had, "as contemporaries alleged - the secret of running the state over head and ears in debt."(Dickson 1967, p. 39)

Success in the public debt market required of more participants like Francis Child than just purchasing Treasury securities. For decades, goldsmith-bankers who invested in tallies had established a specialization in providing mediation between the public and the Exchequer. Many goldsmith-bankers gained advantage as intermediaries in the government debt market by remaining close to the flow of public funds. By participating as bankers and officers in both revenue collection and government expenditure, goldsmith-bankers such as Edward Backwell, Stephen

Evance, Richard Kent and Charles Duncombe gained working knowledge of the flow of public funds in and out of the Exchequer.(CTB, Nichols 1985, Richards 1929) Child's inexperience with public debt meant he needed to develop specific knowledge concerning the intricacies of dealing with public debt.

Beyond information on the Treasury's likelihood of having sufficient funds to make payments, goldsmith-bankers benefited from familiarity with how the Exchequer worked. As Isaac

Meynell, a goldsmith-banker, explained, depositors who were unaware of, "the manner and

154 proceedings of the Exchequer and the way of paying in and receiving money soe advanced," lent their money to the goldsmith-bankers instead.^(Chancery, PRO C9 110/27) With the acquisition of

Exchequer specific human capital, Child could profitably expand his foray into public debt investment.

Child, however, already possessed the most important aspect of public debt intermediation: liquidity. Prior to the Glorious Revolution, Francis Child had managed checking accounts and issued bank notes while maintaining fractional cash reserves. Child was already experienced at providing liquidity to depositors. "In no other respect were the goldsmith-bankers so attractive to the small investor than in their ability to repay on demand or at short notice."(Roseveare 1962, p.

260) The ways of the early modern Exchequer were still medieval, despite the introduction of orders and other modernizations during the second half of the seventeenth century. An investor could not make direct use of the Exchequer, "without much trouble and loss of half a day and no certainty of having the offices open."(Pepys, p. 123)

With high transaction costs and illiquidity, merchant aversion to purchasing Treasury securities directly was wide spread. As D.W. Jones discovered, "Merchants contributed but 5.3 per cent to loans on the Land Tax in 1692-3; yet they accounted for 30 per cent of investments in both the Bank of England and East India subscriptions of 1694 and 1698 respectively."(Jones 1972, p.

340) Goldsmith-bankers such as Francis Child and innovations such as the Bank of England allowed investors to maintain liquidity through withdrawal of deposits, in the case of bankers, or sale of stock, in the case of the Bank of England.

In 1689, the Treasury went to unusual lengths to facilitate goldsmith-banker liquidity. The

Lords of the Treasury explicitly directed the Auditor of the Receipt [Exchequer],

2 The full quote excerpted above is nearly identical to one found by Roseveare in an Exchequer proceeding written by another goldsmith-banker, George Snell.(Roseveare 1962) The statement must have been a legal "boiler plate" explanation of goldsmith-banker intermediation.

155 to accept [as cash] the [bank] notes of Mr. Percivall and Mr. Evans, or of Mr. Atwell and Mr. Courtney, or of Mr. St. John, goldsmiths in Lombard Street, or of Lieut.-Col. Child or Sir Thomas Fowlys, goldsmiths in Fleet Street.(CTB IX, p. 123)

This directive spared goldsmith-bankers, including Child, from presenting specie to the Exchequer and bolstered the reputations of participating goldsmiths. Although direct teller acceptance of goldsmith-banker notes did not become an enduring policy, the Treasury had bent the rules, "at their [the goldsmiths'] request and for their convenience,"(CTB, p. 123) because promoting liquidity was a key factor in the success of early modern financial intermediaries.

II. Limits

While Francis Child responded to the Treasury's expanded debt offerings, he did not abandon lending to his private borrowers. Also, the goldsmith-banker retained substantial capital reserves. To the degree that Child was representative of the other goldsmith-bankers of Lombard

Street and the West End, private bankers limited their purchases of public debt. With a very expensive war on the Continent, the Treasury needed more financing than private intermediaries were willing to provide. Goldsmith-bankers like Child did not offer unlimited funds to the English government because they had to maintain their own positions as viable financial intermediaries. The goldsmith-bankers faced unlimited liabilities. The personal penalties for risk were substantial.(Earle

1989) While Child was arranging new loans with the Treasury, his fellow goldsmith-banker,

Jeremiah Snow, who was still suffering from the crippling blow of the Crown default in 1672, petitioned the same Treasury Lords for clemency. "Sir Jeremiah Snow, now a prisoner in the Fleet, who desires a writ... to [enable him] to go [out or] abroad to sell what estate he has to pay his creditors: same being a compassionate case."(CTB) Francis Child had to balance the expected returns from the marginal investment in tallies against the risk of depleting his reserves or overly concentrating his portfolios and becoming a "compassionate case."

156 CO 2 Pounds O CD O (Thousands) :x a. 8 o oCD 0o0 oo a § J. -I 1 L 5" Q. CO. CD w

157 Child's total loans to all others besides the Treasury peaked in 1688, but, as Chart 4.2 reveals, Child maintained ending outside of the realm of government debt. Child achieved diversity in his lending portfolio by balancing public and private debt. Within tallies, the goldsmith-banker varied his holdings by purchasing debt secured on various revenue streams: Customs, Excise, East

India Goods, Wine and Vinegar, etc. This would help if one revenue stream had funding problems, but did not insure against a general default. The banker also diversified the nature of his public debt by mixing new loans with purchasing tallies closer to maturation. The last example in Table

4.3 shows that Child purchased a £250 tally on the customs and paid an additional £1:1:9 before receiving £5:16:3 two and a half months later. Dickson first noticed Child's increased use of discounting, especially after the founding of the Bank of England in 1694.(Dickson 1967)

Discounting by bankers such as Child facilitated the circulation of long run debt by increasing the liquidity of government debt. Discounting also allowed intermediaries like Child to shorten the term structure of their own government debt holdings and to match it more precisely to the term structure of their liabilities.

Child expanded his wartime volume of credit by lowering his reserves to assets ratio as shown in the first row of Table 4.4. The 1690 and 1694 ratios were ten to fifteen percent lower than both the pre-war years (1685, 1688) and the year of monetary crisis, 1695/6. Child had increased his stock of loans outstanding by five fold from an estimated £9,743 in December 1688 to

£48,000 two years later. Likewise, overdrafts expanded from 1688 to 1690 by £6,000. Over

158 Table 4.4

Breakdown of Francis Child's Balance Sheet Taken from Child's "Casting Up" Accounts

29-May 08-Dec 31-Dec 21-Jul 05-Feb 1685 1688 1690 1694 1696 Reserves/Assets 57.9% 63.2% 48.2% 48.0% 59.1%

Assets, Total 129,191 105,802 161,383 118,031 138,153 Reserves 74,822 66,842 77,861 56,700 81,582 Ledger Balance (-) 24,750 29,487 35,387 24,684 25,945 (Overdrafts) Pawn Acct 29,619 9,473' 48,135 36,647 30,626 (Loans)

Liabilities, Total 124,014 130,186 160,635 109,869 189,666 Abstract Book 76,857 54,269 59,896 Ledger Balance (+) 124,014 130,186 79,478 55,600 69,501

Others 4,300 400 Net Worth 5,178 -24,384 747 8,162 8,355

Reserves, Total 74,822 66,842 77,861 56,700 81,582 Silver coin 40,128 37,225 46,373 38,692 19,749 Gold coin 12,357 5,784 9,763 2,555 19,898 Bills 17,648 Total Jewels, plate, etc. 22,337 23,834 21,726 15,453 24,287

Source: Child's Ledgers

3 Estimated based on the May 29, 1685 Pawn Account balance of £29,619 less net accumulated receipts of £28,325 (£266,796 debits minus £295,121 credits from May 1685 to December 1688 plus £8,179 in interest, credit side, over the same period).

159 the same two years, Child's stock of reserves, mostly silver and gold coins, rose by only £11,000.

Four and a half years later, Child had a similar reserves to assets ratio (48.2% in 1690 and 48.0% in

1694), but his stock of loans had slipped £12,000 while his reserves had slid by £20,000. Rather than crowd out non-government borrowing, Child increased his vulnerability to runs and panic.

It is difficult to say whether Child was typical of London's goldsmith-bankers in either

keeping his reserves to assets ratio in the fifty to sixty percent range or in lowering fractional

reserves to fund expansion into government debt. The breakdown of Child's assets and liabilities comes from a "Casting Up the Shoppe" book, a book separate from his ledgers and loan accounts.

The book is exceptional by providing the only listing of a goldsmith-banker's gold and silver stocks

known to the author. A banker's stock of cash is necessary for the reserve calculations in Table 4.4

so, at the present time, the levels of Francis Child's reserves are an unique insight. Beyond

establishing that goldsmith-bankers were fractional reserve bankers, one can only extrapolate

specific conclusions from Child to other goldsmith-bankers with care.

As a fractional reserve banker, Child could expand his lending by issuing notes or crediting

accounts. The goldsmith-banker had great short run flexibility. The value of his paper debt to the

public was enhanced by clearing arrangements with fellow West End bankers, such as Sir Thomas

Fowle, and Lombard Street Bankers, such as Sir Stephen Evance and Sir Charles Duncombe.

Participation in the inter-banker system of clearing notes and checks (demandable debt) allowed

goldsmith-bankers like Child to float notes more widely. Child appears to have engaged in such

credit creation, especially in 1689 and 1690.

Bearing unlimited liability and specializing in the ease of depositor access, Child was

vulnerable to runs as he decreased his fractional reserves. The constraint for Child, however, was

not his daily cash flow. With fifty to sixty percent cash reserves, Francis Child could cover any

typical day's demand for liquidity. Child was prepared to cover even the highest unexpected

160 withdrawals.(Diamond and Dibvig 1983) Rather, substantial investment in government debt heightened the chance of a panic because of asymmetric information between Child and his depositors.(Calomiris 1993) Even if a banker such as Child knew the future flow of payments

looked promising, he was vulnerable to runs if his depositors feared government default. Political rumor or battlefield news could leave customers worrying about another Stop of the Exchequer.

Not knowing how many tallies each banker held or each bankers' cash reserves, the deposit holding public could run on all goldsmiths known to have lent to the Treasury. Without knowing each banker's particular exposure to default, a macro shock could precipitate a general panic. Runs, for example, occurred in 1667 under the strain of the Second Dutch War.(Roseveare 1962)

Depositors were aware that the goldsmith-bankers were big lenders to the Crown but would have had difficulty knowing exactly how much each had loaned at any particular moment.

Trudging to Westminster to examine the Exchequer's transfer books was a high monitoring cost. In fact, processing the transfer of tallies, collecting interest payments and the other aspects of dealing with the Exchequer were value adding specializations that kept depositors lending to bankers rather than loaning directly to the Treasury.(Roseveare 1962, Chapter 3) Monitoring would diminish a banker's value added as a specialized intermediary.

The liquid nature of Child's liabilities made a run a formidable concern. All of Child's

liabilities were demand account balances. According to Table 4.4, the goldsmith-banker had about enough cash and plate on hand to cover a withdrawal of half of his demand accounts. Again, we cannot say if Child was an especially conservative banker relative to his peers, but the goldsmith-

banker was prepared, even before taking on government debt, for a considerable run.

If Child's banking colleagues also kept substantial reserves, the Crown's ability to expand

government debt was truly limited. While Child did make a substantial shift towards holding tallies,

the goldsmith-banker only bent his balance sheet so far. As the other intermediaries faced their own

161 thresholds, Treasury debt issuance would have been constrained. Moreover, raising the premium of tallies was not a clear solution for the Treasury. Higher interest rates implied a increased burden to the government in making payments, which added to the likelihood of their defaulting. The possibility of default, or at least the public expectation of possible default, was already limiting private lending to the Treasury. Increasing rates might have attracted more funds but also would have heightened the fear of runs.

HI. Conclusion

Francis Child responded to the Nine Years War by substantially reworking his investment portfolio. Lending was expanded and reserves fell as a percentage of assets. Child, however, restrained his exposure to public debt and maintained sufficient reserves to cover a substantial run.

The goldsmith-banker also kept diversity in his portfolio by retaining loans to private individuals.

Child's story suggests London's responsive private financial sector did not provide the war- pressured Treasury with all the credit the army and navy desired because the bankers feared the risk of runs. The limitations of goldsmith-banker lending to the Treasury necessitated experimentation in public borrowing. Lotteries and a land bank were tried, but the Bank of England truly allowed

William III to circumvent Sir Francis Child and London's other existing financial intermediaries.

This chapter has focused on a goldsmith-banker's ratio of reserves to assets, yet Francis

Child's stock of precious metal reserves was the substantial sum of over forty thousand pounds.

The next chapter examines the management of such large cash holdings independent of lending behavior. Although reserves were not earning interest, the relative value of gold and silver did fluctuate. To see how goldsmith-bankers and their merchant customers adjusted their bullion stocks with the movement of international prices, Chapter 5 returns this dissertation to Lombard Street. Sir

Stephen Evance was a Lombard Street goldsmith-banker and part of an international network that

162 shipped bullion and financed armies. The movement of bullion in and out of Evance's shop during the Glorious Revolution suggest how goldsmith-bankers assisted England's progression towards monetary crisis and eventually the gold standard.

163 Chapter 5

Gold, Silver and the Glorious Revolution:

International Arbitrage and Goldsmith-Bankers

The gold and silver reserves goldsmith-bankers kept in their shops was of tremendous value.

Beyond holding bullion stocks, trade of precious metals was also a defining activity of goldsmith- banking. The earliest goldsmith-bankers held deposits of plate for safe keeping and accepted precious metals as pawns. With time, goldsmith-bankers expanded into more sophisticated activities

like international bullion arbitrage. One such goldsmith-banker was Sir Stephen Evance who, during the last decades of the seventeenth century, bought and sold large quantities of bullion from his shop on Lombard Street.

In addition to trading in precious metals, Stephen Evance secured foreign credit for

important customers, most notably the English Treasury. Both specie and credit mattered to

Evance's participation in international markets because bullion and credit were substitute media of

exchange between nations. Gold, silver and credit were separate markets with different exchange

rates, but if their prices diverged, merchants and governments would employ whichever medium

offered the best exchange rate. Financiers such as Evance added arbitrage behavior to the process

of exchange rate adjustment. By trading one substitute for another, seventeenth century style

arbitrage captured the returns created by the differences between the exchange rates of the precious

metals and credit. Arbitrage reinforced the movement of the international bullion and credit markets

towards equilibrium, and bullion trading bankers like Stephen Evance played a central role.

164 The process of equilibrating the international markets was of great importance to late seventeenth century England. The imbalances of exchange rates drove the flow of bullion into and out of the kingdom's monetary stock. Changing costs of international credit affected both commerce and the state's ability to maintain the navy at sea or troops campaigning on foreign soil.

The state of Europe's international financial markets could strengthen or undermine the commercial and military viability of England's emerging great power status. Eventually, the international bullion flows of the late seventeenth century pushed England's monetary regime to the breaking point. The crisis surrounding the Great Recoinage of 1696 brought rigid controls on the price of

English gold and heralded the de facto beginning of a fixed value, bimetallic standard for early modern England. The origins of England's gold standard begins with a crisis in the stock of domestic bullion — a crisis developing out of the bullion transactions of financial intermediaries like the goldsmith-banker Stephen Evance.

This paper presents a portion of Stephen Evance's bullion accounts. These records reveal the micro-level web of transactions within the process of international bullion movements during a critical period in English monetary history. Evance's accounts portray an organization of late seventeenth century specie arbitrage dominated by a select network of bankers and merchants. In particular, the evidence details the change in the behavior of Evance and his associates in response to the first years of William III. The Glorious Revolution of 1688 brought the Nine Years War

(1689-1697) between England, her allies and France. The war required unprecedented capital flows out of Britain to support England's navy at sea and army on the Continent. For a cash starved government, assistance in financing an expensive war was essential. "The Wars now adays [sic] seem rather to be waged with Gold than with Iron, and unless we Pay well, we shall never be able

165 to Punish well."1 Financial experiments such as the Bank of England resulted from this difficult

period. The war also disrupted Britain's overseas trade. With rising military expenditures and

limited exports, Britain faced growing balance of payments problems.

Britain settled her international accounts by exporting bullion, and Evance played a central

role. In 1691, the English Treasury contracted out to Evance and his partners the actual transfer of

military funds.2 These intermediaries moved the funds through the existing international markets.

More than eight million ounces of silver left England over a seven year period from 1689 to 1695.3

Such large movements of bullion had pronounced effects on the international exchanges of the late seventeenth century.4

Large war remittances from England outstripped reciprocal flows of funds into England. In turn, the English pound sterling weakened relative to the Flemish schellingen banco (Amsterdam bank money). The exchange rate between London and Amsterdam, measured as schellingen banco per pound, fell from 1688 until England's monetary and financial crisis in 1696. The pound fell as

England's military payments grew.

The falling exchange rate was for neither gold nor silver but credit. In the seventeenth century, the principal method of moving purchasing power between nations was by bill of exchange

1 Observations Upon the Bank of England, (1695), quoted in Professor Roseveare's (1991) introduction, p. 1.

2 Section 111, below, considers how the military was paid.

3 "The Total recorded export of molten silver between 1689 and 1695, including what the East India Company exported in this form (though not as foreign coin) since clipping was the likelier source of this, amounts to 8,402,078 oz. Compared with this, the total silver known to have been lost from silver coin between these dates amounts to some 9,174,200 oz. So the amount exported is large enough to account for the bulk of the loss through clipping from silver." (Jones 1988, p 231)

4 England's war financing as an important but by no means the sole shock on exchange rates. The French called up their silver by ten percent in 1689. (Li, p. 55). Moreover, "Between 1688 and 1691," the Dutch, "Set about minting massive quantities of schellingen at a lower bullion content than that laid down by the money ordinance of 1681." (Jones 1988, p. 142).

166 (Kerridge, van der Wee). The bill of exchange was a contract promising payment of a foreign currency in a foreign port at a future date. Before coming due, the bill of exchange was negotiable, so bills were used as a medium of exchange. Although bills of exchange were promises to pay in specie, merchants valued bills as a way of avoiding payment in specie. Systems of off-setting debts, especially within exchange banks, lowered transaction costs by minimizing the use of precious metals. Credit, embodied by the bill of exchange, was the dominant means of business payment, so the credit exchange rate was the going price of moving purchasing power between currencies.

Gold and silver, however, could also be used for international payments and had their own effective exchange rates. Given two nations with bimetallic standards, the ratio of each nation's money associated with a fixed amount of bullion created a silver or gold exchange rate. Such

"mint" exchange rates, however, were replaced in importance by the market price ratios for bullion when the specie prices were free to adjust.

Arbitrage between credit, gold and silver played on the differences between these exchange rates.5 If the credit exchange rate fell below silver's, a merchant in London would gain more purchasing power in Amsterdam by exporting silver than by buying a bill of exchange on

Amsterdam. In contrast, a merchant in Amsterdam preferred to purchase a bill on London as the credit exchange rate fell (Sb/£). A weak pound induced the export of silver from London to

Amsterdam and the repatriation of the capital from Amsterdam to London by bill of exchange.

Returns to arbitrage lasted until the gaps between the gold, silver and credit exchange rates closed. The process of market adjustment, however, was complicated by the monetary regimes of the day. Seventeenth century England was on a silver standard, meaning law fixed the unit of account value of silver coins. Consequently, market adjustments for silver coins could not be made

5 Redish (1990) described the instability of a bimetallic system. This paper adds credit as a third element.

167 in terms of the price of silver coins. The fixed price of silver coins meant arbitrage opportunities

were reconciled by large shifts in the quantity of silver supplied to the foreign exchange markets. In

contrast, the unit of account price of gold coins could and did float above their mint value. When

constrained by a price floor, gold likewise adjusted by quantity shifts rather than domestic price

movements. Section I will consider the process of market adjustment in detail.

Stephen Evance's bullion accounts document at the micro level the change in the flow of

bullion from into to out of Britain during the realignment of exchange rates that accompanied

England joining the Nine Years War against France. Over the twenty-seven month period covered

by Evance's records (March 1688 to August 1690), William III came to the throne and England

went to war. Over the same period, the pound fell from a strong position relative to the schellingen

banco. At its height, the credit exchange rate lured gold from Amsterdam to London, and Evance's

accounts to show that he increased his holdings of gold. As the pound fell, Evance dramatically

slowed his accumulation of gold. Moreover, the banker's customers, many of whom were active in

Dutch markets, stopped selling him gold. By 1690, the pound's fall was beginning to reward the export of silver from London. The same merchants with extensive Dutch contacts began buying

silver from Evance.

The behavior of Stephen Evance and his associates confirms the dynamics of the early

modern international bullion markets and set the stage for an English monetary crisis. Merchants and bankers balanced the international gold, silver and credit markets through substitution and

arbitrage. Merchants switched to the lowest relatively priced means of international payment, be that gold, silver or credit. Moreover, arbitrage between this threesome of alternative media of

exchange accelerated inter-market equilibration. The dynamics of the three substitute exchange rates

was driving England towards a monetary collapse as the Nine Years War raged on. The price of

gold had skyrocketed and over half the silver content of the circulating coin had been clipped away.

168 By the end of 1695, after years of recurrent credit shocks, the price of domestic English gold coins

(guineas) had risen fifty percent and the average silver content of circulating coins was less than half the original mint weight.

The monetary crisis of 1695 and 1696 led to the Great Recoinage of 1696 which had two long-lasting effects. First, silver coins were reminted back to full weight and not allowed to fall in weight again. Second, a price ceiling was imposed on gold. Both domestic gold and silver were stripped of their ability to adjust to international pressures except by the demonetization of coins and export of their bullion. Because England fixed her gold to silver ratio higher than France or

Holland did, future credit shocks that weakened (strengthened) the pound exchange rate drew silver out of (drew gold into) Britain. Although England was legally on a silver standard, the interplay of gold, silver and credit was moving her inexorably towards gold. The story of Stephen Evance is representative of the micro level dynamics that over the second half of the seventeenth century and the through the eighteenth century led England to adopt the gold standard.

Before examining Evance's bullion transactions in Section II, Section I will detail the particular background conditions regarding gold and silver necessary to understand the process of adjustment between gold, silver and credit in seventeenth century London. Section III highlights the political economy of Evance's services on behalf of the Crown. Soon after the period covered by the goldsmith-banker's known records, Evance acquired the additional role of transferring English military remittances to the Continent. The market process of bullion/credit arbitrage complemented the political economy of financing England's military. As a financier on behalf of the Crown,

Evance purchased credit on the Low Countries. As an international bullion dealer, Evance sold silver for export to the Low Countries. The two activities, taken together, constituted credit/bullion arbitrage while allowing syndicates, in one of which Evance participated, to provide William Ill's

Treasury with transfers of military funds.

169 I. Gold, Silver and Credit Exchange Rates

What had held the Species at home for some years, but its badnesse; and when they made it good, should they not have taken Some care to keep it here? Did they not know that Nations who cannot Satisfye their Ballance with Commodities, must doe it by Money & Bullion or become Bankrupt? Was it not plaine this overballance agst Us proceeded chiefly from the great Remittance to the Troops abroad? And could the overballance soe arising be any way hindred but by procuring Forreigne Creditt? Was it not evident the Money would goe away as fast as minted? Charles Davenant, 1696*

The theoretical framework of this paper begins with three exchange rates - gold, silver and credit. The three forms of purchasing power were substitutes, though not perfect substitutes.7

Differences between their exchange rates created arbitrage opportunities. Profit seeking behavior shifted the demands and consequently the prices of these substitutes. The economic process of adjusting these rates towards equivalence produced international bullion dislocations and domestic price changes.

Analysis assumes three exchange markets (gold, silver and credit) spanning two countries.

The two nations considered are England and Holland. Amsterdam was the hub of northern Europe's bullion trade and at the center of William Ill's remittance program to the Continent while Evance's book recorded transactions in London.

For gold, silver and credit, the same exchange market bridged London and Amsterdam. The prices of bills of exchange (credit) were expressed in Sb/£ (schellingen banco per pound sterling)

6 Charles Davenant's Memorial Concerning the Coine of England extracted from his Memorials, (1696), reproduced as an appendix in Li (1963), p. 209.

7 Credit was the preferred means of payment throughout the seventeenth century. Gold and silver had higher transportation costs than credit, and metals had to be assayed. Moreover, English coins and bullion could not be legally exported. This paper assumes traders would always prefer credit to specie if both have equivalent purchasing power. Because gold was about fifteen times more valuable by volume than silver, gold had limited uses in retail trade. Silver coins dominated small value transactions. (Supple 1959)

170 terms, so the price of credit in one city will be taken as the inverse of the price in the other. This assumption exaggerates the integration of the seventeenth century credit markets between London and Amsterdam for explanatory purposes.8

The prices of gold and silver in each city were denominated in the local money. The bullion exchange rates have been calculated as the ratio of one nation's price for a given quantity of precious metal over another's price for the same quantity of precious metal. In an initial equilibrium, the exchange rates for all three markets between London and Amsterdam are balanced, or

P P W ^Credit ± ~ p p '

In equilibrium, merchants are indifferent between using credit, gold or silver.

Chart 5.1 begins in equilibrium, but the shock of war financing shifts the demand for credit in London (on Amsterdam) outward. The shift in demand for credit in London raises the price of credit and increases the quantity supplied. In Amsterdam, the lower half of Chart 5.1, the shift in the demand for credit in London registers as a shift in the supply of credit on London.9 The price of credit in Amsterdam falls (the inverse of London's rise) and the quantity supplied rises to match

London's increase in demand.

8 The price of credit exchange rates in Amsterdam and London systematically differed. The gap created an implicit interest rate.(de Roover 1974) The exchange rates of the eighteenth century were highly integrated (Neal 1990), and available evidence supports this view for the later half of the seventeenth century, but more work needs to be done.

9 The same shifts would occur when English agents sold bills of exchange, in Amsterdam, on London.

171 > 3 s o % Q. O CD s 3 CO ICL I 3" 0) 3 (O CD ^ 0) I IS

0) o 3

I—H O) o 3" O O 7T

172 The higher price of bills of exchange in London generates more demand for credit's substitutes. The rising demand for gold and silver in London increases bullion prices (£/Sb) and quantities supplied. In Amsterdam, London's growing demand for specie appears as a shift in the supply of both metals as more bullion entered the international market.

The market responses to the initial credit shock result in the quantity of gold and silver used in international payments increasing. The markets clear when the ratio of the prices of both gold and silver in Amsterdam and London matches the new credit exchange rate. In Chart 5.1,

pi pi ml _ GoW6,**,» _ SUvcr^^ ,,v e C"#

A number of historical conditions altered the functioning of exchange markets in the seventeenth century. In Holland, the prices of gold and silver, relative to credit, were fixed. The

Amsterdam Exchange Bank, called the Wisselbank, only credited accounts based on a fixed volume of gold or silver."(van Dillen) Hence, bank money at the Wisselbank was a guaranteed volume of gold or silver. All but the smallest bills of exchange were legally obliged to clear through the

Wisselbank, so the credit market worked by the Amsterdam Exchange Bank's rules." The

Dutch/English credit exchange rate was measured in schellingen banco, banco meaning in terms of

Wisselbank money. The number of pounds a schellingen banco could purchase on the open exchange varied, but the amount of silver or gold in a schellingen banco did not.

10 For an explanation of the silver and gold content of Amsterdam Exchange Bank money, see the Appendix E.

11 Bills of exchange over 300 guilders (about £25) had to clear through the Wisselbank.(Assar 1987, p. 105)

173 Return to the example of a credit shock from wartime capital transfer, and add fixed prices

(P Bar) in Amsterdam for gold and silver. The same equilibrating process occurs, but only the

pound sterling portion of the bullion exchange rate (price ratio) adjusts. With a fixed denominator,

the higher £/Sb exchange rates for gold and silver derive from higher prices in the English specie

markets. In Chart 5.2, the growing quantity of bullion supplied in the London exchange markets

corresponds with a contracting supply of gold and silver in the English domestic bullion market. In

equilibrium, the price of bullion in England, P1, must clear both the domestic market and the foreign

exchange market. Because the banco prices of gold and silver in Amsterdam do not fall to help

balance exchange rates, more precious metal has to flow out of domestic British use than if banco

prices did vary.

Arbitrage between credit, gold and silver hastened the process of reaching equilibrium

exchange rates. Bullion entered the exchange markets because gold and silver from England

purchased more in Amsterdam than credit did. The reverse also held; credit from Amsterdam

purchased more in London than Dutch bullion did. The misaligned exchange rates rewarded the export of bullion from London and the purchase of credit in Amsterdam (on London) when

returning the purchasing power. The repatriation of capital to London by bill of exchange decreased the supply of credit in Amsterdam on London, which strengthened the pound and closed the gap between the credit and bullion exchange rates. Stephen Evance was an intermediary who facilitated the exchange of bullion for credit and the movement of bullion between the domestic and foreign

exchange markets.

174 <—*•" Q_ 3" C T| 0) CO X 3 CD Q_ 3 Q_ CD 3 "0 o »-«- o' 3 00 8 1 1 o 3" CD o CD > 00 3 3 c_ Tl CO z^z r+ -OT CD o3 CD Q—^. CO ro CD ^ 3 3 0) m CO a g CO cr 0) DO 3 0) CO s3 CD D

175 A. The Domestic Markets

The constraints on England's domestic gold and silver markets also must be considered before examining Evance's behavior. When gold or silver entered the exchange markets, the bullion simultaneously left the domestic markets. However, the domestic legal environment added limits on the prices and movements of domestic coins and bullion. The mint values of coined precious metals interrupted the equilibration of domestic and international markets and created larger inter-market bullion flows than otherwise would have been the case.

The price floor for coined gold was the mint price of twenty English shillings a guinea set in

1663. The mint price, however, soon rose to twenty-one shillings six pence a guinea. As long as gold's market price stayed above 21s 6d a guinea, the mint price was not a constraint on the market price. The prices of English gold bullion and coin were free to float above the price floor. As demand for gold rose in international exchange, gold was drawn from domestic use and the price of gold coin and bullion rose in both foreign exchange and domestic markets. The resulting shift in supply from domestic to international markets kept the price of gold equal in both uses.

In reverse, a strengthening pound on the credit exchanges brought gold into London and lowered the pound sterling prices of bullion and coin. Chart 5.3.1 starts on the left with a strong pound (low £/Sb) on the foreign exchange market. Gold flows to the higher domestic prices, so the supply of gold in foreign exchange contracts while the supply of domestic gold bullion expands

(Step 1 in Chart 5.3.1). When, however, the price of domestic gold falls below the gold coin price floor (P Bar), gold flows out of bullion and into the mint to be coined (Step 2 in Chart 5.3.1). Gold will persist in moving into the English domestic market and then into the English mint until the foreign and domestic markets clear at a price equal to or above the gold price floor, or

176 3" 0) 0 c 3' CD D) > O. —"0t o" c" CD T| 3 CD O 3 O «-f- —? ^^ D) 3 a o O. m 3 0) Tl CO co' 3- 8' 3" O. cn CO 0 O CO 3o" CD_ o. ^ 0) C5O' -1 CD 3 00 CO 0) 3 O o TJ -?

O oCD' it CO

177 _ rGoUBuUion Gold Coin UMlm UmlM (3) CM*. •= p s r r » GoldAmunLm GoldAmsurdlm

When the price of credit on Amsterdam was low (£/Sb), gold flowed into London through intermediaries like Evance as will be detailed in Section II.

The domestic silver market was more complicated than that of gold. The price of silver bullion was free to float, but much of England's silver was in the form of coin. The price of silver coin was fixed. England was on a silver standard, and the pound value of an ounce of silver, in coin form, was set by law.'2 The market price of domestic silver coin could not rise or fall with a changing demand for silver on the international market. Melting silver coins into bullion and the subsequent export of the bullion were both illegal; however, once the gap between the value of silver in the domestic bullion market outstripped the unit of account value of coins by enough to compensate for breaking the law, silver was lured from domestic use.

Chart 5.3.2 introduces the fixed price market for English silver coin into the adjustment process following an exchange shock. As with gold, the supply of domestic silver bullion contracts as silver moves into the foreign exchange market. Unlike gold, the price of silver coins, fixed at P

Bar, can not rise in response to the higher domestic bullion prices. A shortage of silver coins develops as the supply of coins contracts from the melting down of coins into bullion for eventual entry into the foreign exchange markets.

12 Silver coinage was the medium of account in England throughout the seventeenth century. The monetary value of silver coins were fixed in terms of the unit of account, the pound sterling. The penny of sterling silver was defined as 7.8 grains troy. Twelve pennies equaled a shilling and twenty shilling equaled a pound (240 pennies). All silver coins - the halfpenny, shilling, half crown, etc. - were proportional in silver content to the penny.(Feavearyear 1963)

178 o

"U £ 3 c D CO 93. 0) 3 3 CD 3 o. «-•• 19 O X —h o CD O. rn CO C3O O a- 3" co" CD 3" en CO 3 co O CO CD cf ro —I

0) 8" —t 3 7T O CD o oT TJ O CD CO

O

179 To bring the price of silver into balance in all three silver markets a tremendous amount of the stock of silver coin would have to fall into the melting pot. England mitigated the collapse of her money supply by adjusting the relative value of the "fixed" price of silver coin. Rather than raise the price of silver coins, Britain unofficially lowered the precious metal content of coined silver. The clippers brought this about. The effect of lowering the weight of a silver coin by clipping while keeping the unit of account value constant was to increase the price of an ounce of silver. The less silver the coins retained, the higher the unit of account price silver bullion needed to threaten demonetization. The effect of clipping on the price of silver coins relative to bullion can be shown in Chart 5.3.2 as the effective price P*. With clipping, any silver bullion priced between a clipped coin's legal price and effective price would not force the coin out of circulation nor draw silver into the mint to become coinage.13

A weakening pound on the exchanges (rising £/Sb), drew silver out of the English economy not by melting silver coins but by rewarding additional clipping. This paper does not attempt to explain why clipped coins circulated at full unit of account value but simply accepts that they did.14 Other things constant, a higher price for shaved silver produced more shaving activity.

Intermediaries such as Stephen Evance connected silver expropriated from English coin with its higher valued use in the foreign exchanges. A weakening pound on the international credit market

13 While silver could leave the coins, silver bullion could not become a clipped coin. Silver coins minted after 1663 were milled with marked edges. Milled edges made clipping obvious. Milled silver coins were treated as bullion and stopped circulating as money immediately after issue. In contrast, the clipped coins circulating under William III were from the Tudor and Early Stuart years when coins were minted by hammering. Hammered edges were not uniform and easily shaved. Clipped hammered coins remained in circulation. Although clipping effectively raised the price of domestic silver coin under the regime of a silver standard, no new silver could join the ranks of the clipped.

14 "Within half a dozen years after the Revolution, there was scarce a single hammered shilling or half Crown to be met with in any sum of money, but what was extremely clipped and diminished. Yet still such pieces were received and passed as current, as when they were of full weight and value; this encouraged the clippers who increased their numbers and their diligence, and vended the money they had clipped in all places without any check or fear from the laws." (Haynes 1700)

180 produced a flow of silver through Evance's shop on Lombard Street. Moreover, arbitrage sped the balancing of exchange rates. Some of the silver arriving from London would be converted into credit and repatriated back to London. Arbitrage raised the supply and lowered the price of credit in

London which eased pressure on English silver.

B. The Course of Exchange Rates

With the three exchange rates models introduced above, a falling credit exchange rate (Sb/£) rewarded the movement of bullion out of domestic English use and into the international exchange markets. Over the course of the war and the period of Stephen Evance's bullion records, the credit exchange rate was weakening. The series of observations of the actual credit exchange rate between the two cities from 1688 to 1690 is incomplete. Chart 5.4 draws the credit exchange rates in

Amsterdam and London as reported by McCusker (1978). In 1687, the pound was very strong by historical standards against the Flemish schellingen. The observations from May to November 1688 show pound sterling declining relative to schellingen banco. The pound strengthened some in 1692 but thereafter the slide of the pound continued throughout the war and reached a bottom in 1696.

The credit exchange rate observations must be viewed with caution. Gaps exist, so only general trends can be drawn for key periods such as 1689 and 1690. The credit exchange rate in

Amsterdam tended to be lower than the rate in London. As de Roover (1974) explained, the difference in rates between Amsterdam and London created an implicit interest rate. Assuming positive interest rates, London's rate would be higher than Amsterdam's. In 1691 there was a substantial gap between London and Amsterdam credit exchange rates representing high short term rates and likely liquidity crises. The rather high June 1689 exchange rate of 35.88 Sb/£ in London may reflect similar high short term rates. Despite the June outlier, the connect-the-dots trend for the pound exchange rate was downwards during 1689 and late 1690.

181 Schellingen Banco/C

182 Chart 5.5 matches the credit exchange rate with the exchange rates for gold. The gold mint exchange rate was 32.61 Sb/£.15 To gauge the market value of gold in London, Chart 5.5 also plots the exchange rate value for guineas from 1685 until 1696.'" The price of guineas has been converted into schellingen banco based on the same fixed Wisselbank value of gold used for the mint exchange rates, so variation in the series results only from changes in the price of guineas in

London. Until 1692, the series remained within a tight band. Guineas started to rise in price (fall as a Sb/£ exchange rate) in 1693 with an acceleration in 1694 and dramatic climb in 1695.

A substantial gap existed between the credit and guinea exchange rates from 1685 through

1688. The guinea price floor prevented guineas from falling below twenty-one shillings and six pence a coin. Consequently, the strong pound in the credit market attracted gold from overseas.

During those four years the English mint processed £2,225,000 worth of gold, a level that easily outproduced any other four year period in the seventeenth century. 17(Craig 1953, p. 416) In Chart

5.5, 1695 also had a large spread between the price of gold and credit, and, as Table 1 shows, 1695 also had a large influx of gold into the English mint. Because English gold coins had a price floor,

15 See the Appendix E.

16 The price series was constructed from transactions in Guineas recorded in the Ledgers of Sir Francis Child, a prominent London goldsmith-banker. See Appendix F.

17 Craig assumed a guinea worth 21 shillings.

183 Schellingen Banco/E

CO CO CO CO PO 4^ O) 03 c J I UJ L 3" CD

0 c 3" CD 2 Q> Q. 0) 3 Q. o 0 g_ 0) o. []

O 3 g" en o. o 3 3 CO o CD

!

> 3 % a. 0) 3 o ro ro CO CO CO CO CO I ro s % 00 0 ro 4*. 0) 00

184 equilibration between gold and credit had to occur by quantity inflow of gold rather than a fall in the price of gold coins.

Table 5.1

Total Gold Minted by Year at the Royal Mint, London

Year 1684 1685 1686 1687 1688 1689

£Gold 319,220 564,204 624,218 421,370 589,375 134,864

1690 1691 1692 1693 1694 1695 1696

51,159 57,222 120,223 54,094 64,780 753,079 145,548

Source: Craig 1953, p. 416.

During the period of Evance's records, 1688 to 1690, the credit exchange rate fell from a

high rate that encouraged the importation of gold into London to a lower level that made market

indifference between gold and credit. The exchange rates before mid-1688 drew gold into domestic

English use by creating arbitrage returns. The importation of gold into England followed by the

exportation of credit (by way of bill of exchange) produced two month returns of 3.6 percent in

May of 1686, 3.1 percent in May of 1687.'8 As yearly rates, these returns to gold/credit arbitrage

18 These calculations assume two percent transportation costs and a 4.5 percent guinea premium. Jones argues the specie point (transportation cost) threshold was around two percent.(Jones 1988, pp. 123-4) Most seventeenth century transportation costs were to mitigate, "risks of shipwreck, piracy, and brigandage. "(Munro 1979, p. 173 Guineas always carried a premium over gold bullion. Because guineas were milled, clipping would be easily detected, so the gold coins carried a premium for known weight and quality. The premium varied but appears to usually have been between two and three percent based on Houghton's prices. Reporting the gold returns as occurring over two month periods assumes that the shipment of gold overseas took a month. This may well be a conservative estimate, but a one month transportation period simplifies comparison with returns on bills of exchange.

185 translate into 24 percent and 20 percent returns respectively. The implicit returns on credit re- exchange over the same period ranged from 8 percent to negative 3 percent at annualized rates.19

The exchange rate stopped attracting gold to London as the price of credit fell to around

34.5 Sb/£. By October of 1688, the credit exchange rate was down to 34.25 Sb/£ and below this threshold. The annualized returns to gold/credit arbitrage fell to 7 percent by June of 1688 and to only 0.2 percent in August of that year. Section II will report a matching diminishment in our goldsmith-banker's importation of gold in 1688.

Unlike gold, the price of silver coins were fixed by England's silver standard. Chart 5.6 presents the credit exchange rate and the mint exchange rate for silver. The silver mint exchange rate was 36.21 Sb/£.20 Again assuming a two percent transportation cost spread on either side, the top of the silver zone was 36.94 Sb/£. This silver exchange rate topped all but the most extreme credit exchange rate observations. By May 1688 the falling credit exchange rate was already below the transportation cost lower bound for mint silver of 35.48 Sb/£. Depending on the premium associated with coined money, a shift from indifference between credit and minted silver to the active exportation of silver was to be expected after late 1688.

The silver mint exchange rate assumes full weight coins.21 The best available measure

19 A negative rate of return on credit re-exchange occurred when the pound was falling quickly in value. A merchant in Amsterdam purchased a bill of exchange on London expecting to return his capital at a higher Sb/£ rate than he paid originally in Amsterdam. If the pound fell over the month between the drawing of the bill in Amsterdam and the bill's acceptance in London, the investor could suffer a worse rate than expected when returning the capital to Amsterdam. For example, in April of 1686, a bill of exchange could be purchased on London (in Amsterdam) at 36.17 Sb/£. In May, when the bill came due, the exchange rate in London (on Amsterdam) had fallen to 36 Sb/£. Repatriating the capital to Holland by bill of exchange would result in a one half percent loss over the two months (3 percent per annum).

20 For the silver exchange rate's derivation, see the Appendix E.

21 Any freshly minted milled coins; however, left circulation, and most entered the melting pot. "For this practice of melting down our silver coin, there is another lure, which, perhaps, the most tender conscience can hardly withstand: and it is when 7, 8 or ten percent may be got in a few hours, by

186 CO 2 13 o Schellingen Banco/2 (D CD CO CO

T3 sp S Ef. o 3 O 3m" L_ a CO X ? ?T C ^ m CO CD c CO S -t 00

[ ] | CO <" E '« CD £3 C- -(aD 3 Q. m c) CO x 5 7 - CD Hc ; -> o 5 3" 0) § 3 5S + (Q >» CO CD z 7 S3• 5 CD i CO 3• I r ) w3 a a { • -

% < m 2 3 ,-»

throwing the milled Crowns and half Crowns into the melting pot, and casting Spanish or foreign bars." (Haynes 1700)

187 CO o Pound Weight C ro ro co G o D Ol 0 C71 0 CJ CD c 1 1 1 _ 1 O): ff 01oo5- LrPr

8= DiP _v- _n 0) = tf 00- » = LHfr f CO : 0) -±: i-TI -2 0D- -O 03 = LrPr CD - a O 0)- d3 z: co­ H in (0= a < - £si 0) 0; r sT a 169 1

(O IIIII I n f° 7T - (D _u,: D o)E J ITI CD; Ef 0) iv> = • 169 3 11111I I I 3 Ju l O

^ *< : -*- =

4&.: dlP 1 f

§ f / , —1 1— CO cn 8 % 8 cn

188 suggests circulating coins were already under ninety percent of original mint weight by 1685. Chart

5.7 compares the weight of bags collected at the Excise for tax payments.22 The samples were taken in July of each year. The weights of the bags indicated a lightening of the coinage as early as the summer of 1689. The summer of 1690 brought an acceleration of content loss. Weight loss continued until the Great Recoinage in 1696.

Chart 5.6 converts the average annual bag weight in Chart 5.7 into a clipped coin exchange rate.23 The collapse of the content of the coinage over the course of the war is clear. While the credit exchange rate fell during the war, silver's exchange rate fell faster, at least until 1695. The silver released from the shrinking coins would have acted to buoy the exchange rates of gold and credit through an increasing supply of English silver bullion.

Chart 5.6 also plots a few observations of the silver bullion exchange rate based on London bullion prices reported in McCusker (1978). The silver bullion exchange rates suggest that the supply of silver from clippings kept English silver on a par with, if not stronger than, gold and credit in the exchange markets. The falling credit exchange rates were never close to overcoming

the plummeting exchange value of the bullion remaining in clipped coins. Rather than forcing them

out of circulation, the rising price of international credit was luring silver from hammered coins.

With the weakening pound, clipping began to renew itself in 1688, and Stephen Evance was a key juncture in silver's journey from domestic English circulation to the international bullion markets.

22 Over the war years, counterfeiting was an active problem. Measurement by weight could not differentiate between clipping and counterfeiting.

23 The months between the July observations have been constructed as a weighted average of the two bounding observations.

189 II. The Bullion Book of Stephen Evance

From 1688 to 1690, the falling credit exchange rate should have changed the behavior of

London bullion dealers. Did the procurement of gold bullion trail off in 1688? Did the export of silver increase in 1689 and 1690? The bullion business of goldsmith-banker Stephen Evance offers a glimpse into the activities of both the banker and his larger customers during a critical period in

English monetary and financial history. The transactions recorded in one of Evance's bullion books agree with the expectations developed in Section I. The London goldsmith-banker's records demonstrate that Evance and his customers were importing gold when the pound was strong in foreign exchange and exporting silver as the pound weakened.

Bullion Book #5 of Stephen Evance stretched from March 1688 to August 1690.24 In the book Evance recorded his transactions involving gold, silver, plate and a variety of fashioned pieces.

The book also noted fees for services both supplied and demanded by Evance along with his household transactions. Entries were separated into purchases (paid) and sales (received). Typical entries listed the date, a pound sterling value and a description. Sometimes the description included the name of a supplier or customer. Some named entries have been used below to construct the accounts of individual customers.

Stephen Evance was a goldsmith-banker for whom bullion was only a portion of his business.25 Over this period Evance entered the employment of the Excise and loaned funds to the

24 The book does not name its owner (PRO CI 14/179). Evance's ownership was suspected by David Mitchell at the Centre of Metropolitan History based on tax records. The Calendar of Treasury, vol IX, pt 2, matched Evance's loans to the Exchequer with silver transactions in the book in July 1689. The Bullion Book was used as evidence in an unknown legal case in Chancery and has remained there to this day.

25 Stephen Evance was in partnership with Peter Percival at the Blackboy in Lombard Street by 1677 and until at least 1697. By 1702, Evance had taken a new partner, William Hales. Evance suspended payments in 1721. Heal (1935), Hilton-Price (1890) Evance was apprenticed to Henry Nelthorpe, also a banker, from 1669 to 1676.(Goldsmiths Company)

190 Crown. In the war years that followed, Evance became an important financier to the Crown in partnership with Joseph Heme. From May 1691 until October 1694 (now) Sir Stephen Evance and

Sir Joseph Heme were the center of a syndicate that contracted to transfer funds to the Low

Countries. The goldsmith-banker was Governor of the Hudson's Bay Company from 1692 to 1695, a large stock holder and a major purchaser of the company's furs.(Carlos and Van Stone 1994)

Additionally, Evance referred to a Cash Book while some bullion transactions were conducted, "on account." Evance would certainly have had a banking ledger for his deposit customers. Evance even issued his own bank notes.26 That the bullion history of Evance was but a portion of his overall banking business must be kept in mind.

Bullion Book #5 is silent on the form of payment for transactions. Each entry has an amount in pounds sterling that Evance either "paid for" or "received for" bullion, plate, services, rent, wine, etc. Evance could have been paying and receiving domestic coin (full or clipped), bills of exchange, bankers' notes (his own and others), credit on his own ledgers or checks payable on the ledgers of others. Evance only recorded the unit of account value: pound sterling.

Chart 5.8 gives Evance's running total of gold, silver, plate and "gold & silver" purchases by month. "Gold & silver" were entries of both metals that could not be separated. Evance's purchases of gold outpaced purchases of silver until September of 1688. After September of that year, gold purchases slowed while silver purchases accelerated. Total silver purchases surpassed gold in the summer of 1690 with both totals ending in the neighborhood of £225,000 in August.

The "gold & silver" series remained unremarkable at about fifteen percent of either of the larger series. Purchases of plate proved very small relative to the other series.

26 The Auditor of the Receipt at the Exchequer was ordered to accept the notes of Evance along with fellow goldsmith-bankers Percival, Atwell, Courtney, St. John, Child and Fowles.(CTB IX, p 123) Evance also had numerous credit clearing transactions with Sir Francis Child to be found in Child's ledgers.

191 Pound Sterling CO o (Thousands) C

Q.

o. Q. 0> 3 Q. CO cf

*

(Thousands)

192 Evance's sales tell a very different story. In Chart 5.9, sales of silver reach a similar magnitude as purchases of silver. In contrast, gold sales are tiny. Clearly, Evance was not selling gold at the rate he was acquiring gold. Like gold, the "gold & silver" sales series shrank relative to its purchases. Sales of plate matched those of gold, but all three sales series were small relative to silver's.

Chart 5.10 compares the accumulated stock of each metal, purchases minus sales to date.

Evance's initial stock of bullion was not reported, so Chart 5.10 begins with no inherited stocks.

The accumulation of gold by Evance was substantial. The rate of collecting gold clearly slowed after September of 1688. A lesser change also occurred after the following September in 1689.

From March to September 1688, Evance had accumulated £147,095 of gold. In the year ending

September 1689, Evance's stock of gold increased by an additional £48,000. The next eleven months ending August 1690 brought only £23,100 of new gold into Evance's stock.

Chart 5.10 contrasts gold with the accumulating stock of the other metals. Evance did add silver to his bullion stock, but the silver plateau in the neighborhood of £25,000 seems modest compared to gold's mid-1690 level of £225,000. Evance built most of his silver stock in 1689. The spike in the summer of 1689 was created by a £24,000 loan of pieces of eight and an £8,700 loan, also in foreign coin, to the Crown in July of that year.(CTB IX, pp. 40, 44) Although Evance shipped £32,700 worth of foreign coin out of his stock, he purchased £22,500 worth of new silver that same month.

The low stock of silver relative to gold reveals the movement of silver out of, as well as into, the goldsmith-banker's possession. Evance sold little gold. Evance also sold little "gold & silver". This hybrid retained a large stock, as large as silver's by 1690. The retention of "gold & silver" suggests behavior closer to gold than silver. The stock of "gold & silver", however, must be

193 Pound Sterling CO 2 (Thousands)

CO i i1 s 3 OL O CD of O

2 0) 3 Q. tW 8

(Thousands)

194 CO Pounds Sterling 8 (Thousands) 8 m

(Thousands)

195 viewed with great caution. We have no way to know if the mix of "gold & silver" purchased was the same as the mix of "gold & silver" sold.

Evance did sell more plate than he purchased. Perhaps some portion of the gold and silver stocks went into plate. Regardless, any level of metal converted into plate was small relative to the precious metal stocks.

The gold side of Evance's bullion record fits Section I's expectations. Evance's acquisition of gold in 1688 was very large. In the seventeenth century, £150,000 was a fortune, so Evance's collecting of gold was surely intentional. The deceleration of gold stock growth fits with the argument above that the falling credit exchange rate choked off gold imports.

Evance's accumulated gold because the banker had any number of domestic uses for the precious metal. The author suspects, however, that the goldsmith-banker may have retained large amounts of the gold as reserves. Another goldsmith-banker, Sir Francis Child was keeping a fifty to sixty percent reserve ratio during these years.(Chapter 3) The author has no measure of the total size of Evance's business, but, if Evance kept reserve levels similar to Child's, our goldsmith-banker would have held a considerable stock of bullion.

If Evance was a final user of gold when the yellow metal was drawn into England, the goldsmith-banker acted as an intermediary when silver left England. To better assess goldsmith- banker's role as a middleman in the international bullion market, a few important customer accounts have been assembled. Many of Evance's larger gold suppliers were active in international markets and experienced in specie shipment. Moreover, separating Evance's customers reveals that the banker was acting as an intermediary in the silver market as well as in his freight business.27 As the credit exchange rate fell, Evance purchased silver from one set of customers and sold silver to

27 Evance's freight activities are set out in Appendix G. While indicating that the goldsmith-banker did participate in international shipments, the volume of the freight entries cannot account for the bulk of Evance's bullion business.

196 another. Those buying silver were often the same merchants selling Evance gold when the pound was strong on the credit exchange.

Evance's largest customers, as individuals and certainly as a group, were Sephardi Jews.

Sephardi Jews were important merchants in the Atlantic trade. The Spanish and Caribbean trades brought the Sephardi merchants into the bullion market. From Cadiz, the Amsterdam Jews shipped

European linens and woolens and East India spices to the West. The trade returned silver, tobacco, dye-stuffs, vanilla and jewels.(Israel 1983, p. 530) The Sephardi merchants participated in the international bullion market centered in Amsterdam.28

The Jewish community in London had very close contacts with the Sephardim in

Amsterdam. Large families had members in both cities and migration between the cities was common. William III also bound the communities together. Both Amsterdam and London Jews supported William and acted as his agents. "William III employed the firm of Machado and Pereira

[Sephardi of Amsterdam] as the main suppliers of grain, horses and munitions to his armies fighting the French and retained their services as his principal provisioned down to his death in 1702."(Israel

1983, p. 533) Alfonso Rodriguez (Isaac Israel de Sequeira), a London Jew, participated in the provisioning of William III in cooperation with Isaac Pereira of the Dutch firm Machado and

Pereira.(Hyamson 1951, p. 68)

Alfonso Rodriguez, and his father Gomes Rodriguez, had substantial transactions with

Evance. A number of other members of London's Sephardi community also bought and sold bullion through the goldsmith-banker. Evance's business with London's Sephardi merchants started

28 "In 1685, of the four million pesos shipped directly to northern Europe on the return of the New World fleet, three-quarters was consigned to Holland and much of the silver remitted to Italy went to Livomo on account of Dutch firms trading with the Levant; of the bullion brought back on the 1698 fleet, four million pesos was consigned to Holland, less than half of this to France and only 600,000 pesos to England which , before 1713, never succeeded in rivalling France or the United Provinces in this key traffic."(Israel 1983, p. 530)

197 with the beginning of Evance's fifth bullion book in March 1688 and likely predated it. In the summer of 1689, Evance made large loans to the new King. By 1691, Evance and Heme led a remittance syndicate for the King's troops. The goldsmith-banker's dealings with the Sephardi fit into a pattern of financial support for William III by both the banker and his customers.

Four Sephardi accounts, David Penso, Jacob Gabay, Gomes and Alfonso Rodriguez and

Joseph Beuno and Peter Henriques, have been constructed. They behaved very similarly. All four were leading members of the synagogue on Creechurch Lane, London, and undoubtedly knew each other.29 Chart 5.11 and Charts 5.11.1-4 presented the accumulated stock of their joint and individual accounts respectively. All sold Evance large quantities of gold in 1688. As a group the four sold Evance, net a few purchases, £45,468 worth of gold or thirty percent of the goldsmith's accumulation in 1688. They sold Evance virtually no gold after 1688. Taken together, the accounts of these merchants represent an important supply of gold that ended in 1688. Moreover, the change in Sephardi behavior matched the beginning of Evance's overall slowdown in gold acquisition.

In sharp contrast, the Sephardi Jews began buying silver in 1689 and continued into 1690.

By September 1690, they had purchased £61,102 worth of silver from Evance. These four names account for about thirty percent of Evance's silver sales. The silver purchases of the four Sephardi merchants corresponded with the expected exportation of silver as the credit exchange rate fell.

29 Gomes Rodriguez (Abraham Israel de Sequeira) was elected a member of the Mahamad or ruling council of the London synagogue in 1672 and 1678. His son Alfonso Rodriguez was elected to the Mahamad in 1697, 1699-1700 and 1704. Jacob Haim Gabay was elected to the Mahamad in 1707 and 1711 and to the treasurership of the synagogue ( this office being called the Gabay) in 1703. David Penso (alias Alexander Felix), a Dublin Jew, was elected Gabay in 1689-90 and to the Mahamad in 1699-1700. Jacob Beuno Henriques had a prominent place in the synagogue seating arrangement recorded in 1682. P Henriques was a signatory to the civil laws of the congregation in 1677, but Evance sometimes uses P Henriques Junior. (Hyamson 1953, pp. 68, 422, 423, 427-429)

198 Pound Sterling (Thousands)

(Thousands)

199 Pou nd Sterling CO (Thousands) 51 o ^ i § CD n o oi o en o c ^ ii ii s E 1 S =r- \ ! i CD C/)~ o oo- / \ i _CD 00- Z CD Tj - **< «-*• XI _ N. CO O ct c i s- o C - ; ^ =r DO *< " :i o o : 3 CO f § O 0) 'J o PJ_ _ •D 3- CD - 3 C(DO Ol / £ o i 0) a. 3 C_- y' 3 3- §_ ^/^ Q. c _ 0) > —H o CO r. o CQ c - CO" a- i l Q_JL) - i i n A ^ " " o c; sc (Th ousjands )

200 Pound Sterling CO c (Thousands) o ?, JCKD m^ —' L —' L . -k - m 8°i o oi o en o cn < 3"- II II OJ L

ice , 68 8 \ 3 - - \. 1 CO / •--. m <" -o / \ < m J3 0) 3 o - O c CD s- V) c - Z • o

168 9 a

Meta l 3" 1 / CD • a. o 5 i CD —K CJ1 —1 / —\ tZ ' f o g ( • 3 C__ \ 3 C_ 3- g_ / CD ro c _ • o •.- Si- o 0) XT g. SJ i o CO o- CD >- f C - cr ^ a-

O) i i I I w a. §5 s * ° s c n (TrlOU ! sands)

201 Pound Sterling

(Thousands)

202 C/D <-> Pound Sterling C o (Thousands) CD __ •• s. < ^ oi o en o oi o c n g §-. 1 1 i i O _L_ » I : 1 _CD 0)_ N\ 1 CO "0 00-

£ O / 'N. ? ~* : x Z3 \ O

C i CD 5-- co" C - Z i DO *< " %

i§: % PJ_ . i 7o7 "O CD C/) —i ^ " J • 3 s c_: / • 3 _T §_ o 9 c _ . ' g Q. .-* *< to' _JL~ CL 3 OL 8" / TJ CO o- / CD f 3 CQ 8 c -

0) 1 1 i i (0 ' n' i iiO Ol -k - o« o w ° « 1 (Thou;sands )

203 Were these Jewish merchants involved in import and export? A House of Commons committee blamed Jewish merchants for silver exports. The committee was replying to a petition made by the working goldsmiths of London. On May 7, 1690, the committee reported,

That it appears, by a Certificate from the Custom House, dated seventeenth April last, That great quantities of Silver have been of late exported; whereof we had a particular Account for the last Five years: That above Seven Parts of Eight had been shipped off by the Jews, who do any thing for their Profit.30

Some detailed evidence exists in London of the overseas character of Evance's Jewish customers. In 1695, Peter Henriques sold Sir Francis Child two Bills of Exchange on

Amsterdam.31 Gomes Rodriguez, with his son Alfonso, had an earlier history of international dealings in specie and bills of exchange. Gomes Rodriguez had a running cash account with a

Lombard Street banker named Edward Backwell that portrays Rodriguez as a very busy merchant.32 Between May 1670 and March 1671, Rodriguez cleared £48,800 worth of transactions through his banker. Rodriguez bought (£1261) and sold (£3508) foreign specie composed of Pieces of Eight, Rix Dollars, Ducats, Pistol Is and gold bullion. Rodriguez provided Backwell with Bills of

Exchange on Amsterdam and Antwerp. The merchant paid the East India Company £2653 and had dealings with Sephardi Jews and other important merchants.33

Some Low Country evidence of bullion shipment also exists. The Amsterdam Exchange

Bank was an important part of the Dutch bullion market and much of the bank's records are still extant. Arbitrage returns required only Wisselbank prices in the Amsterdam bullion market.

30 The Journal of the House of Commons, reproduced in Li (1963), p. 55.

31 The two Bills of Exchange were for £300 and £200 at 31 Sb/£.(Child's Ledger CH/194/11)

32 Samples of Gomes Rodriguez's account were taken from Alderman Edward Backwell's Ledger "S", folios 40, 303, 406, 412, 551, 583 and 589.(BackwelI's Ledger EB/8)

33 A sample of Rodriguez's associates includes Manuel Lopes Pereira, Francis DeLiz, Simon Henriques, Joseph DeSilva, John Lethuillier, the firm of Burkin and Everson and Richard Mounteney, Receiver of the Customs for the Port of London.

204 Dealers did not have to deposit specie in the bank, so bullion shipments through Amsterdam need not appear in the Wisselbank's records. Moreover, the common use of aliases and foreign agents complicated the linking of Evance's customers with Amsterdam's bullion market. The Amsterdam

Exchange Bank's accounts are full of names similar to Evance's Jewish customers; however, the author was unable to find a match.34

Two other of Evance's bullion customers, however, were found with accounts at the

Amsterdam Exchange Bank. Alexander Henderson and the firm of Boddens & Rumswickle had active accounts at the Wisselbank concurrent with silver purchases in London from Evance (see

Chart 5.12). While Boddens & Rumswickle left no trace in their Wisselbank accounts of a business in precious metals, Henderson did.

Between November 1689 and August 1690, Alexander Henderson purchased £11,500 worth of silver from Stephen Evance in harmony with possible arbitrage behavior. Over the same period,

February 1689 until August of 1690, Henderson supplied the Wisselbank with £7,500 in specie.35

The bank's records do not distinguish between gold or silver nor does the £7,500 match the £11,500 worth of silver Henderson purchased from Evance in London. However, Henderson was clearly active in both the London and Amsterdam bullion markets; he purchased silver in London and deposited specie in Amsterdam.

The behavior of other named entries differed greatly from the merchants mentioned above.

Chart 5.13 demonstrates that Evance's named business, other than the merchants examined above and the Exchequer, consistently supplied the goldsmith-banker with both gold and silver. Evance's loans of £35,000 to the Exchequer in July of 1689 are omitted because Evance loaned the Crown

34 Names related to Evance's Jewish customers, such as David Gabay Henriques, Abraham Penso Felix and Abraham Beuno Henriques, were in abundance.

35 79,583 Florins Banco exchanged at 35 Sb/£.

205 Pound Sterling CO (Thousands) 2 o

3 o O _(D CD "0 CO* X) o Za CO

I^ s9) iCO 3 P, > Ol CD X CD 3 IV) Q. CD -i

CD 3 Q_ CD —i CO o3

(Thousands)

206 CO Pound Sterling o c (Thousands) O m ® -»• ro CO 4^ Cjri O) ^sj < oooooooo (D § I 1 I I 1 1_ 3 3 3M o L O O r-t- 05 3" CD P 00" CD (f) -rj 00- —t 00 <-* z O _) CD U) % 3 c o 00 co 39 3" —c C _ 0) 00 *< 3, 3o CO CO o r-t- en () O £2. 7T CO CD en —h CJ1 o_ ro s Q. CD 3 to 3 3 CO Q. z C . CD CO _, PJ ar-h. 3" 3 r CD CD Q_ 5? O —? > CD c. 3" 3 CQ CD c • CO CD CO . _La_ CD n —I 1 1 1 1 1 1— —i -J- N> CO -fc> Ol O) ^| O=r . o o o o o o o 3 h CO (Thousands)

207 CO E Pound Sterling O CD (Thousands) -^ 3 &) O

Q.

O o_ Q. to 3 OL CO $

(Thousands)

208 foreign silver coins to aid William Ill's military expenditures. The loans were clearly for international payments.

Bullion entries without names also supplied Evance with gold and, from 1689 on, with silver

(see Chart 5.14). The transactions without names tended to be much smaller than the named ones.

Table 5.3 separates named and unnamed entries by the average transaction value and the number of transactions for each metal. Roughly seventy percent of the bullion entries lack a name in the description. Certainly some bias existed towards not bothering to record the customer of a small transaction. Yet, many small transactions were named and entries worth thousands of pounds went anonymous. The smaller, unnamed transactions account for about half the pound value of Evance's bullion purchases. Larger, named transactions represent two-thirds of the goldsmith's silver sales.

A pattern of Evance acting as a silver middleman is evident. Evance bought silver from the many and sold silver to the few. The merchants purchasing silver did so when the credit exchange rate fell. The evidence is consistent with the theoretical expectations mapped out in Section I. The

large customers were worldly merchants capable of exporting precious metal. Some, such as Gomes

Rodriguez and Alexander Henderson, have known histories in the international bullion trade.

209 Table 5.3

Stephen Evance's Bullion Transactions by Metal and Named Entry

Sales Silver Named Unnamed Total Avg£ 808 265 478 Sum £ 142,253 72,637 214,890 Obs 176 274 450

Gold Avg£ 239 70 103 Sum £ 2,638 3,260 5,898 Obs 11 46 57

Gold & Silver Avg£ 640 29 643 Sum £ 5,762 29 5,791 Obs 8 1 9

Purchases Silver Named Unnamed Total Avg£ 306 124 171 Sum £ 106,643 124,058 230,701 Obs 348 998 1346

Gold Avg£ 463 117 262 Sum £ 119,097 107,960 227,057 Obs 257 609 866

Gold & Silver Avg£ 492 97 167 Sum£ 15,749 14,562 30,311 Obs 32 150 182

210 III. The Political Economy of War Remittances

In War and Economy, D.W. Jones demonstrated that the massive export of silver from

England allowed William Ill's government to avert a financial crisis during the early years of the

War of the League of Augsburg. The exodus of silver from clipped coins balanced Britain's current account in a less disruptive way than other possible adjustments. The process of silver export, however, was not cost free. Beyond physical transportation, transaction costs had to be overcome and risk allocated.

The expense of moving purchasing power to the Continent was important because the

Crown had to pay for it, either directly or indirectly. The more expeditious the transfer of funds, the more money William III had for other purposes. From 1688 until May 1696, the Crown contracted out the transfer of funds by six month periods with payments to the forces in current

Flemish money being recompensed in London at a fixed exchange rate. "If market rates proved unexpectedly favourable, the contractors stood to gain; if the opposite proved the case, in theory they would have to bear the loss."(Jones 1988, p. 84) The Treasury sought remittance agents that could acquire bills of exchange on the Low Countries at competitive rates while being low cost bearers of exchange rate risk.

During the period of war covered by Evance's Bullion Book #5, the English relied on the

Dutch Army's paymaster, William van Schuylenberg, to draw bills on London. From 1689 to 1691,

Van Schuylenberg charged a commission and eventually detained funds in the Low Countries, "to answer interest and loss on [such of] his said bills not being duly [in proper time] paid."(CTB IX, pp. 1521-2) In May of 1691, Joseph Heme, a successful merchant, and Stephen Evance offered to transfer the remittances, "without demanding anything for commission money and himself [Heme] to be at all charge and hazard."(CTB IX, p. 1152) Besides not requiring a commission, Heme and

211 Evance agreed to accept repayment in the form of government debt.36 The Lords of the Treasury accepted Heme's and Evance's offer, and the syndicate transferred funds from May 1691 through

October 1694.

The Treasury certainly benefited from the Heme and Evance syndicate, but the goldsmith- banker and his merchant partner did too. If we accept the conclusion (Section II above) that Evance was selling silver for export then the banker could accept bills on Amsterdam as payment for the silver. Evance was likely already accepting credit, since arbitrage gains were created from buying silver with credit. The silver sales created the bills of exchange on the Low Countries that the syndicate needed to meet its obligations to the Treasury.

To the degree that Evance split arbitrage returns with his export customers, the goldsmith- banker profited by accepting credit on the Continent from his customers.37 The goldsmith-banker was in a position to claim a portion of the arbitrage returns. First, Evance had the silver. Bullion clipped from silver coins from all over England was channeled into Lombard Street.38 The

Goldsmiths' Company was a national organization, and the regional goldsmiths recast the silver and,

36 Heme offered, "the constant return [forwarding by bill of exchange] of weekly subsistence for the [English] Forces in Flanders and the monthly payment of £3,600 for the Wolfenbuttel Troops [Heme furnishing bills] upon credit of tallies." (CTB IX, p. 1152, emphasis added)

37 Evance's earlier Bullion Book #5 does not detail the quantity of silver the goldsmith-banker offered per sale, just the price. Evance might have been selling the silver at an above market price (below market quantity). Evance makes no mention of commissions, but he may have written such charges in his ledger or other accounts.

38 "It was organized on a three-sided basis involving the clippers with their skills and instruments; the goldsmiths of such centres as Manchester, Pontefract, Leeds Wakefield, Halifax and York who, with a float of cash (including clipped coin) always on hand, were prepared to buy molten clippings as bona- fide bullion, but at rates of between 4s. and 4s. 6d. per ounce as compared with the market rates of above 5s. 2d. per ounce; and finally, all manner of dealers such as clothiers, dyers, drovers, graziers and even clergymen who, receiving large cash amounts, could cull their receipts for the heavier coin most suitable to be clipped." (Jones 1988, p. 229)

212 "sent it to London as plate melted down."39 Evance's bullion book records a large number of

silver purchases. Although many of the goldsmith-banker's silver suppliers sold in small amounts

(see Table 5.3 above), some of Evance's unnamed silver purchases are quite large with many over

£100 and a few over £1,000.40 Evance and the other goldsmith-bankers on Lombard Street were

the end of this domestic chain of transactions and the gateway to the international markets.

Besides having silver, Evance had a gate keeping position in the process of silver export.

Legal export of bullion required assurance that the precious metal was of foreign origin. The

Goldsmiths' Company had a responsibility to verify that bullion to be exported had first arrived in

England from abroad.4' Goldsmith-bankers such as Evance were important members of the

Goldsmiths' Company and had the power to certify bullion was of foreign origin at the London

Customs House. In fact, Stephen Evance was the Prime Warden (the highest officer) of the

Goldsmiths' Company in 1692. Even those seeking to stop the flow of silver out of England

realized that domestic English silver was, "being melted down, and Shipped Off, . . ., under the

Pretence of Foreign Bullion."(Houghton 1695, p. 28)

Evance's agreement with the Treasury complemented the existing arbitrage behavior of the goldsmith-banker and his customers. English silver was traded for credit on the Low Countries. In turn, the Heme and Evance syndicate could use the bills of exchange, denominated in Flemish units of account, for the Army's support in Flanders. The circle was closed when the Crown repaid

39 PRO Assi 45/16/1/5, quoted in Jones (1988), p. 231.

40 Evance had an interest in not leaving a record of potentially illegal silver purchases. The Bullion Book #5's existence as evidence in Chancery Court bears witness that the goldsmith-banker's accounts were evidentiary. The author has been unable to match any of Evance's named customers to regional goldsmiths.

41 From May 1, 1695, no bullion could be legally exported, "but only as shall be marked and stamped at Goldsmiths-Hall by the Wardens, ... to be Lawful Silver."(6&7 WIII.C 17 Statutes, V. 9, p. 368)

213 Evance for credit provided on the Continent with pound denominated debt in London. Effectively,

Evance converted silver into government debt while clearing his portion of the returns to arbitrage.

Moreover, Stephen Evance appeared content to accept government debt. Before the formation of the Heme syndicate, Evance had already loaned the Exchequer substantial amounts.

From July 1689 to March 1690, Stephen Evance loaned £87,048 on the extra-ordinary revenues being raised for the war.42(CTB IX, pp. 1971-2008) While Evance may not have gotten a better interest rate than the general public (six or seven percent), the goldsmith did gain Royal patronage with an officership in the Excise and the jewelership to the King. At a time of monetary stringency, the Crown benefited by being able to pay Evance in debt and favors.

From both the Treasury's and Evance's perspective, the Heme and Evance syndicate was an improvement over the use of the Dutch paymaster William van Schuylenberg. The syndicate charged no fees, shouldered exchange rate risk and accepted debt as payment. For Evance, the goldsmith banker could pocket his share of arbitrage returns from swapping silver for credit and then convert the credit into the final form he desired, government debt. Moreover, the syndicate could assess risk (exchange and default) as well as anyone and spread the risk amongst the members. Joseph Heme and fellow syndicate member William Scawen were merchants of the largest size and knew the international markets well. Evance's Sephardi customers (see Section II) traded and prayed with the Amsterdam Jews that supplied William Ill's troops. Once made an officer of the Excise, Evance had a first hand knowledge of an important public revenue stream and potential government default.

Though not obliged, the Treasury compensated transfer agents for exchange rate losses. Van

Schuylenberg contracted at 11 gilders current to the pound or 35 Sb/£ and received full

42 This figure covers loans on the Present Aid, the 12d Aid, the 2s Aid, the Additional 12d Aid and the First Poll.

214 compensation by holding on to funds.43(CTB, IX, p. 1521) The Treasury offered 10 gilders and 10

stivers current to the pound sterling (33.4 Sb/£) to Heme and Evance in 1692, and the syndicate

received some compensation.(CTB IX, p. 1426) The Bank of England, however, truly suffered

with the collapse of the pound sterling in 1695.(see Chart 5.4) Having taken over from the Heme

and Evance syndicate in October of 1694, the Bank of England suffered huge losses and borrowed

£67,000 from William van Schuylenberg and £200,000 from the Dutch Estates General in 1695 and

another £64,000 to pay back van Schuylenberg.44 When £200,000 in new bills of exchange on

London fell due in the summer of 1696 contemporaneously with the liquidity crisis caused by the

Great Recoinage, the Bank of England defaulted.(Jones 1988, p. 24) It is unlikely any private

syndicate could have weathered such a fall in exchange rates.

In 1696, the transfer of England's military remittances returned to private hands. The transfers of funds for the rest of the Nine Years War and the later War of the Spanish Succession were, "handled by a succession of London-based syndicates of merchants, goldsmith bankers, and

Huguenot and Jewish financiers."(Jones 1988, pp. 83-4) Heme and Evance were the beginning of an agency structure of funds transfer that lasted for two decades.

IV. Conclusion

Stephen Evance was an important goldsmith-banker during the Glorious Revolution. At his shop on Lombard Street, Evance stockpiled a large quantity of gold in 1688 and added gold by diminishing amounts thereafter. As a silver dealer, Evance bought silver from a broad range of customers and sold to a more select clientele. Half of Evance's silver sales were accounted for by

43 Exchange rate assumes an agio of 4.5 percent.

44 Half of the Estates General's loan was repaid the next month, November 1695. (Jones 1988, p. 21)

215 large loans to the Crown and large transactions with a few merchants dealing with the Low

Countries. The 1689 and 1690 silver purchases of a few Sephardic merchants were simultaneous with the beginnings of a silver outflow from Britain. Evance and his international partners switched from importing gold to exporting silver with the change in arbitrage opportunities. Evance's expertise and network of international contacts made the goldsmith-banker a central player in the financing of England's war with France. King William Ill's troops demanded payment. Clippers provided the silver. Goldsmith-bankers and international bullion dealers transported the silver and balanced the foreign exchanges of Europe.

Stephen Evance's exportation of silver from London as the credit exchange rate weakened agrees with this chapter's vision of the relationship between the international gold, silver and credit markets. The recoinage crisis of 1696 was a product of the actions taken by international players like Evance over the prior years. By the end of 1695 and years of recurrent credit shocks, the price of gold guineas had risen fifty percent and the average content of circulating coins was less than half the original mint weight. The Great Recoinage of 1696 did two things. Silver coins were reminted back to full weight and not allowed to fall in weight again. Also, a ceiling price for gold was imposed. Both domestic gold and silver were stripped of their ability to adjust to international pressures except by demonetization and export. Because England fixed her gold to silver ratio higher than France or Holland did, future credit shocks that weakened (strengthened) the pound exchange rate drew silver out of (gold into) Britain. Although England was legally on a silver standard after 1696, the interplay of gold, silver and credit was moving England inexorably towards gold.

216 Chapter 6

Conclusions

Goldsmith-bankers created a web of intermediation that linked the monetary and financial

institutions of late seventeenth century England. Branching from the core banking services of delegated lending and the issuing of debt as a medium of exchange, individual goldsmith-bankers organized into a system of clearing, monitoring and occasionally assistance. Some innovative goldsmiths also specialized in bridging financial elements of England's public sector (taxes, government debt and war finance) to the private banking system. Still other forms of goldsmith- banker intermediation opened England's bullion markets to the world. The breadth and sophistication of goldsmith-banking meant that early modem London's financial system was more

integrated before the Glorious Revolution of 1688 or the founding of the Bank of England in 1694 than previously understood.

By the Restoration of Charles II in 1660, the circulating debt of goldsmith-bankers was

common in London. Debt issued by or written on goldsmith-bankers facilitated trade by avoiding the transactions costs associated with specie. Deposits with goldsmith-bankers provided liquid

investments in an age when few marketable assets existed. Moreover, transferability enhanced the

efficacy of debt as a substitute for gold and silver, and goldsmith-bankers were major suppliers of

notes, bills and checks payable to bearer or to order. While endorsement allowed goldsmith-banker

debt to pass when parties had limited information, debt payable to bearer (currency) minimized

transactions costs when all parties trusted the issuing goldsmith. The notes, bills and checks of

217 goldsmith-bankers provided essential intermediation that aided London's commerce in overcoming the hazards of dealing with the low quality coinage of the day.

Goldsmith-bankers increased the usefulness of notes and checks by offering intermediation between the public and other bankers. By accepting rivals' debt, goldsmith-bankers exchanged their own debt for that of competitors while creating a network externality. Mutual debt acceptance by goldsmith-bankers improved the attractiveness of all bank debt, and each goldsmith benefited by his colleagues' actions. While mutual acceptance among goldsmith-bankers was a form of intermediation that complemented transferability in minimizing the public's recourse to specie, goldsmiths accepting each others' debts still had to settle debts with each other. Like everyone else, goldsmith-bankers appreciated not moving gold and silver, so a system of retained precautionary debt reserves developed. By building up uncleared debits, goldsmith-bankers only needed to clear net balances. Holding competitors' debts, however, created risk that increased with time, so clearing occurred frequently enough to monitor net positions for abuse. Moreover, small net positions could be rolled over into new reserves for the next round of clearing, so only systematic imbalances had to be dealt with in gold and silver.

Besides improving the circulation of bank debt, the inter-banker clearing of debt that followed from mutual debt acceptance between bankers assisted goldsmith-banking in dealing with the problems of delegated lending. Like issuing debt as a medium of exchange, intermediation between depositors and borrowers was a fundamental service of goldsmith-banking, but one that created asymmetric information, fractional reserves and the potential for runs. From the first instance, demandable debt complemented delegated lending because the potential for runs allowed relatively uninformed depositors to monitor goldsmith-bankers. The additional layer of intermediation that supported the goldsmith-bankers' sophisticated system of inter-banker debt clearing permitted goldsmiths to monitor their rivals for insolvency and assist those rivals suffering

218 from potentially contagious runs. A system of clearing that improved the circulation of debt as a medium of exchange could also improve goldsmith-bankers' ability to maintain asymmetric

information between depositors and bankers.

The most celebrated form of goldsmith-banker delegated lending was bank purchases of

Treasury debt. Goldsmith-bankers were lured to the extraordinary returns available from government debt (called tallies). By collecting deposits, goldsmith-bankers amassed commanding sums which they invested in tallies, especially when wars increased both the Treasury's demand for funds and the interest rates offered. The threat of default, however, was real. For instance, the

Treasury did stop payments in 1672. To protect themselves, goldsmith-bankers maintained substantial fractional reserves by modem standards. The records of Sir Francis Child suggest a reserves to assets ratio of fifty to sixty percent.(Chapter 4) With their intermediation, goldsmith- bankers supplied the Treasury with needed funds but balanced these advances with a fear of default.

Goldsmith-bankers, moreover, were able to link specialized lending to the crown with the supply of debt as a medium of exchange to open the process of tax collection. As with other transactions, the avoidance of specie saved transaction costs in the payment of taxes. Unlike other activities, revenue collectors could demand payment in specie and so themselves avoid the danger of a debt-issuing banker defaulting. Goldsmith-bankers reversed the situation by acquiring Treasury debt (tallies) that tax collectors desired. Because of a peculiarity in how the Treasury accounted for advances on the public revenues, tax collectors paid debt holders directly. When goldsmith-bankers held the very tallies a tax collector had to settle out of the revenue stream, the risk to the tax collector from accepting that banker's debt was eliminated. The tax collector and the goldsmith- banker could swap tallies for bank notes which spared the tax collector from moving specie and opened an important avenue of circulation to the banker's debt. By combining two forms of intermediation, goldsmith-bankers eased the process of tax payments for their customers.

219 Goldsmith-bankers also combined forms of intermediation in response to opportunities from international piarkets. During the Nine Years War that followed the Glorious Revolution of 1688, a goldsmith-banker named Stephen Evance both supplied delegated investments in Treasury debt and bridged England's domestic bullion market with the world markets for gold and silver. Because of these dual services, Evance was able to take advantage of the complementarity between international bullion arbitrage and the financing of England's forces in Flanders. Exchanging English silver for credits owed the banker on the Continent brought the banker and his merchant customers arbitrage returns. Evance then arranged for these same credits in the Low Countries to finance England's forces in exchange for government debt that the banker desired as interest-earning assets. Because the goldsmith-banker mastered two forms of intermediation, the Treasury could pay the troops with debt, silver left England to maintain the island's balance of payments and the goldsmith-banker supplied both services.

The story of the goldsmith-bankers has been an unsung backdrop to the second half of the seventeenth century as a formative period in England's financial history. Goldsmith-bankers, however, were in the middle of both the private and public innovations that have come to be known collectively as the Financial Revolution. England transformed itself from a financially underdeveloped nation in the middle of the seventeenth century to an empire building major power one hundred years later, and goldsmith-bankers played two crucial roles. These unregulated bankers established private banking decades before the Glorious Revolution in 1688. Moreover, the extensive connections goldsmiths fashioned between London's private and public financial sectors set examples for the Bank of England and the Treasury to follow. The later stages of the Financial

Revolution reallocated the rents from collecting taxes, lending to the Treasury and issuing privileged debt from private financiers to government institutions. The growing power of the Bank of England and the Treasury did not fill a vacuum, but rather displaced private bankers.

220 I. Future Research

Although this thesis has addressed important aspects of seventeenth century goldsmith- banking, much research is left to be done. The detailed examination of the data on Backwell's clearing arrangements with his fellow goldsmith-bankers presented in Chapter 2 will eventually be expanded to allow a glimpse at the breadth of note circulation in Restoration London. Likewise, the internal rate of return of Francis Child's loan portfolio in Chapter 4 needs to be calculated. Child's inconsistent accounting methods prove a challenge, but better estimates of seventeenth century returns are needed.

The monetary and legal backgrounds that have been taken for granted deserve independent consideration. The economics of clipping marks the seventeenth century as an interesting period of private production of token money. Likewise, how merchants navigated the jurisdictional competition between courts of law remains unresolved. Considerable work will be required in examining how transferred debt contracts were actually enforced by the various suppliers of law in the seventeenth century.

Moving forward in time, the author intends to consider how the network of goldsmith- bankers contended with the newly founded Bank of England. Few goldsmith-bankers supported the

Bank of England and competition was fierce between the private bankers and the Bank of England during the first few years after the Bank of England's founding. The author will also consider how the existing system of goldsmith-bankers affected the structure and working of the bank.

Finally, research should move across the channel to Holland. Much has been made of

Dutch finance in the seventeenth century, yet far too little is know of how the Wisselbank worked or how private Dutch banking functioned in the seventeenth century. With a better understanding of

Dutch financial practices, we can more accurately assess how much of England's Financial

Revolution was imported from Holland, how much was a product of domestic ingenuity and how

221 much was a product of an innovative maelstrom combining the two strands in the turbulence that was England's financial history in decades surrounding the turn of the eighteenth century.

222 Bibliography

Primary Sources

Backwell, Edward, Ledgers J (March 1663 to March 1663/4), K (March 1664 to September 1664), M (September 1664 to March 1664/5), O (January 1665/6 to December 1666), P (January 1666/7 to March 1667/8), Q (March 1668 to March 1668/9), R (March 1669 to March 1669/70), S (March 1670 to March 1670/1), T (March 1671 to March 1671/2), (EB/1-9, Royal Bank of Scotland, London).

Chancery Court, Equity side, Reynardson's Division, (Public Record Office, Chancery Lane, C9 110/27).

Child, Sir Francis, Customer Ledgers, Pawn Accounts and Casting Up Book and Loose Notes, (CH/194, Royal Bank of Scotland, London).

Evance, Stephen, Bullion Book #5, (Public Record Office, Chancery Lane, CI 14/179).

Goldsmiths Company Records, (Goldsmiths' Hall, London).

Haynes, Hopton, Brief Memoires Relating to the Silver and Gold Coins of England, (Lansdowne MS 801, British Library; 1700).

Houghton, John, A Collection of Letters for the Improvement of Husbandry and Trade, (Gregg International Publishers, 1692-1702, reprinted in 1969).

Houghton, Thomas, The Alteration of the Coin with a Feasible Method to do it, (Hollander Collection, University of Illinois; 1695).

Reasons Humbly Offered to the Honourable House of Commons by Eminent Merchants and Citizens of the City of London: Showing the Inconveniencies that May Arise by the Bank, anonymous, (Hoare's Bank, London, 1694).

Wisselbank Ledgers, Indices and Balances, (Amsterdam Municipal Archives, 5077/4).

Secondary Sources

Anderson, B. L., and Cottrell, P. L., Money and Banking in England, (London, 1974).

Ashley, Maurice, Financial and Commercial Policy under the Cromwellian Protectorate, (Oxford Press University Press, 1935).

223 Assar, W.D.H., "Bills of Exchange and Agency in the 18th Century Law of Holland and Zeeland," The Courts and the Development of Commercial Law, (Duncker & Humblot, Berlin, 1987), pp. 103-130.

Ashton, Robert, The Crown and the Money Market, 1603-1640, (Oxford, 1960).

Attman, Artur, Dutch Enterprise in the World Bullion Trade, 1550-1800, no. 23, (Goteborg: Kungl. Vetenskaps- och Vitterhets-Samhallet, 1983)

Attman, Artur, American Bullion in the European World Trade, 1600-1800, no. 26, (Goteborg: Kungl. Vetenskaps- och Vitterhets-Samhallet, 1986)

Aylmer, The King's Servants, The Civil Service of Charles I, (London, 1974).

Baker, J. H., "The Law Merchant and the Common Law Before 1700," Cambridge Law Journal, Vol. 38, No.2, (November, 1979), pp. 295-322.

Bhattacharya, Sudipto and Thakor, Anjan, "Contemporary Banking Theory," Journal of Financial Intermediation, Vol. 3, (1993), pp. 2-50.

Baumol, W.J., "The Transactions Demand for Cash: An Inventory Theoretic Approach," Quarterly Journal of Economics, LXVI (Nov. 1952).

Baxter, Stephen B., The Development of the Treasury, 1660-1702, (Cambridge, Mass., 1957).

Brewer, John, 77K? Sinews of Power, (New York, 1989).

Calendar of Treasury Books, Vols. VIII and IX, (HMSO).

Cagan, Phillip and Schwartz, Anna, "The National Bank Note Puzzle Reinterpreted," Journal of Money, Credit and Banking, Vol. 23, No. 3, (August 1991), pp. 293-307.

Calomiris, Charles and Kahn, Charles, "The Role of Demandable Debt in Restructuring Optimal Banking Arrangements," American Economic Review, v. 81, no. 3, (June 1991), pp. 497- 513.

Calomiris, Charles, "Regulation, Industrial Structure, and Instability in U.S. Banking: An Historical Perspective," Structural Change in Banking, ed. Michael Klausner and Lawrence J. White, (1993).

Carlos, Ann, and Van Stone, Jill, "Stock Transfer Patterns in the Hudson's Bay Company: Not a

Family Concern 1670-1730," presented to the Business History Conference, (March 1994).

Challis, C.E., A New History of the Royal Mint, (Cambridge University Press, 1992), p. 391.

Chandaman, CD., The Public Revenue, (Oxford University Press, 1975).

224 Clapham, John, The Bank of England: A History, (Cambridge, 1944).

Clark, Dorothy K., "Edward Backwell as a Royal Agent," The Economic History Review, v. IX (1938-39), pp. 45-55.

Clark, Dorothy K., "A Restoration Goldsmith-Banking House: The Vine on Lombard Street," Essays in Modern English History, (Cambridge, 1941), pp. 3-47.

Clay, Christopher, Public Finance and Private Wealth, the Career of Sir Stephen Fox, 1627-1716, (Oxford, 1978).

Craig, Sir John, The Mint, (Cambridge Press, 1953).

Davis, Ralph, A Commercial Revolution, No. 64,(Historical Association, 1967).

Davis, Ralph, "English Foreign Trade, 1660-1700," The Economic History Review, Second Series, Vol. VII, No. 2, (December 1954), pp. 150-166.

DeKrey, G. S.,A Fractured Society, The Politics of London in the First Age of Party 1688-1715, (Clarendon, 1985).

DeRoover, Raymond, "What is Dry Exchange? A Contribution to the Study of English Mercantilism." Business, Banking and Economic Thought, (University of Chicago Press, 1974).

Diamond, Douglas and Dybvig, Philip, "Bank Runs, Liquidity and Deposit Insurance," Journal of Political Economy, (June 1983), pp. 401-19.

Dickson, P.G.M., The Financial Revolution in England: a Study in the Development of Public Credit, 1688-1756, (Macmillan, 1967).

Dowd, Kevin, The State and the Monetary System, (St. Martin's Press, 1989).

Dowd, Kevin, editor, The Experience of Free Banking, (Routledge 1992).

Dowd, Kevin, Laissez-Faire Banking, (Routledge, 1993).

Dowd, Kevin, "Competitive Banking, Bankers' Clubs, and Bank Regulation," Journal of Money, Credit, and Banking, Vol. 26, No. 2, (May 1994), pp. 289-308.

Earle, Peter, The Making of the English Middle Class, (London, 1989).

Feavearyear, Albert, The Pound Sterling, (Clarendon Press, 1963), 332.4942F31p.

Friedman, Milton and Schwartz, Anna, A Monetary History of the United States, (Princeton, 1963).

225 Friedman, Milton and Schwartz, Anna, "Has Government Any Role in Money?" Journal of Monetary Economics, Vol. 17, (1986), pp. 37-62.

George, Mrs. Eric, "Note on the Origin of the Declared Account," The English Historical Review, v. 31 (1916), pp. 41-58.

Glasner, David, Free Banking and Monetary Reform, (Cambridge University Press, 1989).

Goodhart, Charles, 77K? Evolution of Central Banks, (MIT Press, 1988).

Goodhart, Charles, "Are Central Banks Necessary?" Unregulated Banking: Chaos or Order, Capie and Wood editors, (St. Martin's Press, 1991), pp. 1-22.

Gorton, Gary, "Clearinghouses and the Origin of Central Banking in the United States," Journal of Economic History, Vol XLV, No. 2, (June 1985), pp. 277-284.

Gorton, Gary and Mullineaux, Donald, "The Joint Production of Confidence: Endogenous Regulation and Nineteenth Century Commercial-Bank Clearinghouses," Journal of Money, Credit and Banking, Vol. 19, No. 4, (November 1987), pp. 457-468.

Gorton, Gary and Pennacchi, George, "Financial Intermediaries and Liquidity Creation," Journal of Finance, Vol. XLV, No. 1, (March 1990), pp. 49-71.

Gorton, Gary, "Reputation Formation in Early Bank Debt Markets." National Bureau of Economic Research Working Paper, (No. 4400, July 1993).

Heal, Ambrose, London Goldsmiths, (Cambridge, 1935).

Hilton Price, F.G., ^ Handbook of London Bankers, (London, 1890A).

Hilton Price, F.G., "Some Account of the Business of Alderman Edward Backwell, Goldsmith and Banker in the 17th Century," Transactions of the London and Middlesex Archaeological Society, Vol. VI (London, 1890B), pp. 191-230.

Hilton Price, F.G., "Some Notes of the Early Goldsmiths and Bankers," Transactions of the London and Middlesex Archaeological Society, Vol. VI (London, 1881), pp. 225-281.

Hoffman, P., Rosenthal, J. and Postel-Vinay, G., "What Do Notaries Do? Overcoming Asymmetric Information in Financial Markets: The Case of Paris, 1751," (Working Paper, June 1994)

Holden, J. Milnes, The History of negotiable Instruments in English Law, (University of London, 1955).

Holdsworth, W. S., A History of English Law, Vol, VIII, (Little, Brown and Company, 1926).

Homer, Sydney, A History of Interest Rates, Second Edition, (Rutgers University Press, 1977).

226 Horsefield, Keith, British Monetary Experiments, 1650-1710, (London, 1960).

Horsefield, Keith, "The "Stop of the Exchequer" Revisited," Economic History Review, v. 35 (1982), pp. 511-528.

Horwitz, Stephen, Monetary Evolution, Free Banking, and Economic Order, (Westview Press, 1992).

Hyamson, Albert M, The Sephardim of England: a History of the Spanish and Portuguese Jewish Community, 1492-1951, (London: Methuen & Co, 1951).

Israel, Jonathan, "The Economic Contribution of Dutch Sephardi Jewry to Holland's Golden Age, 1595-1713," Tijdschrift voor Geschiedenis, 96 (1983), pp. 505-535.

Jones, D.W., "London Merchants and the Crisis of the 1690s," Crisis and Order in English Towns 1500-1700, ed. Peter Clark and Paul Slack, (Toronto, 1972), pp. 311-356.

Jones, D.W., War and Economy in the Age of William III and Marlborough, (Basil Blackwell, 1988).

Kaufman, George, "Bank Contagion: A Review of the Theory and Evidence," Journal of Financial Services Research, Vol. 8, No. 2, (April 1994), pp. 123-150.

Kenyon, Robert Lloyd, The Gold Coins of England, (London, 1884), pp. 178-9.

Kerridge, Eric, Trade & Banking in Early Modern England, (Manchester University Press, 1988).

Li, M.H., The Great Recoinage of 1696 to 1699, (Weidenfeld & Nicolson, 1963).

Macaulay, T.B., The History of England from the Accession of James II, Vol. IV, (Lippincott, 1873).

McCusker, John and Gravesteijn, Cora, The Beginnings of Commercial and Financial Journalism, (NEHA, Amsterdam, 1991).

McCusker, John, Money and Exchange in Europe and America, 1600 - 1750, A Handbook, (University of North Carolina Press, Chapel Hill, 1978), pp. 13-55.

Melton, Frank, "Goldsmith Notes, 1654-1655," Journal of the Society of Archivists, Vol. 6, No. 1, (April 1978), pp. 30-31.

Melton, Frank, Sir Robert Clayton and the Origins of English Deposit Banking, 1658-1685, (Cambridge, 1986).

Miller, Merton, and Orr, Daniel, "A Model of the Demand for Money by Firms," Quarterly Journal of Economics, v. 80, (1966), pp. 413-435.

227 Mitchell, David M., "'Mr Fowle, Pray Pay the Washwoman': The Trade of a London Goldsmith- banker, 1660-1692," Business and Economic History, Vol. 23, No. 1 (Fall 1994), pp. 27-38.

Munro, John, "Bullionism and the Bill of Exchange in England, 1272-1663: A Study in Monetary Management and Popular Prejudice," 77K? Dawn of Modern Banking, (Yale University Press, 1979), pp. 169-240.

Neal, Larry, The Rise of Financial Capitalism, (Cambridge University Press, 1990).

Neal, Larry, "For God's Sake, Remitt Me: The Adventures of John Law's London Goldsmith- Banker, 1712-1729," Business and Economic History, Vol. 23, No. 2, (forthcoming, 1994).

Newbold, Paul and Bos, Theodore, Introductory Business Forecasting, (South-Western, 1990).

Nichols, GO., "The Development of English Repayment and Interest Rates: An Examination of Government Borrowing and the Loans of Sir Stephen Fox, 1667-74," Revue Internationale d'Historie de la Banque, v. 30-31, (1985), pp. 247-265.

North, Douglass and Weingast, Barry, "Constitutions and Commitment: The Evolution of Institutions Governing Public Choice in Seventeenth-Century England," 77K? Journal of Economic History, v. XLIX, no. 4, (December, 1989), pp. 803-832.

North, Roger, The Lives of the Norths, Vol. Ill, (London, 1926).

Ogilvie, Charles, The King's Government and Common Law, 1471-1641, (Oxford, 1958).

The Diary of Samuel Pepys, ed. by R.C. Latham and W. Mathews, vol. VIII, (University of California Press, 1974).

Palasek, Karen, "The Case for a Laissez-Faire Monetary System," a comment on White (1989), Cato Journal, v. 9, n. 2, (Fall 1989), pp. 399-404.

Piergiovini, "Courts and Commercial Law at the Beginning of the Modem Age," 77K? Courts and the Development of Commercial Law, (Duncker & Humblot, Berlin, 1987).

Posthumus, N.W., Inquiry into the History of Prices in Holland, Vol. I, (E.J. Brill, 1946).

Redish, Angela, "The Evolution of the Gold Standard in England," Journal of Economic History, Vol. L, no. 4, (December 1990).

Richards, R.D., "A Pre-Bank of England English Banker - Edward Backwell," 77K? Economic Journal, (January 1928), pp. 335-355.

Richards, R.D., 77K? Early History of Banking in England, (Westminster, 1929).

Richards, R.D., "The Stop of the Exchequer," Economic History: Supplement to the Economic Journal, II (1930), pp. 45-62.

228 Rolnick, A. and Weber, W., "New Evidence on the free Banking Era," American Economic Review, Vol. 73, (1983), pp. 1080-1091.

Rolnick, A. and Weber, W., "The Causes of Free Bank Failures: A Detailed Examination," Journal of Monetary Economics, Vol. 14, pp. 267-291.

Rolnick, A. and Weber, W., "Explaining the Demand for Free Bank Notes," Journal of Monetary Economics, Vol. 21, (1988), pp. 47-71.

Roseveare, Henry G., The Advancement of the King's Credit 1660-1672, (unpublished Ph.D. Thesis, Cambridge, 1962).

Roseveare, Henry G., The Treasury, (London, 1969).

Roseveare, Henry G., 77K? Treasury 1660-1870, The Foundations of Control, (London, 1973).

Roseveare, Henry G., The Financial Revolution, 1660-1760, (London, 1991).

Samuel, Edgar R., "Sir Francis Child's Jewellery Business," 77K? Three Banks Review, no. 113, (March 1977), pp. 43-55.

Schwartz, Anna, "Are Central Banks Necessary?" Critical Review, Vol. 7, No. 2-3, (Spring-Summer 1993), pp. 355-370.

Selgin, George, The Theory of Free Banking, (Rowman and Littlefield, 1988).

Selgin, George, "Legal Restrictions, Financial Weakening, and the Lender of Last Resort," 77K? Cato Journal, v. 9, n. 2, (Fall 1989), pp. 429-60.

Selgin, George, "Free Banking and Monetary Control," (Working Paper, University of Georgia, May 1994).

Selgin, George, "The Rationalization of Central Banks," Critical Review, Vol. 7 No. 2-3, (Spring- Summer 1993), pp. 335-354.

Selgin, George and White, Larry, "How Would the Invisible Hand Handle Money?" forthcoming, Journal of Economic Literature, (1994).

Shaw, William A. ed., State Papers (Domestic), Interregnum, xvi, No. 100, in "A Selection of Commonwealth State Papers Illustrative of the Monetary Movements of 1649-1651," Select Tracts and documents Illustrative of English Monetary History, 1626-1730, (George Harding, London, 1935).

Spencer, P.D., "Speculative and Precautionary Balances as Complements in the Portfolio," Journal of Banking and Finance, Vol. 13, (December 1989), pp. 811-830. (London, 1934), pp. 149-176.

229 Supple, B.E., Commercial Crisis and Change in England, 1600-1642, (Cambridge University Press, 1959).

Sutherland, Stuart, "The Law Merchant in England in the Seventeenth and Eighteenth Centuries," Transactions of the Royal Historical Society, Fourth Series, Vol XVII,

White, Lawrence, Free Banking in Britain, (Cambridge, 1984).

White, Lawrence, "Accounting for Non-interest-bearing Currency: A Critique of the Legal Restrictions Theory of Money," Journal of Money, Credit and Banking, Vol. 19, No. 4, (November 1987), pp. 448-455.

White, Lawrence, Competition and Currency, (New York University Press, 1989).

White, Lawrence, "What Kinds of Monetary Institutions Would a Free Market Deliver?" Cato Journal, v. 9, n. 2, (Fall 1989), pp. 367-393.

Wilson, Thomas, A Discourse Upon Usury, with an historical introduction by R.H. Tawney, (Harcourt Brace, original 1572).

van der Wee, Herman, "Monetary, Credit and Banking Systems," The Cambridge Economic History of Europe, the Economic Organization of Early Modern Europe, Vol. V, (Cambridge University Press, 1977).

van Dillen, J. G., "The Bank of Amsterdam," History of the Principal Public Banks, (Martinus Nijhoff: the Hague, 1934), pp. 79-123.

Zahedieh, Nuala, "London and the Colonial Consumer in the Late Seventeenth Century," Economic History Review, Vol. XLVII, No. 2, (1994), pp. 239-261.

230 Appendix

Appendix A

Notes on the Legal History of Transferability on the European Continent

The rules constraining commercial contracts in the Middle Ages centered around Europe's fairs and commercial hubs. The medieval law merchant was a local enforcer of commercial dealings deriving authority from the merchants themselves or operators of particular markets.

Medieval commercial contracts were enforced under local custom, and up to the fifteenth century, mercantile law was not part of any codified or consistent legal body. Likewise, enforceable contracts under local merchant customs were usually unrestricted in terms of transferability. Eighth and ninth century Lombards were including clauses making wills, conveyances and forms of

indebtedness transferable to nominees.(Holdsworth, pp. 115-6) The most flexible medieval transferability clause was a promise to pay "to producer" or "to bearer". The producer could sue in his own right with standing independent of the creditor, and the producer did not need to show how he come into possession of the contract.(Holdsworth, p. 117) Moreover, the document itself became the contract, so producing the document established a valid contract.

Medieval law was practical but lacked a developed body of legal theory. The revival of

ancient works during the Renaissance provided a developed body of law - Roman civil law - that

the continental powers found inviting. However, Justinian law was hostile to that the merchant

customs of transferability.1 As Roman law was adopted, especially in France and Italy, legal

1 Holdsworth describes the legal change in perspective from the Middle Ages to the Renaissance jurist as a shift from Germanic procedure, "Is the defendant bound to pay? and are his obligations to pay the bearer the direct consequence of the form of his promise to pay?" to the civil and cannon law procedure, "Has the plaintiff a right of action?" Under the older legal procedure, if A promises to pay

231 recognition of transferability diminished, and merchant customs were disenfranchised from the law.

"They [Renaissance jurists] did not hesitate, therefore, to sacrifice commercial convenience on the shrine of legal orthodoxy."(Holdsworth, p. 122) Merchants could and did still enforce transferable contracts amongst each other, but merchants could not turn to civil law for enforcement of transferable commercial debts.

It is hardly necessary to say that the commercial world was seriously inconvenienced by these developments of legal doctrine. The older instruments made payable to the creditor or his nominee had disappeared,...and now the negotiable character of these instruments payable to bearer had been destroyed.(HoIdsworth, p. 125)

The structure in which civil law affected enforcement of commercial agreements varied by

Continental nation. In the sixteenth and seventeenth centuries, Italian commercial courts were replacing "non-technical judges" with "learned lawyers."(Piergiovini, p. 21) Italian commercial courts were adopting civil procedure, and commercial law was becoming systematic and rigorous.

Merchant tribunals in France provided expediency and familiarity with merchant customs. Between

1547 and 11609, the Dutch codified their commercial (customary) law.(Assar, p. 106) Moreover, the Dutch had no separate merchant courts like the French or Italians. "Claims concerning bills of exchange have never been subject to a specialized court in Holland and ZeeIand."(Assar, p. 109)

Up through 1600, transferable financial contracts of a commercial nature were not recognized throughout Europe.(DeRoover, p. 194) French merchants, however, were able to maintain some transferability by leaving the name of the creditor blank.(Holdsworth, p. 125)

X or bearer, A has to pay. Under the later legal procedure, A owes X but the bearer was not party to the contract, so the bearer can only claim payment if acting as an agent of X. The transferable properties of the contract are significantly weakened under the Roman procedure since the bearer would have to prove he ia acting as X's agent and accept X's standing in court. For example, if X died, the bearer has no claim.

232 Appendix B

Daily Balances of Goldsmith-Bankers with Edward Backwell

Thomas and Robert Vyner

30

20

10 in 1 •TwTr JHH^ ' '"ll T

-20

-30

-^40 1663 16641665166616671668 1669 16701671

233 Francis and Isaac Meynell, Gross

35

30

25

20 ij: 5

0 ii—*rhhi ' '^^tyfN^i'djl— -5 w -10 1663 1664-1665 1666 1667 16681669 1670 1671

Francis and Isaac Meynel, Net Dunkirk Money

1663 1664 1665 16661667 1668 1669 16701671

234 Thomas Row

I!

16631664166516661667166816691670 1671

Thomas Row, Net East India Co. Loan

6000

-4000

-6000 1663 1664 1665 1666 1667 1668 1669 1670 1671

235 John Lindsay

5000

4000

3000

2000 1000 I 1

0

-1000 •-M-w*

-2000

-3000

-4000

-5000 1663 1664 1665 1666 1667 1668 1669 1670 1671

John Welstead

L-flr flityJ iff TF

166316641665 1666166716681669 1670 1671

236 Jeremiah Snow

ftf\\ ^M^\

166316641665 166616671668 166916701671

George Snell

15-

10

-X yU—*" ''1 '''y^'#&''^'''«^

-5-

-10-

-15- 166316641665 166616671668166916701671

237 I

John Colvill

10

... II t. -5

-10

-15 16631664166516661667 1668 166916701671

Dorothea Colvill

3500

3000

2500

2000

1500 3 I 1000 500

0

-500

-1000

-1500 1663 1664 1665 1666 1667 1668 1669 1670 1671

238 Edmund Hinton

4000

3000

2000

1000 0 \nil# -1000 • hi.

-2000 I'

-3000

-4000

-5000

-6000 1663 1664 1665 1666 1667 1668 1669 1670 1671

Benjamin Hinton

1663 1664166516661667 16681669 1670 1671

239 John Portman

3000

-3000 166316641665166616671668166916701671

John Mawson

4000

-2000 166316641665 166616671668 166916701671

240 Joseph Horneby

20

15

10

iK nTj TTlpB

-10

-15 16631664166516661667 1668 16691670 1671

Thomas Cook and Nicholas Carey

500

-500 I -1000

-1500

-2000 1663 1664 1665 1666 1667 1668 1669 1670 1671

241 John Hinde

600

400

200

•8 § -200 S. -400

-600

-800

-1000 1663 1664 1665 1666 1667 1668 1669 1670 1671

Thomas Kirwood

500

0 Jll 1.1.11.. 1, .1

-500 "1 1 11 -1000 II » -1500 S. -2000 1 -2500

-3000

-3500

-4000 1663 16641665166616671668166916701671

242 \

Bernard Turner

166316641665166616671668 166916701671

Thomas Williams

500- JPI|| -500-

w -1000- c | *- -1500

-2000 •

-2500•

-3000 • 166316641665166616671668 166916701671

243 Appendix C

How Clearing in Backwell's Ledger Worked

Backwell's clearing accounts were recorded in the extant ledgers lettered "J" in 1663 through "T" in 1671 with ledgers "K" and "N" missing. The ledgers were large tomes that acted as the final record of the banker's activities. The ledgers were constructed from information in Day

Books, Work Books (for worked pieces) and any other as yet unknown sources that lay closer to the original transaction. None of these sources survives. Each ledger folio recorded debits on the left hand page and credits on the right hand page. Each page was broken into columns reading, from left to right: the date, a reference entry, a description of the transaction, an additional reference entry and the pound sterling value of the transactions. When the transaction was within or between ledgers, the reference entries would give the matching folio number. Both intra- and inter-ledger connections were infrequent. Most transactions occurred before the ledger stage, and, consequently, most reference entries are individual letters referring to day books and other potential sources which no longer exist.

Under the clearing system, debits occurred when Backwell paid funds on behalf of a fellow banker's account. Although the account holder sometimes withdrew funds himself, most often

Backwell paid a third party. Credits occurred when money or notes were given to Backwell by the account holder or by third parties.

The ledgers often did not detail on what authority Backwell adjusted a fellow banker's account, especially after 1665. Money (silver coin), gold and plate transactions were usually described as such. Specie deposited or withdrawn was a credit or debit, respectively. Most entries, however, were transactions by paper debt. The two common types of seventeenth century short term debt were promises to pay and orders to pay. The promise to pay (called promissory notes)

244 was an IOU, and a banker's promissory note came to be called a bank note. An order to pay was similar to a modem style check with the bill of exchange being a specialized form.

In the seventeenth century, both promises and orders were called "notes" except for formal bills of exchange. Many entries on the credit side referred to notes, mostly Backwell's bank notes and not orders on Backwell to pay. In Backwell's early ledgers, 1663-64, he often wrote, "my note" which increasingly became simply "note." These notes were not orders to pay because Backwell did not mention who would be debited. The goldsmith-banker would credit a bank note depositor's account with no offsetting debit to anyone else's account. An order (check) drawn on Backwell, however, would begin as a debit on the drawer's account. The order would only appear as a credit if the drawee wanted to take the drawer's debit as a ledger credit on Backwell rather than money or a bank note. Credit entries by orders on Backwell had to have a matching debit.

Chapter 2, Table 5, presents an example from the accounts of goldsmith-banker Thomas

Row for two months, March 1665 and January 1666. On the right side, Row was credited for £400 on March 24 by Francis Meynell. On March 24, Meynell's account with Backwell was debited

£400 with the description, "to him [Meynell]." Row had deposited an order drawn by Meynell on

Backwell and Meynell was duly debited. The descriptions debiting Meynell did not mention the location or even existence of the off-setting entry within Backwell's ledgers. This information was not recorded because Row need not have generated a credit. Instead, Row could have walked away with cash or a note on Backwell which would not have created an off-setting ledger entry. Inter- account clearing of orders on Backwell were an infrequent part of the goldsmith-banker's business.

Also in March of 1665, Thomas Row was credited by Sir Thomas Vyner for £273:15:-.

Vyner's account makes no mention of a debit on March 20 or any debit that month for £273:15:-.

Row was not presenting Backwell with an order drawn on Vyner's account. Vyner's credit to

Row's account could have been a note on Vyner, a bill of exchange, etc. Again, most credits

245 lacked an off-setting debit within Backwell's bank and so were either Backwell's promissory notes or the notes of others.

The debit side is simpler. Except when drawing down a credit, Backwell only accepted the debit upon the note (promise or order) of the account holder. Returning to the sample from the accounts of Thomas Row (Chapter 2, Table 5), on March 1 1665, £1,000 was paid to William

Home out of the account of Thomas Row. For Backwell to know to deduct the thousand pounds from Row, Home had to present Backwell with a note of Row's. The entry on March 8, 1665, in which Hugh Mason collects £86 from Backwell, "ordered the 13th day of February," provides a rare example of a description specifying the type of note (promise or order). The order could have been a simple check or, given the time between writing and receiving, a bill of exchange.

246 Appendix D

Econometric Results of Edwared Backwell's Inter-Banker Clearing

with Benjamin Hinton, John Lindsay and Thomas Row

Table D.l

Benjamin Hinton

Dependent Variable: LN(Days to Clear; Same Day=l)

Percent Parameter Variable T for HO: Change" Variable Estimate'' Mean fi=0 Prob>|T|

3.7' INTERCEPT 1.543827 11.884 0.0001 -7.4 Debit t+1 -0.000350 211 -3.527 0.0005 4.5 Debit t+6 0.000216 209 1.980 0.0490 7.0 Balance t-5 -0.000120 -589 -1.196 0.2331 -12.5 Balance t-6 0.000210 -593 2.243 0.0259 -17.5 Monday -0.175043 dummy -1.677 0.0949 25.2 Saturday 0.251556 dummy 2.618 0.0095 -38.7 Time Trend -0.002140 181 -4.518 0.0001 -62.0 February -0.619848 dummy -3.687 0.0003 -38.0 March -0.380178 dummy -2.514 0.0127 31.8 May 0.318354 dummy 2.268 0.0244 9.8 Total t+3 -0.00004063 -2423 -2.182 0.0302 -7.8 Total t+4 0.000032345 -2416 1.740 0.0832

N=223, F Value=7.500, Prob>F 0.0001, Adj R2=0.2600, R2=0.3000 " These values are the percentage change in the dependent variable given the mean value of each independent variable. b The coefficient values (*100) are the percentage change in the dependent variable for a one unit change in each independent variable. c This value is the anti-logarithm value of the intercept minus one, so same day clearing equals zero.

Independent variables in bold type are significant at the ninety-five percent confidence interval.

247 Table D.2

John Lindsay

Dependent Variable: LN(Days to Clear; Same Day=l)

Percent Parameter Variable T for HO: Change* Variable Estimate*' Mean B=0 Prob>|T|

2.9C INTERCEPT 1.363604 9.941 0.0001 -11.2 Debit t -0.000253 443 -3.119 0.0021 -6.1 Debit t+1 -0.000215 286 -2.637 0.0091 4.0 Debit t+4 0.000165 240 1.882 0.0615 -3.5 Debit t+6 -0.000130 272 -1.477 0.1415 5.6 Balance t-3 -0.000126 -441 -1.474 0.1422 -10.1 Balance t-4 0.000234 -430 2.887 0.0044 -32.6 Monday -0.325734 dummy -2.815 0.0054 -14.1 Tuesday -0.140702 dummy -1.317 0.1895 -17.0 Thursday -0.170385 dummy -1.496 0.1364 -14.8 Time Trend -0.000828 179 -1.164 0.2461 -18.5 May -0.185248 dummy -1.194 0.2343 29.0 June 0.289922 dummy 1.780 0.0768 35.8 August 0.358460 dummy 2.482 0.0140 -3.2 October -0.031797 dummy -0.183 0.8549 -60.7 November -0.606853 dummy -3.109 0.0022 -37.6 December -0.375637 dummy -1.722 0.0869 -5.2 Total t-3 0.000017986 -2792 1.561 0.1203 9.5 Total t+1 -0.000037228 -2565 -2.777 0.0061

N=194 , F Value=4.716, Prob>F=0.0001, Adj RW.2564, R2==0.325 4

* These values are the percentage change in the dependent variable given the mean value of each independent variable. b The coefficient values (* 100) are the percentage change in the dependent variable for a one unit change in each independent variable. c This value is the anti-logarithm value of the intercept minus one, so same day clearing equals zero.

Independent variables in bold type are significant at the ninety-five percent confidence interval.

248 Table D.3

Thomas Row

Dependent Variable: LN(Days to Clear; Same Day=l)

Percent Parameter Variable T for HO: Changie " Variable Estimate'' Mean B=0 Prob>|T|

2.6C INTERCEPT 1.290809 10.723 0.0001 3.7 Balance t-3 0.000315 116 2.650 0.0088 1.9 Balance t-6 0.000223 83 1.952 0.0526 -15.8 Tuesday -0.151855 dummy -1.305 0.1937 20.7 Friday 0.206961 dummy 1.690 0.0928 24.3 Saturday 0.243363 dummy 2.227 0.0273 -53.2 Time Trend -0.002876 185 -6.016 0.0001 -21.5 April -0.214782 dummy -1.268 0.2064 46.9 May 0.468614 dummy 2.491 0.0137 45.6 September 0.455860 dummy 2.841 0.0050 -14.1 Total t-4 0.000057214 -2460 2.134 0.0343 22.3 Total t-3 -0.000087262 -2556 -2.756 0.0065 -9.2 Total t-2 0.000038444 -2383 1.745 0.0828 6.1 Total t+4 -0.000025897 -2373 -1.348 0.1796 -6.9 Total t+6 0.000030241 -2285 1.770 0.0785

N=183 , FValue=5.153, Prob>F=.0001, Adj R .2=0.2411, R2==0.299 2

° These values are the percentage change in the dependent variable given the mean value of each independent variable. b The coefficient values (*100) are the percentage change in the dependent variable for a one unit change in each independent variable. c This value is the anti-logarithm value of the intercept minus one, so same day clearing equals zero.

Independent variables in bold type are significant at the ninety-five percent confidence interval.

The strong intercepts for all three bankers carry the greatest explanatory power and suggest debits were cleared in some regular manner. Row's and Lindsay's intercepts are 2.6 and 2.9 days

249 respectively while Hinton's is about a day longer at 3.7 days. These agree with the findings in

Table 2.6 that Hinton consistently cleared more slowly than did the other two bankers in the sample.

Because of the semi-log nature of the regressions, interpretation of the coefficients is by

percentage change. The coefficients themselves are the percentage change of the Days to Clearing

for a one unit change in the independent variable. The "Percent Change" column on the left of each

regression calculates the percentage change in Days to Clearing given the mean value of the

independent variable to facilitate comparisons.

The regression results in Tables D.l-3 do not support any particular day of the week as a

universal clearing day. Lindsay and Hinton do have strong negative coefficients for Mondays. If

Mondays were a regular clearing day, Tuesdays should be positive, but Lindsay's Tuesday variable

is negative. Moreover, Lindsay's Tuesday variable lacks significance at even ten percent while

Hinton's model did not even select the Tuesday variable. Mondays were not selected for Row's

model, and Tuesdays had a negative sign and low level of significance. A charitable interpretation

could argue Hinton and Lindsay often cleared on Mondays while Row did not and that none of the

other weekdays were more or less likely to have clearing occur.

Like Mondays, Saturdays were the other standout day, but for Row instead of Lindsay. For

both Thomas Row and Benjamin Hinton, debits arriving on Saturday took twenty-five percent

longer to clear than those that arrived on the control day of Wednesday. Since no activity was

recorded for Sundays, the Saturday coefficients likely capture the Sunday effect. Curiously,

Saturdays were not selected for John Lindsay's model which weakens the argument that Lindsay

and Hinton worked under the same clearing regime.

The interactive clearing model finds support because Hinton and Lindsay's models show

signs of Backwell reacting to large debit inflows while Row's model does not. Recall Row

generally ran positive overall balances with Backwell in contrast to the other two bankers' negative

250 balances. The highly significant, negative, and large coefficients for the original (time t) Debits variable for Lindsay and the next days Debits variable for Hinton and Lindsay indicate Backwell may have pushing for faster clearing when large debits arrived and when more debits followed for both bankers with negative balances.

In contrast, the model selection criteria did not recommend including the original or following day's debits in Thomas Row's model. Rather, Row's model included his account's balance from three and six days prior to the day Backwell accepted the note. Both Balance variables had positive coefficients which meant the greater Row's positive balance the longer the average Days to Clearing took for debits originating a few days later. If positive balances were a function of 1) clearing between Row and Backwell and 2) Row having more of Backwell's notes to clear than Backwell had of Row's, larger positive balances would mean Row and Backwell had cleared leaving Row with surplus holdings of Backwell notes. With clearing meaning more credits for Row, Backwell would have been in no hurry to initiate clearing. Row may well have found such precautionary reserves with Backwell to be beneficial. The lack of Debit variables in Thomas

Row's regression agrees with the hypothesis that Backwell was not actively decreasing clearing time with Row.

Hinton and Lindsay also had lagged Balances variables in their models. Hinton's and

Lindsay's results, however, do agree with the description of Backwell working to reduce large amounts of uncleared debts he carried. Lindsay and Hinton each have a positive and significant coefficient for balances at t-4 and t-6 respectively. The effect of both coefficients, however, are mitigated by interactive negative coefficients one time period later. Both Lindsay's t-3 and Hinton's t-5 negative coefficients are smaller and much less significant than the earlier positive coefficients.

The interactive effect of the coefficients is still positive meaning large (in absolute value) negative

251 balances promoted quicker clearing. At least for Hinton and Lindsay, Backwell appeared to respond to growing uncleared debt levels by speeding up the clearing process.

Combining the debits and the balances coefficients, the decentralized clearing hypothesis works for Hinton and Lindsay and partially succeeds for Row. Large overdrafts, large originating debits and large debits at t+1 shortened the clearing time for Hinton and Lindsay. This agrees with the view that Backwell sought to quicken clearing as overdrafts grew. Thomas Row was very different. No variables were even selected to suggest Backwell tried to diminish the clearing time of Row's debits which agrees with interpreting Row's positive balances as a signal that Backwell had little to gain in clearing with Row. That Row's large positive balances would retard the promptness of future clearing indicates Row and Backwell's relationship defied the simple clearing model proposed above. Thomas Row had been Backwell's apprentice and Backwell was purchasing bullion from Row over the course of the year. Recall also that a £42,000 pound bullion transfer has been excluded from the data but did occur in 1670. To conclude that Backwell and Row had a special relationship seems warranted.

The regressions also included three other categories of variables to control for seasonality, time and Backwell's overall business. No clear seasonality emerged. Rather, the bankers were a hodgepodge of months and coefficient signs. For example, all three models included May but John

Lindsay's coefficient was negative and insignificant unlike Row's and Hinton's May which were positive and quite significant. The end of the year brought much faster clearing. Both Hinton's and

Row's Time Trends were negative and substantial. While John Lindsay's Time Trend variable was of little consequence, both his November and December dummy variables were negative and large in absolute value. Why the pace of clearing increased over the year is unclear, but the effect does parallel the diminishment of Backwell's overall balance with all goldsmith-bankers (see Chart 2.6).

252 Backwell's total goldsmith-banker balances mentioned above were introduced as a proxy for the state of Backwell's overall business but is a very incomplete measure of the breadth of

Backwell's business. Unfortunately, more comprehensive variables have yet to be collected. The measure of aggregated banker balances suffers a composition problem since Lindsay, Hinton and

Row are all goldsmith-bankers included in the total. None of the models included Backwell's overall balances for the day debits originated (Total for time = t) suggesting the correlation between each banker's debits and that day's total balances were not overpowering. Moreover, Benjamin

Hinton's model lacked any lagged Total variables. John Lindsay's included a three day lag implying negative overall balances lead to shortened clearing times for Lindsay. Row had a similar result with three interactive Total variables at t-4, t-3 and t-2.

The regressions provide reasonable fits given the limited sample sizes. In time, the author intends to expand the samples to additional years and additional goldsmith-bankers. The overall results, however, do support the hypothesis that Backwell acted to reduce large uncleared balances.

While the above regressions are the best fits given the available data series, they are susceptible to the charge of comparing apples with oranges. Tables D.4-6 rectify the different model structures for the three bankers by imposing the same model on all three. The results from Tables D.4-6 are reported in Tables 2.7 and 2.8 in Chapter 2. Each variable with a ninety-five percent significance or better from each banker's model in Tables D.l-3 above have been pooled. While each model below loses explanatory power relative to the models presented above, they all do share the same structure.

253 Table D.4

Benjamin Hinton

Under an Imposed Common Model

Dependent Variable: LN(Days to Clear; Same Day=l)

Percent Parameter Variable T for HO: Change" Variable Estimate"" Mean 8=0 Prob>|T|

3.6' INTERCEPT 1.530942 10.992 0.0001 1.8 Debit 0.000056983 328 0.603 0.5475 -7.1 Debit t+1 -0.000336 212 -3.278 0.0012 4.0 Debit t+6 0.000190 209 1.710 0.0888 0.7 Balance t-3 -0.000011142 -603 -0.109 0.9135 4.3 Balance t-4 -0.000072824 -594 -0.695 0.4878 -10.0 Balance t-6 0.000168 -593 2.053 0.0413 -19.6 Monday -0.195649 dummy -1.835 0.0679 23.7 Saturday 0.236803 dummy 2.383 0.0181 -40.1 Time Trend -0.002218 181 -4.137 0.0001 -64.0 February -0.640407 dummy -3.721 0.0003 -38.8 March -0.388370 dummy -2.475 0.0141 26.1 May 0.261081 dummy 1.771 0.0780 02.2 August 0.022726 dummy 0.189 0.8504 10.9 November 0.109468 dummy 0.749 0.4547 -0.4 Total t-3 0.000001601 -2792 0.146 0.8838 3.6 Total t+1 -0.000013979 -2565 -0.801 0.4240 -0.2 Total t+3 0.000000863 -2473 0.061 0.9513

N=223 , F Value=5.034, Prob>F=0.0001, R2=0.2945, Adj R2=0.236 0

" These values are the percentage change in the dependent variable given the mean value of each independent variable. b The coefficient values (*100) are the percentage change in the dependent variable for a one unit change in each independent variable. c This value is the anti-logarithm value of the intercept minus one, so same day clearing equals zero.

Independent variables in bold type are significant at the ninety-five percent confidence interval.

254 Table D.5

John Lindsay

Under an Imposed Common Model

Dependent Variable: LN(Days to Clear; Samei Day=l)

Percenlt Parameter Variable T for HO: Change" Variable Estimate"" Mean B=0 Prob>|T|

3.4' INTERCEPT 1.485334 9.942 0.0001 -9.3 Debit -0.000211 443 -2.531 0.0123 -5.3 Debit t+1 -0.000187 287 -2.190 0.0299 -4.4 Debit t+6 -0.000106 272 -1.169 0.2438 5.6 Balance t-3 -0.000128 -441 -1.483 0.1400 -10.6 Balance t-4 0.000247 -430 2.708 0.0074 -0.0 Balance t-6 0.000000273 -421 0.004 0.9970 -21.8 Monday -0.210755 dummy -1.844 0.0668 14.0 Saturday 0.139547 dummy 1.364 0.1744 -32.7 Time Trend -0.001829 179 -3.235 0.0015 1.2 February 0.012198 dummy 0.073 0.9420 -19.5 March -0.194664 dummy -1.221 0.2239 -21.3 May -0.212582 dummy -1.320 0.1884 36.2 August 0.361917 dummy 2.694 0.0077 -49.2 November -0.491698 dummy -3.045 0.0027 -3.6 Total t-3 0.000012765 -2792 1.130 0.2600 7.5 Total t+1 -0.000020945 -2565 -1.191 0.2351 0.9 Total t+3 -0.000003702 -2473 -0.255 0.7993

N=194 , F Value=4.161, Prob>F=0.0001, Adj R2=0.2169, R2==0.285 5

" These values are the percentage change in the dependent variable given the mean value of each independent variable. b The coefficient values (* 100) are the percentage change in the dependent variable for a one unit change in each independent variable. c This value is the anti-logarithm value of the intercept minus one, so same day clearing equals zero.

Independent variables in bold type are significant at the ninety-five percent confidence interval.

255 Table D.6

Thomas Row

Under an Imposed Common Model

Dependent Variable: LN(Days to Clear; Same Day=l)

Percent Parameter Variable T for HO: Change" Variable Estimate"" Mean 6=0 Prob>|r

2.8' INTERCEPT 1.345108 8.474 0.0001 -2.9 Debit -0.000097414 299 -0.695 0.4879 -0.3 Debit t+1 -0.000017804 176 -0.113 0.9100 -2.6 Debit t+6 -0.000168 156 -1.140 0.2560 2.3 Balance t-3 0.000204 116 1.425 0.1561 0.2 Balance t-4 0.000016928 93 0.115 0.9084 1.8 Balance t-6 0.000218 84 1.698 0.0913 -1.3 Monday -0.013227 dummy -0.108 0.9139 24.4 Saturday 0.243606 dummy 2.062 0.0408 -44.2 Time Trend -0.002389 185 -3.931 0.0001 12.1 February 0.121177 dummy 0.481 0.6313 -22.9 March -0.228933 dummy -1.165 0.2457 31.0 May 0.309871 dummy 1.660 0.0988 3.3 August 0.032774 dummy 0.204 0.8388 -25.0 November -0.250118 dummy -1.527 0.1287 0.5 Total t-3 -0.000002120 -2556 -0.164 0.8702 -0.5 Total t+1 0.000001810 -2530 0.086 0.9317 -2.7 Total t+3 0.000010847 -2453 0.619 0.5370

N=183, F Value=2.750, Prob>F=0.0005, Adj R2=0.1398, R2=.2197

° These values are the percentage change in the dependent variable given the mean value of each independent variable. b The coefficient values (*100) are the percentage change in the dependent variable for a one unit change in each independent variable.

' This value is the anti-logarithm value of the intercept minus one, so same day clearing equals zero.

Independent variables in bold type are significant at the ninety-five percent confidence interval.

256 Appendix E

Calculation of Exchange Rates

The Dutch/English mint exchange rates for gold and silver are premised on a few assumptions.

1) The Dutch mint value for silver is taken as banco. Citing Van Dillen, the Wisselbank accepted silver at the "old" Rix dollar value. Accordingly, 25.70 grams fine silver comprised a fifty stiver Rix dollar at the Exchange bank. In 1659, lighter Rix dollars went into general circulation at

24.37 grams fine silver. These current Rix dollars were accepted at the Wisselbank at 48 stivers rather than fifty. The full loss in silver content was 5.18 percent rather than the four percent expressed by the 48 to 50 stiver difference. Van Dillen argues the Exchange Bank kept the difference, "as a slight premium when supplying commercial coinage, and as such also the ducatoon

[gold] and the new rixdollar [24.37 grams fine silver] come into use."(van Dillen, p. 90) In effect, bringing in a new rixdollar bought 48 stivers banco when the silver content should have purchased

47.4 stivers banco. Those bringing silver bullion at 25.70 grams fine silver at fifty stivers banco paid for the pleasure of withdrawing full weight commercial rix dollars.

In 1681, the Dutch introduced a silver guilder (Florin) at 9.61 grams fine silver passing current at 20 stivers. To take the old rixdollar ratio of 25.7 grams fine silver to 50 stivers banco would have doubled the Exchange Bank's fee for minting and raised the value of silver in stivers banco. Unsure as to how the Wisselbank handled bullion after the introduction of the Florin, the author made the conservative assumption. To hold the fee constant, analysis assumes the

Wisselbank continued to accept 50 stivers current as 48 stivers bank with 1.2 percent fees, so two and one half 9.61 gram guilders equate in current stiver purchasing power with a new rixdollar. In

257 turn, 24.025 (9.61*2.5) grams fine silver purchase 47.4 stivers banco. This is the same as the new rixdollar.

2) Dutch gold, the ducat was consistently 3.466 grams fine gold from 1583 to 1816. From

1645 to 1749, the ducat was worth 95 stivers current. Aside from a fee of around one percent, no difference between a ducat current and a ducat banco is known. Adjusting for minting fees, the value of gold is 3.43134 grams fine gold for 95 stivers banco.

3) The English mark was 248.83 grams while the Dutch used a 246.07 gram mark.(Jones

1988, p. 69, footnotes 8 and 9) Adjusting the Dutch silver and gold mint banco values to the

English mark (248.83/246.07) results in 24.294 grams fine silver for 47.4 stivers banco and 3.47 grams fine gold for 95 stivers banco. With 248.83 grams per English mark, an ounce was 31.10375 grams. Converting grams to ounces (dividing by 31.10375) results in .7811 ounces fine silver for

47.4 stivers and .1116 ounces fine gold for 95 stivers. So one ounce fine silver was 60.684 stivers banco, and one ounce of fine gold was 851.254 stivers banco.

4) The Dutch monetary system had two integrated families of coins. One guilder or Florin equaled 20 stivers or 320 penningen (16 penningen = 1 stiver). Also, one pound Flemish equaled

20 schellingen or 240 grooten (12 grooten = 1 schellingen). One pound Flemish equaled six guilders. Converting stivers to schellingen (dividing by 6) brings one ounce fine silver to 10.114 schellingen banco and one ounce fine gold to 141.876 schellingen banco.

5) English silver was minted at 62 pence per ounce troy sterling. At 240 pence per pound sterling, 3.8709 ounces sterling equaled one pound. With sterling being 92.5 percent pure, one ounce fine silver was valued at 67.03 pence at the mint. One ounce fine silver equaled .2793 pounds sterling.

6) The guinea was first minted in 1663 at 22ct. The guinea was originally minted at 20 shillings, but it value was free to rise on the open market. The price floor of gold at the mint was

258 set at twenty shillings. This translates into 970.9 pence per ounce troy fine gold or 4.047 pounds

per ounce fine gold. The guinea immediately carried a market premium of one shilling and soon

after a premium of one shilling, six pence. Jones (1988) argues that 21s 6d had become the official

guinea rate. What authority made this price official remains unclear to this author. The actual

guinea price series, however, clearly shows a 21s 6d price floor for guineas. With each guinea

containing 118.6 grains of fine gold (4.047 guineas equaling an ounce fine gold), the market price of

fine gold based on the guinea was Pguine)* 4.047. Using 21s 6d as the official guinea price, the mint

price of one ounce of fine gold becomes £4.3505.

7) England abolished mint fees in 1666. Of course time was a cost, but no adjustment is

made for coinage costs in England.

The silver mint exchange rate (10.114 Sb/£.2793 ) works out to be 36.2 schellingen banco/pound

sterling for an ounce of fine silver.

The gold mint exchange rate (141.876 Sb/£4.3505 ) works out to be 32.61 schellingen banco/pound

sterling for an ounce of fine gold.

The Dutch Exchange Bank gold to silver ratio calculates to 1:14.03.

The British gold to silver ratio calculates to 1:15.58.

Sources: Feavearyear (1963) and Posthumus (1946).

259 Appendix F

Monthly Average Price of Guineas in Shillings

With 21.5 (21s 6d) Being the Price Floor

1685 Jun 21.625 1691 Jan 21.500 1685 Jul 1691 Feb 21.708 1685 Aug 1691 Mar 21.611 1685 Sep 1691 Apr 21.694 1685 Oct 1691 May 21.681 1685 Nov 1691 Jun 21.514 1685 Dec 21.500 1691 Jul 21.583 1686 Jan 21.583 1691 Aug 21.667 1686 Feb 21.500 1691 Sep 21.708 1686 Mar 21.500 1691 Oct 1686 Apr 21.500 1691 Nov 21.624 1686 May 21.500 1691 Dec 21.750 1686 Jun 21.500 1692 Jan 21.750 1686 Jul 21.500 1692 Feb 21.833 1686 Aug 21.500 1692 Mar 21.750 1686 Sep 21.550 1692 Apr 21.874 1686 Oct 1692 May 21.500 1686 Nov 21.528 1692 Jun 21.694 1686 Dec 21.583 1692 Jul 21.667 1687 Jan 21.528 1692 Aug 21.667 1687 Feb 21.549 1692 Sep 21.667 1687 Mar 21.565 1692 Oct 21.722 1687 Apr 21.542 1692 Nov 21.583 1687 May 21.536 1692 Dec 1687 Jun 21.526 1693 Jan 1687 Jul 21.500 1693 Feb 21.750 1687 Aug 21.555 1693 Mar 21.750 1687 Sep 21.549 1693 Apr 21.750 1687 Oct 21.542 1693 May 21.750 1687 Nov 1693 Jun 22.083 1687 Dec 21.578 1693 Jul 1688 Jan 21.583 1693 Aug 21.792 1688 Feb 21.583 1693 Sep 21.750 1688 Mar 21.583 1693 Oct 1688 Apr 21.500 1693 Nov 21.750 1688 May 21.533 1693 Dec 21.833 1688 Jun 21.550 1694 Jan 21.854 1688 Jul 21.575 1694 Feb 21.854 1688 Aug 1694 Mar 22.000 1688 Sep 21.652 1694 Apr 22.000

260 1688 Oct 21.833 1694 May 22.000 1688 Nov 21.667 1694 Jun 22.000 1688 Dec 21.500 1694 Jul 22.000 1689 Jan 21.500 1694 Aug 22.000 1689 Feb 21.500 1694 Sep 22.000 1689 Mar 21.500 1694 Oct 22.000 1689 Apr 21.500 1694 Nov 22.100 1689 May 21.611 1694 Dec 22.100 1689 Jun 21.542 1695 Jan 22.656 1689 Jul 21.591 1695 Feb 26.512 1689 Aug 21.667 1695 Mar 25.000 1689 Sep 21.833 1695 Apr 25.000 1689 Oct 21.667 1695 May 27.517 1689 Nov 1695 Jun 30.000 1689 Dec 1695 Jul 30.000 1690 Jan 21.667 1695 Aug 30.000 1690 Feb 21.500 1695 Sep 30.000 1690 Mar 21.555 1695 Oct 30.000 1690 Apr 21.500 1695 Nov 30.000 1690 May 21.667 1695 Dec 30.000 1690 Jun 21.624 1696 Jan 30.000 1690 Jul 21.666 1696 Feb 30.000 1690 Aug 21.667 1696 Mar 1690 Sep 21.500 1696 Apr 24.640 1690 Oct 1696 May 1690 Nov 1696 Jun 22.000 1690 Dec 21.500

All Observations come from the Ledgers of Sir Francis Child held by the Royal Bank of

Scotland (CH/149/7,9,10,11). Observations were taken from demand accounts, and the author could find no difference in pricing between credit and debit transactions. Months with multiple observations have been averaged with equal weighing.

261 Appendix G

Stephen Evance's Freight Business

The descriptions of the bullion entries in Evance's Bullion Book were silent regarding geographical sources and destinations; however, Evance did have entries for freight charges.

Evance both paid and charged for freight. A few entries said explicitly, "freight of silver," or

"freight of gold." Most entries simply reported freight. Freight entries were also often accompanied by charges. These were likely custom duties.

In Table 2, columns A and B report the pound sterling sums and the number of observations, per month, of freight payments by Evance. The spring and summer of 1688 was

Evance's busiest period. From March to September of 1688 the goldsmith paid £290 in freight charges over thirty-nine transactions. The period of high freight activity corresponded with

Evance's gold build up. The next spring and summer, Evance again paid for freight and only in slightly smaller amounts. From March through September 1689, Evance spent £258 in only eighteen transactions. Unfortunately, the entries did not differentiate between payments for shipments out and shipments in. The first eight months of 1690 reported little activity. If Evance had been importing gold in 1688 and 1689, he had stopped by 1690.

Evance also charged for freight. Table 2, columns C and D reported the monthly sums and number of observations of freight payments to Evance. Evance's charges to his customers shadowed charges on him, both in pound value and in the number of transactions. The similarities in freight transactions charged to and charged by Stephen Evance suggest the goldsmith-banker acted as a middleman. The few named suppliers of freight services to Evance tended to be captains like Captains Parker, Rowke, Hudson, Stubbs, Guttridge and Peryman. The occasional named

262 purchasers of Evance's freight services were important customers like Alderman Heme, David

Penso, John Lethuillier, Alfonso Rodriguez and fellow goldsmith-banker William Atwell.

Table 2

Stephen Evance's Freight Charges by Month

A B C D Paid for by Evance Received by Evance £ Sums obs £ Sums obs

March, 1688 23.43 3 5.20 2 April 19.29 11 64.73 14 May 6.99 5 9.73 5 June 2.46 4 2.38 5 July 13.23 5 1.15 1 August 28.88 4 September 5.15 3 7.20 3 October 3.41 4 5.00 3 November December 0.45 1 1.45 1 January, 1689 14.30 2 February March 5.23 2 0.73 1 April 8.10 3 May 86.25 5 23.40 6 June 3.12 7 25.79 7 July 22.73 1 August 56.00 1 26.28 2 September 59.88 2 October 8.60 1 November 7.75 2 December January, 1690 February 4.80 1 4.58 4 March 7.42 4 12.58 2 April 1.30 1 May June July August 5.24 1 Totals 381 69 204 60

263 STEPHEN FRANCIS QUINN

429 Prince George Street, #1 Department of Economics Willliamsburg, VA 23185 123 Morton Hall Ph. (804) 229-9597 College of William and Mary E-Mail, [email protected] Williamsburg, VA 23185 Ph. (804) 221-2384 FAX (804) 221-2390

EDUCATION Ph.D., 1994, University of Illinois, Economics M.S., 1994, University of Illinois, Economics B.A., 1988, Rockford College, Economics

PH.D. FIELDS Economic History, Econometrics and Law & Economics

DISSERTATION "Banking Before the Bank: London's Unregulated Goldsmith-Bankers, 1660-1694" Topics of this thesis include: the endogenicity of decentralized inter-banker debt clearing; international arbitrage and bullion flows; analysis of the fractional reserves to assets rations; English law and the transferability of private credit; taxes and the development of private bank notes.

DISSERTATION Professor Larry Neal (Chair), Professor Charles Calomiris and COMMITTEE Professor Paul Newbold

PUBLICATIONS "Tallies or Reserves? Sir Francis Child's Balance Between Capital and Extending Credit to the Crown, 1685-1695," Business and Economic History, Vol. 22, No. 1, (Fall 1994), pp. 39-52.

WORKING "Uncleared Balances and Goldsmith-Bankers: The Endogenicity of PAPERS, Decentralized Inter-Banker Debt Clearing," earlier versions presented PRESENTATIONS, at the Cliometerics Society Meetings (May 1994) and the University AND of London and the Victoria and Albert Museum (Nov. 1993). CONFERENCES "Gold, Silver and the Glorious Revolution; International Bullion Arbitrage and the Origins of the English Gold Standard" to be presented at the American Economic Association Meetings (January 1995), previously presented at Northwestern and Indiana Universities (October 1993) and the London School of Economics (February 1993).

"Tallies or Reserves? Sir Francis Child's Balance Between Capital and Extending Credit to the Crown, 1685-1695," presented at the Business and Economic History Meetings (March 1994).

264 TEACHING AND 1994-95 Visiting Assistant Professor, College of William and Mary. PROFESSIONAL EXPERIENCE 1994 Summer Research Fellow, Institute for Humane Research

1993-94 Research Assistant, Professor Larry Neal, University of Illinois, Analyzed East India Stock Transfers During the South Sea Bubble.

1992-93 Social Sciences Research Council Western Europe Dissertation Fellowship, London.

1990-92 Research Assistant, Bureau of Economic and Business Research, University of Illinois, responsible for published econometric forecasts of state level economic variables.

1989-90 Teaching Assistant, University of Illinois.

AWARDS AND 1988-89 Roberts Trust Fellow, University of Illinois. HONORS 1988 Phi Beta Kappa and Magna Cum Laude, Rockford College.

OTHER "Are You Better Off than You Were Four Years Ago?" Illinois PUBLICATIONS Business Review, (June 1992).

"Forecast," Illinois Economic Outlook, (1992), with Paul Newbold.

REFERENCES Lee Alston, Professor of Economics, Univ. of Illinois, 225 David Kinely Hall 1407 Gregory Street Urbana, IL 61801 Ph. 217-333-4678

Charles Calomiris, Associate Professor of Finance, Univ. of Illinois, 95 Commerce West 1206 South Sixth Street Champaign, IL 61820 Ph. 217-244-8143

Larry Neal, Professor of Economics, Univ. of Illinois, 328A David Kinely Hall 1407 Gregory Street Urbana, IL 61801 Ph. 217-333-4678

265