Domestic Responses to Capital Market Internationalization Under the Gold Standard,1870-1914 Daniel Verdier The internationalization of finance in recent years has brought the world economy to the level it had reached in 1913.With this has come a political debate about the vices and virtues of globalization and an analytic debate about its causes.This article presents an analysis of the earlier period that highlights the importance of political choices in bringing about the internationalization of finance and stresses the variabil- ity of choice among countries experiencing the same global phenomenon. One can hardly open a news magazine nowadays that does not feature an editorial warning against,or urging some kind of adjustment to,capital market globalization. Underlying this"global talk"are the beliefs that capital market internationalization is inevitable,uniform,and irreversible.The scientific debate shows more nuances, focusing mainly on the respective roles played by political and nonpolitical factors. One group of scholars see capital internationalization originating in changes in tech- nology or the international power system or both.They trace its distortionary impact on existing wealth distribution,with the relative immiseration of unskilled labor in the West and,more generally,of holders of immobile factors of production or sectors using these factors intensively.2 Some of them see internationalization as resulting in a weakening of state bureaucracies.3 Another group of scholars place the emphasis instead on state-borrowing preferences as the primary vehicle for global finance,4 on coordination among states as a facilitating mechanism,5 and on the mediating role of state institutions and resulting divergent policy responses. I am pleased to acknowledge the invaluable research assistance of Elizabeth Paulet.I thank John Odell, Louis Pauly,Jonathan Zeitlin,Peter Gourevitch,David Lake,and two anonymous reviewers for valuable suggestions.The research on which this article is based was financed by a grant from the Research Council of the European University Institute.An earlier draft was presented at the annual meeting of the Interna- tional Studies Association,Toronto,in March 1997,and published as an EUI Working Paper,RSC No.97. 1.See Loriaux 1991;Goodman and Pauly 1993:Andrews 1994;and Frieden and Rogowski 1996. 2.See Bates and Lien 1985;and Frieden and Rogowski 1996. 3. See Strange 1986:Webb 1991:Andrews 1994;and Cerny 1995. 4.Haggard and Maxfield 1996. 5.Helleiner 1994.16. 6.See Garrett and Lange 1995;and Garrett 1995. International Organization 52,1.Winter 1998.pp.1-34 1998 by The IO Foundation and the Massachusetts Institute of Technology
Domestic Responses to Capital Market Internationalization Under the Gold Standard, 1870–1914 Daniel Verdier The internationalization of nance in recent years has brought the world economy to the level it had reached in 1913. With this has come a political debate about the vices and virtues of globalization and an analytic debate about its causes. This article presents an analysis of the earlier period that highlights the importance of political choices in bringing about the internationalization of nance and stresses the variability of choice among countries experiencing the same global phenomenon. One can hardly open a news magazine nowadays that does not feature an editorial warning against, or urging some kind of adjustment to, capital market globalization. Underlying this ‘‘global talk’’ are the beliefs that capital market internationalization is inevitable, uniform, and irreversible. The scienti c debate shows more nuances, focusing mainly on the respective roles played by political and nonpolitical factors. One group of scholars see capital internationalization originating in changes in tech- nology or the international power system or both.1 They trace its distortionary impact on existing wealth distribution, with the relative immiseration of unskilled labor in the West and, more generally, of holders of immobile factors of production or sectors using these factors intensively.2 Some of them see internationalization as resulting in a weakening of state bureaucracies.3 Another group of scholars place the emphasis instead on state-borrowing preferences as the primary vehicle for global nance,4 on coordination among states as a facilitating mechanism,5 and on the mediating role of state institutions and resulting divergent policy responses.6 I am pleased to acknowledge the invaluable research assistance of Elizabeth Paulet. I thank John Odell, Louis Pauly, Jonathan Zeitlin, Peter Gourevitch, David Lake, and two anonymous reviewers for valuable suggestions. The research on which this article is based was nanced by a grant from the Research Council of the European University Institute. An earlier draft was presented at the annual meeting of the International Studies Association, Toronto, in March 1997, and published as an EUI Working Paper, RSC No. 97. 1. See Loriaux 1991; Goodman and Pauly 1993; Andrews 1994; and Frieden and Rogowski 1996. 2. See Bates and Lien 1985; and Frieden and Rogowski 1996. 3. See Strange 1986; Webb 1991; Andrews 1994; and Cerny 1995. 4. Haggard and Max eld 1996. 5. Helleiner 1994, 16. 6. See Garrett and Lange 1995; and Garrett 1995. International Organization 52, 1, Winter 1998, pp. 1–34 r 1998 by The IO Foundation and the Massachusetts Institute of Technology
2 International Organization It is not the first time that cross-border capital flows grow out of ordinary propor- tions.A century ago,during the period of the gold standard,the world experienced levels of capital internationalization comparable,if not higher,than current ones.Yet hardly any of the most extreme predictions associated with today's occurrence were realized:internationalization was neither inevitable,uniform,nor notably successful- Britain and France,who embraced internationalization,grew more slowly than Ger- many and the United States,who accepted lower levels of capital market interdepen- dence.And,of course,internationalization was reversed. Two lessons can be drawn from the nineteenth-century stab at capital internation- alization.First,internationalization was a political choice informed by redistribu- tional considerations between rival domestic interests and decided by coalitions on which governments were dependent for support.The choice in favor of openness re- flected the economic preferences of large commercial banks,in alliance with savers in creditor countries,and large firms in debtor countries,all of whom expected to benefit from openness.In contrast,the choice in favor of lesser capital interdepen- dence reflected the preferences of sectors that were expected to lose from openness, including agriculture and sectors with a high density of small-and medium-sized firms. Second,the domestic institutional structure in each country determined the iden- tity of the politically dominant coalition.Decentralized structures allowed potential losers to curb public policies favorable to capital market internationalization,whereas centralized structures allowed expected winners to promote such policies.As a re- sult,economies with centralized states ended up being the most dependent on the international capital market,whereas economies with decentralized states took a less active part in the globalization of finance. This inquiry into the functioning of turn-of-the-century capital markets innovates on two further counts.First,unlike most studies of internationalization,this study parts with comparative statics.7 As generally recognized,internationalization is a process that feeds on itself,calling for a dynamic model.Second,this study modifies the standard approach to the redistributional effects of capital flows.8 Economic in its inspiration,the standard approach points to the cleavage between savers and non- savers,two politically inept groupings on account of size and diffusion.Yet concerns over the wealth effects of financial fows have not remained uniformly unvoiced. They were articulated in a majority of countries along another line of cleavage-the center-periphery cleavage-pitting each central government against its respective local governments. The first part of the article presents the theoretical framework,the second part the argument,and the third part the evidence.The conclusion will summarize the find- ings and amplify the themes of this introduction. 7.See Andrews 1994;and Frieden and Rogowski 1996. 8.Frieden 1991
It is not the rst time that cross-border capital ows grow out of ordinary proportions. A century ago, during the period of the gold standard, the world experienced levels of capital internationalization comparable, if not higher, than current ones. Yet hardly any of the most extreme predictions associated with today’s occurrence were realized: internationalization was neither inevitable, uniform, nor notably successful— Britain and France, who embraced internationalization, grew more slowly than Ger- many and the United States, who accepted lower levels of capital market interdepen- dence. And, of course, internationalization was reversed. Two lessons can be drawn from the nineteenth-century stab at capital internation- alization. First, internationalization was a political choice informed by redistributional considerations between rival domestic interests and decided by coalitions on which governments were dependent for support. The choice in favor of openness re- ected the economic preferences of large commercial banks, in alliance with savers in creditor countries, and large rms in debtor countries, all of whom expected to bene t from openness. In contrast, the choice in favor of lesser capital interdepen- dence re ected the preferences of sectors that were expected to lose from openness, including agriculture and sectors with a high density of small- and medium-sized rms. Second, the domestic institutional structure in each country determined the identity of the politically dominant coalition. Decentralized structures allowed potential losersto curb public policies favorable to capital market internationalization, whereas centralized structures allowed expected winners to promote such policies. As a re- sult, economies with centralized states ended up being the most dependent on the international capital market, whereas economies with decentralized states took a less active part in the globalization of nance. This inquiry into the functioning of turn-of-the-century capital markets innovates on two further counts. First, unlike most studies of internationalization, this study parts with comparative statics.7 As generally recognized, internationalization is a process that feeds on itself, calling for a dynamic model. Second, this study modi es the standard approach to the redistributional effects of capital ows.8 Economic in its inspiration, the standard approach points to the cleavage between savers and non- savers, two politically inept groupings on account of size and diffusion. Yet concerns over the wealth effects of nancial ows have not remained uniformly unvoiced. They were articulated in a majority of countries along another line of cleavage—the center–periphery cleavage—pitting each central government against its respective local governments. The rst part of the article presents the theoretical framework, the second part the argument, and the third part the evidence. The conclusion will summarize the ndings and amplify the themes of this introduction. 7. See Andrews 1994; and Frieden and Rogowski 1996. 8. Frieden 1991. 2 International Organization
Capital Market Internationalization 3 The Model Internationalization has a dynamic,historical character:it feeds on itself.Although all studies of internationalization acknowledge this feature,they fail to draw the appropriate methodological consequence. When inquiring into the origins of internationalization,all authors concur in listing two sets of determining factors:(1)technological innovations yielding reduc- tions in cross-border transaction costs,and (2)government policies easing cross- border capital flows.Although technological innovations may,in some circum- stances,be viewed as exogenous shocks,government policies are unequivocally endogenous to the mechanism of internationalization.Causal models of the compara- tive-statics type cannot supply the proper explanation;internationalization would be serving both as independent and dependent variable-an axiomatic non sequitur for this kind of model.That circularity must instead be explicitly tackled through a dynamic model. The need for a dynamic approach is far from being universally shared.Studies of internationalization that seek to explain internationalization,instead,try to fit it into the Procrustean bed of comparative statics,with circularity being avoided in one of two ways:(1)technological determinism,which makes technological innovation exogenous and uses it to determine the model:10 and(2)structural determinism, which sees internationalization as the suboptimal outcome of states'competitive bidding for international capital.Technological determinism,however,goes against recent developments in "new growth"theory,making innovation a process that is endogenous to firms'profit-maximizing strategies,which states can influence through diverse policies.2 With respect to the second claim only the future will tell whether the deregulatory race between states for capital is structural or contingent on revers- ible domestic changes.The fact that the same countries already went through a simi- lar race under the gold standard weakens considerably the claim that today's compe- tition is here to stay. A simple dynamic model features a two-period decision process,allowing for a change in the state of nature in between.In the first period,the government is con- fronted with a technological innovation that promises to ease internationalization in the second period if the regulatory status quo is left unchanged and if the innovation is allowed to move down its learning curve,that is,be adopted,diffused,and im- proved through learning by doing.The government decides on the basis of expected return and opportunity cost,taking into account what other countries might do,whether to check the innovation by means of countervailing policies or let it mature.If the 9.See Goodman and Pauly 1993;Andrews 1994;Frieden and Rogowski 1996;and Haggard and Maxfield 1996. 10.Frieden and Rogowski 1996. 11.Andrews 1994. 12.Roemer 1994
The Model Internationalization has a dynamic, historical character: it feeds on itself. Although all studies of internationalization acknowledge this feature, they fail to draw the appropriate methodological consequence. When inquiring into the origins of internationalization, all authors concur in listing two sets of determining factors: (1) technological innovations yielding reductions in cross-border transaction costs, and (2) government policies easing cross- border capital ows.9 Although technological innovations may, in some circumstances, be viewed as exogenous shocks, government policies are unequivocally endogenous to the mechanism of internationalization. Causal models of the comparative-statics type cannot supply the proper explanation; internationalization would be serving both as independent and dependent variable—an axiomatic non sequitur for this kind of model. That circularity must instead be explicitly tackled through a dynamic model. The need for a dynamic approach is far from being universally shared. Studies of internationalization that seek to explain internationalization, instead, try to t it into the Procrustean bed of comparative statics, with circularity being avoided in one of two ways: (1) technological determinism, which makes technological innovation exogenous and uses it to determine the model;10 and (2) structural determinism, which sees internationalization as the suboptimal outcome of states’ competitive bidding for international capital.11 Technological determinism, however, goes against recent developments in ‘‘new growth’’ theory, making innovation a process that is endogenousto rms’ pro t-maximizing strategies, which states can in uence through diverse policies.12 With respect to the second claim only the future will tell whether the deregulatory race between states for capital is structural or contingent on reversible domestic changes. The fact that the same countries already went through a similar race under the gold standard weakens considerably the claim that today’s competition is here to stay. A simple dynamic model features a two-period decision process, allowing for a change in the state of nature in between. In the rst period, the government is confronted with a technological innovation that promises to ease internationalization in the second period if the regulatory status quo is left unchanged and if the innovation is allowed to move down its learning curve, that is, be adopted, diffused, and im- proved through learning by doing. The government decides on the basis of expected return and opportunity cost, taking into account what other countries might do, whether to check the innovation by means of countervailing policies or let it mature. If the 9. See Goodman and Pauly 1993; Andrews 1994; Frieden and Rogowski 1996; and Haggard and Max eld 1996. 10. Frieden and Rogowski 1996. 11. Andrews 1994. 12. Roemer 1994. Capital Market Internationalization 3
4 International Organization innovation is aborted,then internationalization will not ensue,and the degree of openness will remain unchanged in the second period.If,instead,the innovation is allowed to mature,internationalization will proceed,and the degree of openness of the capital market will be higher in the second period than in the first. I now amend the story to make space for coalitions and institutions.Assume that the government decision is the outcome of a policy process in which the most orga- nized interests get to impose their policy preferences.Private interests in the first period anticipate the future distributional effects of the initial innovation were it to run its course in the second period.Anticipated losers will try to nip the innovation in the bud if they can politically organize.Whether or not they can organize depends on the nature of extant domestic institutions (or a subset thereof).To the extent that countries have different institutions,the degree of internationalization chosen by each government will differ,reflecting institutional variation. One advantage of setting up the problem this way is not to confuse the outcome- the degree of openness to capital flows achieved by each country-with the cause-an exogenous innovation promising gains and losses tomorrow to interests that can anticipate its wealth effects and act accordingly now.Internationalization is not or- dained in the present formulation,but unlikely to proceed very far if potential losers enjoy political power.Another advantage is to differentiate the initial technological innovation,which may be treated as exogenous to politics,from the price shock that will result from the widespread adoption of the innovation,which is endogenous.3A possible drawback of the present formulation comes from its perhaps excessive sim- plicity;the process is reduced to only two periods,with actors graced with the gift of perfect foresight.Reality may afford many more periods,with individuals and gov- ernments exhibiting a present foresight limited to the next period alone and a present latitude constrained by decisions made in the prior period.The two-stage set-up, however,with its perfect foresight implication,makes the presentation of the mate- rial clearer. Applying this model to the case of capital market internationalization under the gold standard will require completing three successive steps:(1)extract from the late-nineteenth-century historical reality the exogenous technological changes that had the potential to increase cross-border investment in all countries;(2)derive the potential domestic losers from this innovation,assess their nonmarket options in light of their institutional power,and then derive each country's policy response;and (3)derive the predicted degree of openness to international capital flows that each country should have eventually reached according to the model.The next part of the article presents the three-step argument,and the third part confronts it with the his- torical record. 13.The price shock is taken as exogenous in Rogowski's 1989 setup and also in Frieden and Rogowski 1996.In Rogowski's story the exogenous price shock increases the wealth and power of the supporters of internationalization,thereby leading to greater policy openness.In the present story,the price shock comes too late,if at all,to help the partisans of internationalization prevail over their opponents
innovation is aborted, then internationalization will not ensue, and the degree of openness will remain unchanged in the second period. If, instead, the innovation is allowed to mature, internationalization will proceed, and the degree of openness of the capital market will be higher in the second period than in the rst. I now amend the story to make space for coalitions and institutions. Assume that the government decision is the outcome of a policy process in which the most orga- nized interests get to impose their policy preferences. Private interests in the rst period anticipate the future distributional effects of the initial innovation were it to run its course in the second period. Anticipated losers will try to nip the innovation in the bud if they can politically organize. Whether or not they can organize depends on the nature of extant domestic institutions (or a subset thereof). To the extent that countries have different institutions, the degree of internationalization chosen by each government will differ, re ecting institutional variation. One advantage of setting up the problem this way is not to confuse the outcome— the degree of opennessto capital ows achieved by each country—with the cause—an exogenous innovation promising gains and losses tomorrow to interests that can anticipate its wealth effects and act accordingly now. Internationalization is not or- dained in the present formulation, but unlikely to proceed very far if potential losers enjoy political power. Another advantage is to differentiate the initial technological innovation, which may be treated as exogenous to politics, from the price shock that will result from the widespread adoption of the innovation, which is endogenous.13 A possible drawback of the present formulation comes from its perhaps excessive sim- plicity; the process is reduced to only two periods, with actors graced with the gift of perfect foresight. Reality may afford many more periods, with individuals and gov- ernments exhibiting a present foresight limited to the next period alone and a present latitude constrained by decisions made in the prior period. The two-stage set-up, however, with its perfect foresight implication, makes the presentation of the material clearer. Applying this model to the case of capital market internationalization under the gold standard will require completing three successive steps: (1) extract from the late-nineteenth-century historical reality the exogenous technological changes that had the potential to increase cross-border investment in all countries; (2) derive the potential domestic losers from this innovation, assess their nonmarket options in light of their institutional power, and then derive each country’s policy response; and (3) derive the predicted degree of openness to international capital ows that each country should have eventually reached according to the model. The next part of the article presents the three-step argument, and the third part confronts it with the historical record. 13. The price shock is taken as exogenousin Rogowski’s 1989 setup and also in Frieden and Rogowski 1996. In Rogowski’s story the exogenous price shock increases the wealth and power of the supporters of internationalization, thereby leading to greater policy openness. In the present story, the price shock comes too late, if at all, to help the partisans of internationalization prevail over their opponents. 4 International Organization
Capital Market Internationalization 5 The Argument Changes in Banking Technology and the Demand for Short Assets14 The surge in capital flows witnessed under the gold standard,I argue in this and the next two sections,originated in a demand for foreign investments,not merely long, as usually noted,but more importantly short.Banks in the late-nineteenth century had a need for short assets,which the international capital market could supply.In this and the next sections I focus on the demand and supply side of short-term assets. Banking until the mid-nineteenth century relied on personal connections.Bankers would borrow from and lend to individuals whom they knew well,either because they lived in the same towns or because borrowers and bank shareholders were often the same people-a relation that Naomi Lamoreaux has appropriately dubbed"in- sider lending."15 Philip Cottrell wrote of the English country banks: Until the 1880s English country banks were products of the localities and regions that they served;customers and shareholders were frequently the same people. The bank's constituencies both owned the banks and did business with them. Directors and managers knew their customers well and with prudence and local knowledge were prepared to go beyond the bounds of short-term lending.16 Where local,personal connections were unavailing,banks would simply not lend to enterprises.Gustav Mevissen,a co-director of the Bank of Darmstadt,made the point with utmost clarity in an instruction to the bank management written at midcentury: The task of our bank is not to attract the business of industrial and commercial enterprise in general.On the contrary,it will be our mission to establish contact with all government institutions,joint-stock companies,and wealthy private per- sons in the hope of obtaining as large a share of the business of governments,of princes and principates,as well as joint-stock companies and wealthy private persons as possible.17 By the middle of the century banking evolved into a more impersonal and profes- sional activity under the pressure of two circumstances.The first circumstance was the rise in individual deposits and the simultaneous decline of bank equity and note issuing.Until midcentury,there were only two main ways of procuring capital in 14.Assets are the left-hand side of a balance-sheet,and liabilities are the right-hand side.Assers are investments,which banks finance with resources or liabilities.Assets and liabilities are arranged ac- cording to maturity.Short assets typically include cash,loans to the stock market,short-term government debt.three-to-six-month credit advances (also called overdrafis),and commercial paper (bills of ex- change,acceptances).A bill of exchange is a buyer's promise to pay in three months:the seller can cash it immediately with a bank.An acceptance is an international bill of exchange.Long assets include long- term government debt,participations in other joint-stock companies,and all loans or advances with a maturity longer than six months.Short liabilities include deposits,positive current accounts.and,in some cases,notes.Long liabilities include equity (capital and reserves). 15.Lamoreaux 1994. 16.Cottrel11992.53 17.Tily1986,121
The Argument Changes in Banking Technology and the Demand for Short Assets 14 The surge in capital ows witnessed under the gold standard, I argue in this and the next two sections, originated in a demand for foreign investments, not merely long, as usually noted, but more importantly short. Banks in the late-nineteenth century had a need for short assets, which the international capital market could supply. In this and the next sections I focus on the demand and supply side of short-term assets. Banking until the mid-nineteenth century relied on personal connections. Bankers would borrow from and lend to individuals whom they knew well, either because they lived in the same towns or because borrowers and bank shareholders were often the same people—a relation that Naomi Lamoreaux has appropriately dubbed ‘‘insider lending.’’ 15 Philip Cottrell wrote of the English country banks: Until the 1880s English country banks were products of the localities and regions that they served; customers and shareholders were frequently the same people. The bank’s constituencies both owned the banks and did business with them. Directors and managers knew their customers well and with prudence and local knowledge were prepared to go beyond the bounds of short-term lending.16 Where local, personal connections were unavailing, banks would simply not lend to enterprises. Gustav Mevissen, a co-director of the Bank of Darmstadt, made the point with utmost clarity in an instruction to the bank management written at midcentury: The task of our bank is not to attract the business of industrial and commercial enterprise in general. On the contrary, it will be our mission to establish contact with all government institutions, joint-stock companies, and wealthy private per- sons in the hope of obtaining as large a share of the business of governments, of princes and principates, as well as joint-stock companies and wealthy private persons as possible.17 By the middle of the century banking evolved into a more impersonal and professional activity under the pressure of two circumstances. The rst circumstance was the rise in individual deposits and the simultaneous decline of bank equity and note issuing. Until midcentury, there were only two main ways of procuring capital in 14. Assets are the left-hand side of a balance-sheet , and liabilities are the right-hand side. Assets are investments, which banks nance with resources or liabilities. Assets and liabilities are arranged ac- cording to maturity. Short assets typically include cash, loans to the stock market, short-term government debt, three-to-six-month credit advances (also called overdrafts), and commercial paper (bills of ex- change, acceptances). A bill of exchange is a buyer’s promise to pay in three months; the seller can cash it immediately with a bank. An acceptance is an international bill of exchange. Long assets include longterm government debt, participations in other joint-stock companies, and all loans or advances with a maturity longer than six months. Short liabilities include deposits, positive current accounts, and, in some cases, notes. Long liabilities include equity (capital and reserves). 15. Lamoreaux 1994. 16. Cottrell 1992, 53. 17. Tilly 1986, 121. Capital Market Internationalization 5
6 International Organization large quantity-note issuing,which in many countries already was,or about to be- come,regulated by government,and equity;deposits played a marginal role.By midcentury,however,the spread of industrialization led to a relative enlargement of the saving public and to a shift of the public's preferences from cash to checks(or credit transfers)for transaction purposes.Demand for deposit accounts,long and short,grew so much that it became thinkable for private bankers to finance lending with deposits taken from numerous individuals with whom they had no prior or other dealings.Banks saw in deposit-taking a way of improving profitability.Depositors typically earned less than bank shareholders;by increasing the share of deposits relative to capital,banks could increase earning on capital.The second part of the nineteenth century thus saw in most countries a rush toward deposit banking.Lead- ing in this new type of banking were the clearing banks in England and Wales,the Credit Lyonnais in France,and the Deutsche Bank in Germany. Deposits grew in the economy as a whole relative to gross national product (GNP)and in the banking sector relative to other banking resources.I8 The rising importance of deposits created a liquidity problem for the banks for two reasons. First,deposits were short-term assets.Although banks tried to lengthen the maturity of deposits by creating term deposits,according to which early withdrawals carried penalties,they could never prevent depositors confronted with the danger of a bank run from cashing their savings rather than facing the risk of losing them all.Second, unlike stockholders,depositors had no insider information on the good management and solvency of the bank.They could not monitor the management nor draw a reliable assessment of the bank's solvency.They relied instead on rumor,with the result that banks were subject to "sunspot"panics,that is,runs on deposits with no other rationale than each depositor's fear of being the victim of other depositors'fear of runs.A run on a bank would trigger a run on other banks if it were believed that the collapse of the first bank would weaken the liquidity of the others,as was often the case.19 The liquidity problem arising from the generalization of deposits was com- pounded by another circumstantial change,taking the form of the progressive replace- ment of the bill of exchange by overdrafts.20 The substitution was caused by multiple separate changes,including the reduction in transport costs,changes in sale and payment practices (buyers paying cash to take advantage of discounts),the tele- graphic transfer of payments,and firms relying on checks in general to effect pay- ment.21 Overdrafts were better remunerated than bills,but they were easily renewed and thus less liquid.Unlike bills,moreover,advances could not be readily recycled through rediscounting at the central bank. 18.Data on commercial and savings bank deposits are found in Mitchell 1983,1992;for Australia, Butlin,Hall,and White 1971;and,for Denmark,in Johansen 1985.Data on financial assets are found in Goldsmith 1969. 19.The liquidity problems arising from the greater importance taken by deposits in banks resources are underscored in Lamoreaux 1994.107. 20.The terms bill of exchange and overdraft are defined in footnote 14. 21.Cottrell1980,204
large quantity—note issuing, which in many countries already was, or about to be- come, regulated by government, and equity; deposits played a marginal role. By midcentury, however, the spread of industrialization led to a relative enlargement of the saving public and to a shift of the public’s preferences from cash to checks (or credit transfers) for transaction purposes. Demand for deposit accounts, long and short, grew so much that it became thinkable for private bankers to nance lending with deposits taken from numerous individuals with whom they had no prior or other dealings. Banks saw in deposit-taking a way of improving pro tability. Depositors typically earned less than bank shareholders; by increasing the share of deposits relative to capital, banks could increase earning on capital. The second part of the nineteenth century thus saw in most countries a rush toward deposit banking. Leading in this new type of banking were the clearing banks in England and Wales, the Cre´dit Lyonnais in France, and the Deutsche Bank in Germany. Deposits grew in the economy as a whole relative to gross national product (GNP) and in the banking sector relative to other banking resources.18 The rising importance of deposits created a liquidity problem for the banks for two reasons. First, deposits were short-term assets. Although banks tried to lengthen the maturity of deposits by creating term deposits, according to which early withdrawals carried penalties, they could never prevent depositors confronted with the danger of a bank run from cashing their savings rather than facing the risk of losing them all. Second, unlike stockholders, depositors had no insider information on the good management and solvency of the bank. They could not monitor the management nor draw a reliable assessment of the bank’s solvency. They relied instead on rumor, with the result that banks were subject to ‘‘sunspot’’ panics, that is, runs on deposits with no other rationale than each depositor’s fear of being the victim of other depositors’ fear of runs. A run on a bank would trigger a run on other banks if it were believed that the collapse of the rst bank would weaken the liquidity of the others, as was often the case.19 The liquidity problem arising from the generalization of deposits was com- pounded by another circumstantial change, taking the form of the progressive replace- ment of the bill of exchange by overdrafts.20 The substitution was caused by multiple separate changes, including the reduction in transport costs, changes in sale and payment practices (buyers paying cash to take advantage of discounts), the tele- graphic transfer of payments, and rms relying on checks in general to effect pay- ment.21 Overdrafts were better remunerated than bills, but they were easily renewed and thus less liquid. Unlike bills, moreover, advances could not be readily recycled through rediscounting at the central bank. 18. Data on commercial and savings bank deposits are found in Mitchell 1983, 1992; for Australia, Butlin, Hall, and White 1971; and, for Denmark, in Johansen 1985. Data on nancial assets are found in Goldsmith 1969. 19. The liquidity problems arising from the greater importance taken by deposits in banksresources are underscored in Lamoreaux 1994, 107. 20. The terms bill of exchange and overdraft are de ned in footnote 14. 21. Cottrell 1980, 204. 6 International Organization
Capital Market Internationalization 7 Relying on more volatile resources (deposits)to finance less liquid assets (over- drafts),banks were caught in a liquidity squeeze.They became aware of it in the wake of a string of banking crises,during which deposits were withdrawn in ex- change for coin and central bank notes.Hence,Michael Collins notes that after each crisis in England and Wales,the most severe being the crash of the City of Glasgow Bank in 1878,the banks tended to maintain a higher proportion of very liquid as- sets.22 Jean Bouvier notes that the crash of 1882 in France served to disqualify loans to industry in the eyes of Henri Germain,the director of the Credit Lyonnais.2 The standard response to the liquidity crisis was for banks to move to a form of banking that was safer.This meant developing standard lending procedures and thus more interchangeable and negotiable instruments,which could be used as secondary forms of liquidity.But since standardization could more easily be achieved in short- term lending than in long-term lending,standardization amounted to shortening the maturity of most assets:commercial banks would abandon their initial universality, specializing instead in short-term lending.24 Short,standardized assets had the advan- tage of being readily disposable in periods of crisis.But they had two drawbacks. First,they yielded lower profits.Second,safe paper was hard to find,especially now that overdrafts were displacing trade bills.In London,Paris,Milan,and Berlin,bank- ers complained about a persistent shortage in"good"paper,increasingly limited to international acceptances,that is,to bills generated by the settlement of international trade.25 The important role played by good paper in the smooth functioning of the monetary market placed these international centers into competition for the natural- ization of the market for acceptances.2 This shortage was also responsible for the revival of competition,noted in several countries,between the central bank and the deposit banks.27 The higher demand for good paper elicited new profit-making strategies amalgamation,centralization,and internationalization.All three aimed at relieving the need for good paper through greater productivity and higher volume.Amalgam- ation allowed banks to take advantage of the internal scale economies released by the move toward standardization.It is important to note that no such economies of scale existed during the first half of the century,when banking was still a matter of per- sonal connections and when profits sanctioned investments in high-yield,low- volume loans to local industries.Only after banks had been forced to abandon their long-term positions in local firms and to compensate for low yield through high volume did amalgamation become a profitable strategy.Amalgamation reduced bank capital requirements,improving earning potential.Amalgamation also allowed merg- 22.Collins1991,41. 23.Bouvier 1968,221.See also Levy-Leboyer 1976.462. 24.See Bouvier 1968,162;and Lamoreaux 1994.89. 25.See Conti 1993,311:Polsi 1996.127;and Riesser 1911,306. 26.The Deutsche Bank was organized in 1870 by a group of private bankers to capture a greater share of the foreign short-term credit and payments business:Tilly 1991.93.Broz argues that the Federal Reserve Bank was established to develop a market for acceptances in New York:Broz 1997. 27.On Britain,see De Cecco 1974,101;and Ziegler 1990,135;on France,see Bouvier 1973.160:and Lescure 1995,318;and on Belgium,see Kauch 1950,235,260
Relying on more volatile resources (deposits) to nance less liquid assets (over- drafts), banks were caught in a liquidity squeeze. They became aware of it in the wake of a string of banking crises, during which deposits were withdrawn in ex- change for coin and central bank notes. Hence, Michael Collins notes that after each crisis in England and Wales, the most severe being the crash of the City of Glasgow Bank in 1878, the banks tended to maintain a higher proportion of very liquid assets.22 Jean Bouvier notes that the crash of 1882 in France served to disqualify loans to industry in the eyes of Henri Germain, the director of the Cre´dit Lyonnais.23 The standard response to the liquidity crisis was for banks to move to a form of banking that was safer. This meant developing standard lending procedures and thus more interchangeable and negotiable instruments, which could be used as secondary forms of liquidity. But since standardization could more easily be achieved in shortterm lending than in long-term lending, standardization amounted to shortening the maturity of most assets: commercial banks would abandon their initial universality, specializing instead in short-term lending.24 Short,standardized assets had the advantage of being readily disposable in periods of crisis. But they had two drawbacks. First, they yielded lower pro ts. Second, safe paper was hard to nd, especially now that overdrafts were displacing trade bills. In London, Paris, Milan, and Berlin, bank- ers complained about a persistent shortage in ‘‘good’’ paper, increasingly limited to international acceptances, that is, to bills generated by the settlement of international trade.25 The important role played by good paper in the smooth functioning of the monetary market placed these international centers into competition for the naturalization of the market for acceptances.26 This shortage was also responsible for the revival of competition, noted in several countries, between the central bank and the deposit banks.27 The higher demand for good paper elicited new pro t-making strategies— amalgamation, centralization, and internationalization. All three aimed at relieving the need for good paper through greater productivity and higher volume. Amalgam- ation allowed banks to take advantage of the internal scale economies released by the move toward standardization. It is important to note that no such economies of scale existed during the rst half of the century, when banking was still a matter of per- sonal connections and when pro ts sanctioned investments in high-yield, low- volume loans to local industries. Only after banks had been forced to abandon their long-term positions in local rms and to compensate for low yield through high volume did amalgamation become a pro table strategy. Amalgamation reduced bank capital requirements, improving earning potential. Amalgamation also allowed merg- 22. Collins 1991, 41. 23. Bouvier 1968, 221. See also Le´vy-Leboyer 1976, 462. 24. See Bouvier 1968, 162; and Lamoreaux 1994, 89. 25. See Conti 1993, 311; Polsi 1996, 127; and Riesser 1911, 306. 26. The Deutsche Bank was organized in 1870 by a group of private bankers to capture a greater share of the foreign short-term credit and payments business; Tilly 1991, 93. Broz argues that the Federal Reserve Bank was established to develop a market for acceptances in New York; Broz 1997. 27. On Britain, see De Cecco 1974, 101; and Ziegler 1990, 135; on France, see Bouvier 1973, 160; and Lescure 1995, 318; and on Belgium, see Kauch 1950, 235, 260. Capital Market Internationalization 7
8 International Organization ing banks to rationalize their asset portfolio,taking over the best paper held by their competitors and liquidating less desirable items.28 Amalgamation naturally led to centralization-the relocation of bank headquar- ters in financial centers.Centralization allowed banks to capture external scale econo- mies:central clearing allowed banks to economize on working balances,and the greater breadth of the market increased the liquidity of security issues.29 Moreover. centralization allowed banks to enter lucrative lines of activity,such as the underwrit- ing of government and railroad loans.Centralization finally led to internationaliza- tion,since among these government loans figured those to foreign governments,until then the exclusive province of prestigious private banking houses.30 In sum,the liquidity squeeze that characterized commercial banking during the second half of the nineteenth century created a demand for short assets and led banks to pursue a profit-making strategy geared to the capture of a larger share of the relatively diminishing supply of short assets. The Gold Standard and the Supply of Short Assets The gold standard gave a boost to international capital markets,making possible an absolute increase in the coveted short instruments.It did so directly,though to a small extent,by assisting the market for acceptances,and indirectly,yet to a greater extent,by giving a boost to long-term credits. The gold standard first assisted the market for short-term capital,that of interna- tional acceptances,by reducing the currency risk.We do not know to what extent. Surely,the currency risk was already low under preceding bimetallism.Moreover, the market for international acceptances was,from 1870 on,monopolized by Lon- don;international acceptances did play substitute for vanishing bills of exchange in Britain,but not elsewhere.The greatest contribution to the uniform supply of short assets across financial centers,I believe,was more indirect;it was a spin-off of the boom in long-term foreign investment.I first develop the impact of the gold standard on long-term foreign investment and then its related effects on banks'short assets. The gold standard stimulated the long-term financial market.Operating as a com- mitment rule,according to which gold countries pledged to maintain a fixed parity between one unit of their currency and a given quantity of gold,the gold standard made possible the systematic transfer of capital from capital-rich and slow-growing economies to capital-poor and fast-growing economies.31 Countries seeking long- 28.Lamoreaux 1994,144. 29.Kindleberger 1978,72-75. 30.See Bouvier 1968:and Cameron 1991.14-16. 31.The gold standard is viewed by Bordo and Kydland 1995 as a solution to the time-inconsistency problem analyzed by Kydland and Prescott 1977.In the initial story,a government with discretion over the formulation of monetary policy will have an incentive to engineer a surprise inflation to stimulate employ- ment.Absent a binding commitment,the public will come to anticipate the outcome,leading to an infla- tionary equilibrium.A solution to the dilemma is for the government to waive discretion and pledge to abide by a binding rule.A variation on that story,one that makes time-inconsistency relevant to the gold standard,runs like this:a government with discretion over its monetary and fiscal policy will have an
ing banks to rationalize their asset portfolio, taking over the best paper held by their competitors and liquidating less desirable items.28 Amalgamation naturally led to centralization—the relocation of bank headquartersin nancial centers. Centralization allowed banks to capture external scale econo- mies: central clearing allowed banks to economize on working balances, and the greater breadth of the market increased the liquidity of security issues.29 Moreover, centralization allowed banks to enter lucrative lines of activity,such asthe underwriting of government and railroad loans. Centralization nally led to internationalization, since among these government loans gured those to foreign governments, until then the exclusive province of prestigious private banking houses.30 In sum, the liquidity squeeze that characterized commercial banking during the second half of the nineteenth century created a demand for short assets and led banks to pursue a pro t-making strategy geared to the capture of a larger share of the relatively diminishing supply of short assets. The Gold Standard and the Supply of Short Assets The gold standard gave a boost to international capital markets, making possible an absolute increase in the coveted short instruments. It did so directly, though to a small extent, by assisting the market for acceptances, and indirectly, yet to a greater extent, by giving a boost to long-term credits. The gold standard rst assisted the market for short-term capital, that of international acceptances, by reducing the currency risk. We do not know to what extent. Surely, the currency risk was already low under preceding bimetallism. Moreover, the market for international acceptances was, from 1870 on, monopolized by Lon- don; international acceptances did play substitute for vanishing bills of exchange in Britain, but not elsewhere. The greatest contribution to the uniform supply of short assets across nancial centers, I believe, was more indirect; it was a spin-off of the boom in long-term foreign investment. I rst develop the impact of the gold standard on long-term foreign investment and then its related effects on banks’short assets. The gold standard stimulated the long-term nancial market. Operating as a com- mitment rule, according to which gold countries pledged to maintain a xed parity between one unit of their currency and a given quantity of gold, the gold standard made possible the systematic transfer of capital from capital-rich and slow-growing economies to capital-poor and fast-growing economies.31 Countries seeking long- 28. Lamoreaux 1994, 144. 29. Kindleberger 1978, 72–75. 30. See Bouvier 1968; and Cameron 1991, 14–16. 31. The gold standard is viewed by Bordo and Kydland 1995 as a solution to the time-inconsistency problem analyzed by Kydland and Prescott 1977. In the initial story, a government with discretion over the formulation of monetary policy will have an incentive to engineer a surprise in ation to stimulate employ- ment. Absent a binding commitment, the public will come to anticipate the outcome, leading to an in ationary equilibrium. A solution to the dilemma is for the government to waive discretion and pledge to abide by a binding rule. A variation on that story, one that makes time-inconsistency relevant to the gold standard, runs like this: a government with discretion over its monetary and scal policy will have an 8 International Organization
Capital Market Internationalization 9 term foreign capital paid lower interest rates on loans contracted in London,Paris, Berlin,and other financial centers if they adhered to the gold standard.32 In the Rus- sian and Austrian empires,partisans of industrialization thought that industrialization could be a speedy process if foreign capital intervened to stimulate it-foreign capi- tal would come by adopting the gold standard.33 Reflecting on the experience of Spain,which suspended gold convertibility in 1883,Pablo Martin-Acena argued that,by staying out of the gold standard,"Spain missed on growth."34 Not only did imports of foreign capital cease from 1883 until 1906,when a new administration finally opted for a return to gold,but yields on the public debt were "'consistently maintained above British,French,and even Italian yields."35 The generalization of the gold standard coincided with a rise in international capi- tal outflows to levels that were never approached before and have never been ap- proached since.Paul Bairoch's estimates for capital fows for all net creditor coun- tries show a slowdown in the depression decades of the gold standard,followed by an unprecedented surge after 1900: 1840-1870:2.5-3.5 percent GNP 1870-1900:1.5-2.0 percent GNP 1900-1913:5.5 percent GNP36 Comparable data for the 1920s,1960s,and 1970s were below 1 percent. Most of this investment,about three-fourths,came from three countries (United Kingdom,France,and Germany),which were running persistent current account surpluses by generating savings in excess of domestic investment.Relative to total domestic savings,net capital outflows in 1910 represented 52 percent for the United Kingdom and 15 percent for France.37 The rest was contributed by the Netherlands, Belgium,Switzerland,and,toward the end of the period,Sweden.Most of this invest- ment went to a few countries-the United States,Canada,Australasia,Argentina, Brazil,Mexico,Russia,Spain,Portugal,Italy,Austria-Hungary,and the Scandina- vian countries.38 What made foreign investment so popular among savers in Britain,France,Ger- many,and other creditor countries was its greater safety,at equivalent yield,than domestic paper.In the case of Britain,Michael Edelstein found that overseas returns exceeded home returns over the years 1870-1913;he also found that overseas re- incentive to borrow and then default on its debt through inflation or suspension of payments.Anticipating default,bond holders will either ask for a higher interest rate or not purchase govemment debt.A solution to the dilemma is for the government to commit to gold convertibility at a fixed rate-a transparent and simple rule:Bordo and Kydland 1995.Bordo and Schwartz found that those countries that adhered to the gold standard rule generally had lower fiscal deficits,more stable money growth,and lower inflation rates than those that did not;Bordo and Schwartz 1994. 32.Bordo and Rockoff 1995,18. 33.De Cecco 1974,52 34.Martin-Acena 1994,160. 35.Ibid.,144. 36.Bairoch1976.103 37.Green and Urquhart 1976.241,244. 38.Cameron1991.13
term foreign capital paid lower interest rates on loans contracted in London, Paris, Berlin, and other nancial centers if they adhered to the gold standard.32 In the Russian and Austrian empires, partisans of industrialization thought that industrialization could be a speedy process if foreign capital intervened to stimulate it—foreign capital would come by adopting the gold standard.33 Re ecting on the experience of Spain, which suspended gold convertibility in 1883, Pablo Martin-Acen˜a argued that, by staying out of the gold standard, ‘‘Spain missed on growth.’’ 34 Not only did imports of foreign capital cease from 1883 until 1906, when a new administration nally opted for a return to gold, but yields on the public debt were ‘‘consistently maintained above British, French, and even Italian yields.’’ 35 The generalization of the gold standard coincided with a rise in international capital out ows to levels that were never approached before and have never been ap- proached since. Paul Bairoch’s estimates for capital ows for all net creditor countries show a slowdown in the depression decades of the gold standard, followed by an unprecedented surge after 1900: 1840–1870: 2.5–3.5 percent GNP 1870–1900: 1.5–2.0 percent GNP 1900–1913: 5.5 percent GNP 36 Comparable data for the 1920s, 1960s, and 1970s were below 1 percent. Most of this investment, about three-fourths, came from three countries (United Kingdom, France, and Germany), which were running persistent current account surpluses by generating savings in excess of domestic investment. Relative to total domestic savings, net capital out ows in 1910 represented 52 percent for the United Kingdom and 15 percent for France.37 The rest was contributed by the Netherlands, Belgium, Switzerland, and, toward the end of the period, Sweden. Most of thisinvest- ment went to a few countries—the United States, Canada, Australasia, Argentina, Brazil, Mexico, Russia, Spain, Portugal, Italy, Austria-Hungary, and the Scandina- vian countries.38 What made foreign investment so popular among savers in Britain, France, Ger- many, and other creditor countries was its greater safety, at equivalent yield, than domestic paper. In the case of Britain, Michael Edelstein found that overseas returns exceeded home returns over the years 1870–1913; he also found that overseas reincentive to borrow and then default on its debt through in ation or suspension of payments. Anticipating default, bond holders will either ask for a higher interest rate or not purchase government debt. A solution to the dilemma is for the government to commit to gold convertibility at a xed rate—a transparent and simple rule; Bordo and Kydland 1995. Bordo and Schwartz found that those countries that adhered to the gold standard rule generally had lower scal de cits, more stable money growth, and lower in ation rates than those that did not; Bordo and Schwartz 1994. 32. Bordo and Rockoff 1995, 18. 33. De Cecco 1974, 52. 34. Martin-Acen˜a 1994, 160. 35. Ibid., 144. 36. Bairoch 1976, 103. 37. Green and Urquhart 1976, 241, 244. 38. Cameron 1991, 13. Capital Market Internationalization 9
10 International Organization turns were not significantly riskier than domestic returns,but in fact tended to be less so.39 The greater safety of foreign investments relative to home investments is easily explained;it derived from the nature of these investments,which,according to Arthur Bloomfield,"depended directly or indirectly on government action."40 Loans either went to foreign governments(Russia and countries in central and southern Europe), or,even when loans went to private companies,as in the case of railroad construction and other public investments (utilities,roads,bridges,harbors,telegraph and tele- phone networks),they were made possible by government assistance in the form of guarantees,land loans,and cash grants.Finally,the bulk of this investment was portfolio;a generous estimate places the relative share of direct investment of the total long-term international debt in 1914 at only 35 percent.4 The higher yield of foreign over domestic investments holding risk constant,albeit empirically established,is more difficult to explain.Edelstein offered two interesting hypotheses.42 A first hypothesis,which the author thought to be valid in the case of the United States,views foreign returns constantly running ahead of expectations: "Overseas regions had a tendency to generate greater amounts of profitable innova- tions and new market opportunities,periodically fostering greater disequilibria,which in turn left their mark on realized returns."A second hypothesis looks for higher returns in market imperfection:"Overseas areas evinced a tendency to generate more circumstances involving imperfect competition and,possibly,greater monopoly rents."The active role played by host governments in attracting foreign capital pre- dictably was a consequential source of monopoly rents. A third hypothesis,I venture,was the relative backwardness of receiving coun- tries.With the exception of the Netherlands,creditor countries(Britain,France,Ger- many,Belgium,Switzerland)were generally more advanced industrially than debtor countries.The differential timing of industrialization triggered a product-cycle ef- fect:high-growth sectors in debtor countries were already stable-(or low-)growth sectors in creditor countries.Although yields on new ventures may have been the same,risks were lower in newly industrializing economies.In contrast,investing in an advanced economy meant putting one's money into new ventures with untested rates of return. The boom in long-term flows would supply banks with the short assets they were so desperately looking for in two ways.First,the joint-stock commercial banks on the continent,and later in Britain,took over the floating and placement of long-term bonds.Although these bonds were nominally long term,and banks standardly held onto such bonds no longer than it took to place them among their clienteles,the safety and trading volume of these instruments made them easily disposable assets, easily convertible into cash,and thus de facto substitutes for short-term paper. Second,commercial banks would float a government or government-guaranteed long-term bond issue provided that they be given a share in the more lucrative short- 39.Edelstein 1982.138. 40.A.Bloomfield 1968,4. 41.Dunning1992,116. 42.Edelstein 1982.140
turns were not signi cantly riskier than domestic returns, but in fact tended to be less so.39 The greater safety of foreign investments relative to home investments is easily explained; it derived from the nature of these investments, which, according to Arthur Bloom eld, ‘‘depended directly or indirectly on government action.’’ 40 Loans either went to foreign governments (Russia and countries in central and southern Europe), or, even when loans went to private companies, asin the case of railroad construction and other public investments (utilities, roads, bridges, harbors, telegraph and tele- phone networks), they were made possible by government assistance in the form of guarantees, land loans, and cash grants. Finally, the bulk of this investment was portfolio; a generous estimate places the relative share of direct investment of the total long-term international debt in 1914 at only 35 percent.41 The higher yield of foreign over domestic investments holding risk constant, albeit empirically established, is more difficult to explain. Edelstein offered two interesting hypotheses.42 A rst hypothesis, which the author thought to be valid in the case of the United States, views foreign returns constantly running ahead of expectations: ‘‘Overseas regions had a tendency to generate greater amounts of pro table innovations and new market opportunities, periodically fostering greater disequilibria, which in turn left their mark on realized returns.’’ A second hypothesis looks for higher returnsin market imperfection: ‘‘Overseas areas evinced a tendency to generate more circumstances involving imperfect competition and, possibly, greater monopoly rents.’’ The active role played by host governments in attracting foreign capital pre- dictably was a consequential source of monopoly rents. A third hypothesis, I venture, was the relative backwardness of receiving countries. With the exception of the Netherlands, creditor countries (Britain, France, Ger- many, Belgium, Switzerland) were generally more advanced industrially than debtor countries. The differential timing of industrialization triggered a product-cycle effect: high-growth sectors in debtor countries were already stable- (or low-) growth sectors in creditor countries. Although yields on new ventures may have been the same, risks were lower in newly industrializing economies. In contrast, investing in an advanced economy meant putting one’s money into new ventures with untested rates of return. The boom in long-term ows would supply banks with the short assets they were so desperately looking for in two ways. First, the joint-stock commercial banks on the continent, and later in Britain, took over the oating and placement of long-term bonds. Although these bonds were nominally long term, and banks standardly held onto such bonds no longer than it took to place them among their clienteles, the safety and trading volume of these instruments made them easily disposable assets, easily convertible into cash, and thus de facto substitutes for short-term paper. Second, commercial banks would oat a government or government-guaranteed long-term bond issue provided that they be given a share in the more lucrative short- 39. Edelstein 1982, 138. 40. A. Bloom eld 1968, 4. 41. Dunning 1992, 116. 42. Edelstein 1982, 140. 10 International Organization