On Trade, Sovereign Defaults and Economics: New Evidence from the 19th century (I)
This draft: September 2009 Very preliminary !! Please do not quote without permission of the author Juan H. Flores1 University of Geneva
Abstract This paper revisits the long studied question on why Governments repay their debts, and focus on trade reduction in post-default periods. I argue that a main link through which sovereign defaults affected trade in the late 19th century were financial intermediaries’ credit restrictions. By analysing the functioning and industrial organization of sovereign debt markets and the business for trade finance, we conclude the following. First, this paper demonstrates that the most important underwriters for sovereign debt were also very active in international trade finance. Second, looking at archival evidence from Baring Brothers, I show that the bank strongly reduced its exposure to Argentina after the 1890 crisis, held less Argentinean securities in its portfolio and reduced its trade credit—overall and in particular towards Argentina. And third, for the period of 18701913 this paper argues that those countries with stronger bilateral trade with Great Britain and who had one underwriter for their sovereign debt issues also defaulted less. It is thus not about sanctions, but about economics, pretty much as it is today.
JEL Classification Numbers: F34, N20, N23 Keywords: sovereign defaults, financial crises, trade finance, merchant banks
1
Email:
[email protected]. I thank Antonio Tena for sharing his database on bilateral trade. A complementary paper co-authored with him is in progress. I also thank Marc Flandreau and Olivier Accominotti for providing me further data; Moira Lovegrove and Clara Harrow, archivists in ING Baring Brothers, for their invaluable help. I also thank the project “The Economic History of the European Periphery” from the Ministry of Spain.
“But if your baker owed you £5 would you refuse to buy his bread (the best and cheapest bread in town), because otherwise you would be enabling him to repay you-not a smaller sum than he owes you, but—in a manner convenient to himself, as well to you?” The Economist, 20 November 1909.
Introduction Recent financial turbulences have renewed the interest of economists on sovereign debt defaults, as expectations of fiscal deterioration of Governments worldwide increase. Though many factors may influence a Government’s decision to default, the economic literature has advanced two main arguments why Governments would have an incentive to repay their debt. On the one hand, the Eaton and Gersovitz (1981) argue that reputation is a main incentive for Governments to honour their obligations. A default, by damaging a borrower’s reputation, would exclude a Government from international capital markets. On the other hand, the classical paper of Bulow and Rogoff (1989) demonstrates that reputation alone does not explain why countries repay their debts. A country may have the incentives to borrow in one market, take the proceedings, invest in other market, and default. This later argument has given way to a new literature trying to explain why sovereign debt markets exist at all, and to understand the economic rationality of Governments as repeated players in international capital markets.
Empirical analysis has followed. A main finding is that sovereign defaults periods emerge in clusters (Sturzenegger and Zettelmeyer, 2006; Suter, 1990,1992). Looking back to the early 19th century, these periods are: 1820s, the 1870s, the 1890s, the 1930s, the early 1980s and then again the 1990s. But besides this fact, long term patterns have been difficult to identify, and some empirical works have served to leave more open questions than those that they solved.2 This is why economists and economic historians have analysed specific periods, only to find that results differ in time and space, something that hints to the importance of particular financial and institutional structures. However it may be, the identification of the factors that explain the booms and busts periods of sovereign debt markets remains a main challenge in the economic literature. 2
See for instance, Tomz and Wright (2007) on the weak long term relationship between sovereign defaults and below trend GDP growth. Other long term patterns include also the geographic nature of defaulting: so, for instance, Latin America has been a persistent defaulter since the 1820s.
2
In this paper, we focus on the late 19th century and make a step forward towards the comprehension of the economics of sovereign defaults. This period has been a favourite field for economic historians as capitals flowed freely across borders and financial integration reached levels similar to today’s. The literature so far has mainly dealt with 1)“sanctions”, understood as gunboat diplomacy and trade embargoes (for instance Tomz, 2007; Mitchener and Wiedenmier 2004, 2005); 2) the cost for Governments of defaulting in the form of exclusion from financial markets or of higher borrowing costs (Eichengreen and Portes, 2000; Flandreau and Zumer, 2004; Tomz, 2007) and 3) on the role of the Corporation of Foreign Bondholders for the management of sovereign defaults (Estevez, 2007, Mauro and Yafeh, 2003). Though important, none of these studies has dealt with the microeconomics of defaulting and thus, each issue has been treated as isolated from the rest. So for instance, trade declines after defaults (when they exist) are interpreted as evidence of “supersanctions” (See Mitchener and Weidenmier, . Others have interpreted exclusion from capital markets as a difficult cooperation task that was only facilitated through the emergence of an institutional actor, the Corporation of Foreign Bondholders. To my perspective, however, these pieces of evidence can be better understood if we look at the functioning and the industrial organization of debt issuing and trade finance, and how they intervened in a Government’s decision to default.
This paper argues that trade declines, trade credit and exclusion from capital markets were important to the development of 19th century capital markets. In particular, I identify the link between trade and sovereign default within the financial intermediaries, understood as those agents who participated in the financing of international trade, and those who participated in the issuing and underwriting bonds of foreign Governments in the London capital market. We find that both markets were highly concentrated and that some of the dominating financial intermediaries in issuing sovereign bonds were also the dominating banks in the issuing of acceptances, the main vehicle for trade finance. This means that a sovereign default could affect the activities through a variety of channels. For instance, if the financial intermediaries held many securities of a defaulting country, the need for liquidity obliged the bank to reduce its credit, including advances for trade purposes. If the sovereign default was accompanied by an economic crisis in that country, the bank would have incentives to reduce further 3
its credit to merchants from the country due to their higher risk. Besides, the concentration of both markets implied that these financial intermediaries were also empowered to impose “sanctions” to defaulters, through direct trade credit penalties or through the exclusion from capital markets. As a result, we should see that those countries who traded more with the UK or which depended more on trade and trade finance had less incentives to default; this effect should be stronger if, besides this fact, the issuance of bonds of a particular country was done by one or few financial intermediaries, as this empowered them to impose penalties or obliged them to limit transactions with the defaulting countries.
We proceed as follows. Section I briefly reviews the literature of sovereign defaults focusing on the theoretical link with trade and the empirical evidence from the 19th century. Section II revisits the role of merchant banks in the business for trade finance and loan issues. Section III shows how both markets were correlated and analyses the industrial organization for both markets. Section III focus in one particular case, Baring Brothers, and how the bank reacted to the 1890 Argentinean default.
3
Section IV tests the expected correlation between probability of default, trade intensity and underwriters’ turnover, and shows that those countries who had higher bilateral trade with England and whose bonds were mainly underwritten by one bank were also those countries who defaulted less. Section V concludes. Section I. Literature review: The economics of sovereign defaults in the 19th century Sovereign defaults have made their come back since the 1990s and this has accompanied a resurge of the economic literature on sovereign debt. In a recent paper that makes a comprehensive survey on the economics and law on sovereign debt and defaults, Panizza, Sturzenegger and Zettelmeyer (2008) observe that a mainstream part of the literature is the role of the costs of default as a necessary condition for the sovereign debt market to exist. These costs include economic costs and sanctions. The former refers to the exclusion from capital markets and to future higher borrowing costs. The later refers to other kind of enforcement, such as legal sanctions, gunboat 3
This crisis caused a sudden-stop in capital exports which triggered in a number of defaults during the 1890s. The countries than defaulted in that decade were: Argentina and Uruguay (1890), Portugal (1892), Greece (1893), Guatemala and Nicaragua (1894), Costa Rica (1895), Brazil (1898), Dominican Republic (1899).
4
diplomacy and trade embargoes. Each of these issues has been analyzed in the economic history literature, and the more so for the 19th century, as financial integration increased to very high standards even comparable to today’s (Flandreau and Zumer, 2004) and London emerged as the financial centre of the world, closely followed by the end of the century by Paris and Berlin. Portfolio foreign investment was highly dominated by the issues of Government bonds (Obstfeld and Taylor, 2003), which meant that a sovereign default was an important event with which investors had to cope.
Flandreau and Flores (2009a) analyse the first boom and bust period of the 1820s and argue that neither trade nor gunboat diplomacy affected the decisions of Governments to repay their debts, as British authorities were reluctant to protect investors from speculative investments and because they considered that it would be harmful for the British economy to reduce trade with countries which were to become important markets for British products (Flandreau, Flores, 2009). The authors find that countries who issued their bonds through prestigious financial intermediaries did not default, while those issued by “wildcat banks”, financial intermediaries that occasionally participated in the issuing of foreign bonds, defaulted all. Analyzing European financial markets since the end of the 18th century, Tomz (2007) argues that reputation was a main incentive for countries to repay their debts, and argues that borrowing costs increased after a country defaulted. In another paper, Flandreau and Flores (2009b) argue that the market share of the main underwriters of sovereign debt during the boom of the 1860s up to the 1880s permitted them to impose certain conditionalities to countries that defaulted, organizing bailouts to permit them to reaccess capital markets. Financial intermediaries with high reputation had the ability to resolve information asymmetries –these agents had the incentives only to issue good loans or otherwise they would lose reputation and market share— guarantee that their issues would not inflict losses to investors.
During the period of the classical gold standard, Flandreau and Zumer (2004) argue that defaults increased borrowing costs in the form of higher spreads over U.K. consols in secondary markets: 500 basis points during debt renegotiation; 90 basis points after the first year and even 45 points after 10 years; in general, they find that agents rated lower countries who defaulted or whose debt burden diminished due to debt restructure. Estevez (2007) has analyzed the institutional framework focusing on the Corporation of 5
Foreign bondholders, an organization formed in 1868 to defend the interests of investors. The author argues that its existence diminished the time a country was excluded from capital markets, as the pressure from an official body obliged countries to negotiate and restructure its debt. Finally, other works have revisited the role of direct penalties in the form of supersanctions, whether trade embargoes or gunboat diplomacy. Mitchener and Wiedenmier (2004) argue that there is no evidence of trade decline after a sovereign default unless this was accompanied by supersanctions. These authors also argue that the Monroe doctrine, by credibly threatening intervention in the event of default, obliged Central American Governments to restructure their debts with British investors. Tomz (2007) has challenged these views arguing that interventions during the Gold standard period had other motivations than merely debt recollecting, such as civil wars or territorial conflicts.
Though much importance has been given to this debate, Panizza et al (2008) correctly assert that there is no conclusive evidence on the role of supersanctions. Besides, the evolution of capital markets lack of any presence of this kind of penalty, perhaps due to the evolution of legal modifications after World War II (Panizza et al., 2008: 47). This is why the economic literature has continued to focus on the economic costs of sovereign defaults. In particular, the empirical evidence hints to a negative relationship between default and trade. For instance, Rose (2002) argues that trade declined after defaults in the aftermath of sovereign Paris club rescheduling. Further evidence shows that trade between creditor and debtor countries decrease after a default, at least temporarily, as shown by other papers (Rose, 2002), but the main motivation for this results were direct sanctions following by the Paris club negotiations. Still, these papers do not deal with the precise mechanism on how a default affects trade, and, as Panizza et al. (2008:49) assert, “the channel through which defaults affect trade remains something of a puzzle”.
Further research has focused on trade credit restrictions. Cline (1987) argues that Bolivia and Peru suffered reductions in their flows of short-term credits after debt renegotiations. Rose and Spiegel (2004) argue that the patterns of trade and lending follow the same path, as borrowers with strong bilateral trade would have incentives to repay their debts or otherwise suffer from trade reductions. Their findings relate to
6
bilateral bank debt stocks and thus, consistent with the importance of trade and trade credit sanctions. As far as we know, none of these hypotheses has been tested for the 19th century. Not only does empirical evidence between trade and sovereign defaults remain inconclusive; the present state of the literature still leaves open the understanding of how this link worked. This paper argues that trade and trade credit penalties have to be regarded as individual decisions from those agents who could eventually be empowered to do so, i.e. financial intermediaries. In the next section we provide a deeper view on the functioning of trade finance and sovereign debt markets, and analyse under which conditions the 19th century financial architecture provided incentives to repay sovereign debts. Section II. The role of merchant banks in trade finance and in 19th century Sovereign debt markets We have previously mentioned that recent works on 19th century economic history have focused on the relationship between trade and sovereign defaults providing a macroeconomic link of both variables, namely, a gravity model including a set of instrumental variables on the right hand of the equation and a dummy variable for controlling for the event of default, following the set of studies advanced by Rose (2002), Rose and Spiegel (2004). The results, which can be regarded as weak and inconclusive at best 4, are interpreted as a result of the presence (or absence) of sanctions from the creditor countries. There is, nonetheless, an alternative explanation on whether a sovereign default affects trade and whether we are to talk about sanctions. I aim to analyse the behaviour of financial intermediaries, and how they could have influenced this link. Although the absence of this kind of analysis may appear as a striking empty in the literature, it has been a recurrent issue in the historiography of British banking. We now turn to describe some of these main findings and complement them with archival evidence and quantitative measures.5
4
Some of these works assume that Britain was the only creditor country, although France, since the 1880s and later Germany, since the 1890s became major capital exporters (see Estevez 2007 for new German figures on capital exports for the period 1870-1913). 5 To our knowledge, the only recent papers that deal with the financing of trade is Flandreau and Jobst (2005, 2008) in the context of global currencies and the emergence of leading international currencies.
7
Merchants, or merchant bankers as they came later to be known, were banks whose original function was as merchants of specific commodities (such as wool, cotton, sugar or dry goods). During the 18th and 19th centuries, as international trade increased along with the integration of capital markets, some of these institutions also assumed functions as “accepting houses” and “issue houses”, acquiring a prominent role for both commerce and the development of capital markets. According to Richard Roberts, the main activities of accepting houses referred to “the endorsement of bills of exchange, the means by which trade transactions were often paid, by highly regarded merchants on behalf of lesser known firms” (Roberts, 1993: 23). On the other hand, issuing houses activities included mainly the issuing of long term loans on behalf of foreign countries. By the second half of the 19th century, London had taken the lead as the main financial centre of the world, overshadowing Amsterdam and making the sterling pound the international currency per excellence (Flandreau and Jobst, 2005, 2007). Through the issue of bills of exchange, merchants all over the world could finance their transactions through a confident and efficient mean. Chapman (1983) studied the evolution of merchant banking and how their functions combined accepting activities, by which they “earned their bread and butter” (Chapman, 1983: 82), and the issuing activities. Though Amsterdam had already been a major financial centre to provide loans to foreign countries, London appeared in the international scene after the loan on behalf of the Prussian government in 1818, which was issued by Rothschild. After the independence of most Latin American states during the 1820s, they also became major borrowers opening the first major boom of capital exports, which ended in a major bust by 1826 (see Gille, 1965; Marichal, 1989; Dawson, 1990; Flandreau and Flores, 2009a). During the first half of the 19th century, many of these merchant banks operated internationally through agents. Hidy (1941, 1949) analyses the operation of British merchant banks in the United States, and argues that the main activities of these agents were to “sign contracts, borrow money, receive payments and consignments, and grant credits in the name of their principals” (Hidy, 1941:55). This author identifies by mid19th century that agents were beginning to be supplanted by the selection of special firms as representatives (Hidy, 1941: 56). In Argentina, Marichal (1984) argues that French banks were following this kind of procedures, so that Paribas, who actively 8
participated in the commerce between France and Argentina operated through the house of Bemberg in Buenos Aires. Baring, who became very active in South America, also operated through direct connections with specific Argentine merchants and bankers: Zimmerman, Frazier & Co (1832-1856), Edward O. Madero (1860-1886), and through S.B. Hale between 1879-1899 (Orwell, 2000: 94). These connections allowed merchant banks to be informed on the state of the world markets, and to directly participate in particular operations. Within these operations issuing activities became important for some of these houses. Although the specific relationship between issues and acceptances remains unclear, several arguments can be advanced. First, merchant banks could propose to issue loans on behalf of Governments of prosperous countries, as they offered a safer business with rapid gains.6 Second, trade expansion coincided with stronger economic activity and higher funds availability, and thus, the opportunities for new business in both activities were highly correlated. Third, specific links with railway contractors or other merchants (and even without them) often meant that the funds raised from the loans would translate into the exports of specific goods. For instance, Chapman (1983) quotes 1866 James Rothschild ascertaining that “seven-eights or fifteen-sixteenths of any loan issued are employed in buying goods” (Chapman, 1983:104). Marichal (1984) argues that much of the expansion of trade between France and Argentina was nourished by the loans raised in Paris by Argentina’s Government by the 1880s. Finally, issuing a loan also meant the expansion of commercial business, as official recognition of foreign governments in order to obtain commercial connections (Chapman, 1983: 104).
Finally, the participation in both activities also involved risks. Chapman (1983) evocates the paradox that though merchant banks were obliged to act upon high prudence when granting credit and advances for trade purposes, in many cases they became gamblers when acting as issuing houses. So, the 1820s saw the bankruptcy of implanted houses such as B.A. Goldschmidt or Barclay, Herring, Richardson. This is a main reason why many of these houses preferred to avoid participating in these activities. Much had to do with the activities involved in issuing a loan. Sometimes the 6
Chapman (1983) observes that issuing activities were generally regarded as risky and with low profitability. However, a precise analysis of the evolution of underwriting fees in the 19th century for “safe” Governments indicate that fees were a permanent, albeit low source of profits. Fees for risky countries were high, up to 15% in some cases. To avoid the high risk that this involved, syndicalisation became recurrent use by the end of the century.
9
banks assumed the risk of the issuance by taking the bonds “firm”, and then try to place them in the market a bet that often involved heavy losses if investors refused to buy.7 In other cases, they had to sustain the prices by intervening in the market, so as to keep the illusion of a successful placement. Finally, an extreme case was of course, when a country defaulted. Flandreau and Flores (2009) and Flandreau et al. (2009) show that some merchant banks, particularly those with the major market share and reputation cared about the countries whose bonds they issued and to avoid by all means losses to investors, as this would entail the loss of market share. This could be costly in the short term: Baring, for instance, as it undertook Mexican agency after Barclay’s bankruptcy in 1828 did finance coupon payments during six months hoping the Mexican Government to arrange its finance. This was not the case, Mexico defaulted and Baring abandoned the agency for more than four decades.
The bills of exchange were also issued by merchant banks—who operated through agents or merchant houses in the countries with which they dealt. Some of these merchant banks became highly specialized and competitive in the commodities and geographical regions where they operated, and thus, the acceptance business was a relatively highly concentrated market. Unfortunately, we lack precise and continuous series of the total acceptances issued by British merchant banks. One of the few datasets available is Chapman’s (1983) database on total acceptances issued for the period 19001913 by British merchant banks or accepting houses. That database also includes figure for individual cases, those which Chapman considers the most important and representative merchant banks and accepting houses. These were: Barings, Rothschilds, Kleinworts, Schröders, Hambros, Gibbs, Brandts. They represented about one third of the total market.
I aim to focus on the link between trade and default through a main agent in the sovereign debt market and which has been overlooked in the economic history literature. They were mainly merchant banks, i.e., banks whose original function was the trade of specific commodities such as cotton, wool or coffee. These “merchants” moved at the same time to trade finance as their activities expanded in volume and in space. Historians have described how their functioning evolved during the 18th and 19th 7
On the description of the different issue systems in the 19th century see Flores (2007) and Flandreau et al. (2009).
10
century, and how these activities became complementary with the activity of underwriting sovereign bonds for foreign Governments. These agents, whose functions moved from pure merchant activities, to merchant finance and finally to the underwriting of securities and public bonds in London, were also present in other markets such as Paris or Amsterdam. Both underwriting and trade finance became very soon highly concentrated markets with some few merchant banks dominating the market for sovereign debt and looking for new international trade opportunities (Chapman, 1983). The participation in both markets of the same agents –big houses and big names who dominated in both markets--provide a link which has not been explored so far in the recent literature of trade and sovereign trade. Nevertheless, from my perspective, this is precisely the link that explains why trade fell in response of sovereign default and excludes the possibility of a direct penalty in the form of a “sanction”. In any case, if we were to write about sanctions, we should analyse whether merchant banks imposed sanctions on defaulting governments as they were the agents who had the power to do so; and in fact, as frequently the merchant banks possessed quantities of securities from Governments or countries whose bonds they underwrite, deterring them from defaulting yielded in a particular gain for them. This economic rationality, which to our perspective seems essential to the understanding of 19th century sovereign defaults, has been completely ignored so far.
As these activities became soon highly concentrated markets, some banks underwrote the major part of the bonds issued in the main financial centres of Europe. This fact provides the connection between trade and capital exports booms. The market of sovereign debt and of trade finance were both highly concentrated market, with some main participans, for instance, the Rothschilds, the Barings or the Hambros being very active in both markets. These intermediaries could exclude countries to participate in the London capital markets, and could even “restrict” the credit necessary for trade.
Section III. Lending Booms, trade expansion and the Industrial Organization of Sovereign debt markets and trade finance
In an article published in June 1876, at the peak of a World economic crisis that followed the major trade boom of the 19th century, the Banker’s Magazine argued that defaults from a number of foreign governments would strongly affect the international 11
trade (in particular the exports) of Great Britain—the more so as arrangements to resume payments seemed very difficult—though it recognized that the effects would be temporal and that it would be a matter of time to have a new take off. The logic of a typical sudden-stop situation was well described by the fact that many countries who recurred to the London market did so to procure goods which would otherwise be not affordable to them, at least at their present state of development. But some loans were also used to pay the interest of previous loans, and thus, many of these countries, who fostered the exports from the U.K., were obliged to default.
Economists in Great Britain were well aware of the risks of foreign governments’ defaults. The U.K. being an export economy and the main creditor, defaults not only inflicted losses to its investors; it also shrunk its trade with defaulting countries as demand fell. Besides, as we mentioned, the British financial system was international in nature, and underwriting banks also lost from defaults, as it was the case in the 1820s but mainly in 1890, as Baring Brothers required a bailout from the Bank of England to continue operations and avoid a banking panic in the city.
To deter such events to happen was not an easy task. Since the 1820s, Colombia was the first country, before the loan boom had actually taken place, to default on its 10% debentures issued to pay for new armaments. The affected merchants decided to stop exporting new armament until the Colombian Government resume payments. In fact, the first Colombian loan of 1822, the first of the loans that were to come, had as a first objective to repay former debts. Besides, once the storm of foreign government loans passed by 1825, and defaults spread one after the other, negotiations almost immediately followed for two countries: Argentina (Buenos Aires) and Mexico. Others had to wait until the new countries stabilized and even separate, as in the case of Central America of New Granada; or had to wait for the will of every new government, or the end of civil war, until they succeeded to reach an agreement.
Committees were formed in order to defend the interests of the bondholders. A constant research was initially the support from the Government in the form of military pressure, although these demands were constantly rejected (On the beginnings of these initiatives, see Flandreau and Flores, 2009b). In one of the first speeches, Isidor Gerstenberg, considered the founder of the British Corporation for Foreign 12
Bondholders, also recalled the idea that British interests ought to be defended by their Government. The CFB should consider military intervention as a legitimate way to defend British interests; however, they understood that this would be a real mean only in extreme cases. In fact, in the first annual reports on the activities of the CFB, it was explicitly understood that the realistic means to enforce payments were actually the exclusion from the quotation on the Stock Exchange, the cooperation with investors from other countries, and to press upon the Committee for General Purposes to prevent any new issues from defaulting states (something that was to become a law in the later functioning of financial markets both in Europe and in the US). During the 19th century, financial integration reached standards which have not been reached even today; capital could flow among international borders, and investors were free to decide on the best destiny for their money, seeking the highest rentability. These investments were not without risk. Bankruptcies, wars, and sovereign defaults inflicted occasionally heavy losses to adventurous investors; and thus, the effects on investment, trade, and economic activity had to be evaluated. In fact, it was no mystery to 19th century economists that booms in capital exports were accompanied by booms in international trade. Much of the capital issued in London market was later used to finance railway equipment for the development of contemporary emerging countries. More generally, many of the exports from Great Britain were financed precisely by the means of the financial market of London. Business cycles were thus self-inforcing: a stop in capital exports also put a halt in the exports from Great Britain, which had negative consequences in the country thereby exporting the negative cycle to the world. In fact, British and international historiographies have a long tradition analysing the relationship between Great Britain’s business cycles with the rest of the World: Great was the main trade partner for the majority of European countries and to many other countries in all continents. Besides, it was the major capital export country, only challenged by the end of the century by France and Germany.8 The economic cycles in Great Britain were exported to other countries. Ford (1971) has analysed these cycles for the period 1870-1914 and linked them with the operation of the gold standard. This author has also analysed how Argentina also suffered from the sudden stop of British
8
For a recent review on business cycles for the period 1870-1913 see Flandreau et al. (2009).
13
capitals and how it negatively affected the economic conditions in Argentina that gave way to the 1890 crisis.9
The fact that capital flows and exports were accompanied is nothing new. In Figure 1 I show the evolution of total trade between Britain and Argentina (from Ferns, 1963) and the evolution of capital flows from Britain to that country (from Stone, 1999). The evolution of both variables follows the same path, with a strong boom during the 1880s and then a bust in 1890 with the crisis. It is clear that we are not looking at a case of supersanctions: British marines did not invade Buenos Aires nor was there any embargo at the ports. However, British Banks (the Rothschild committee) and Argentina’s Government did negotiate a bailout by the end of 1890 to end the default and restructure Argentina’s debt. It was in the interest of the London capital market, of Baring, and of Argentina to deal with the problem.
To understand thus the link between trade and capital exports we have to look at the industrial organization of both markets. Figure 2 shows the evolution of the Herfindahl index of the market for sovereign bonds underwriting for the 19th century. A main feature is that, overall, the market was highly concentrated, with two main leaders, the Rothschilds and the Barings, as already noted by Chapman (1983). However, this is not stable, and a second feature is the occasional peaks of concentration that follow capital booms. In fact, after the 1820s, the 1860s, and the 1880s, where the markets become highly competitive, there is a return to concentration as during bust times, only some countries (and some underwriters) are able to issue new bonds. This is explained by the fact that boom periods are followed by a number of defaults (represented as the dots for every year) which trigger the bust and sudden stop in capital exports. Default times provoke thus the closing in the capital markets, and the Herfindahl index increase as a result.
Figure 3 shows how these two markets were linked. The same agents who were leading the markets for acceptances issues were the same leading agents who dominated the markets for sovereign issues. The Figure takes into account the market share of the six banks (Barings, Rothschilds, Kleinworts, Schröders, Hambros, Gibbs, Brandts) 9
On sudden stops in the 19th century see Catao (2006). Prebisch (1919) makes an exhaustive analysis on trade specialization and the effect of cycles in the “core” countries on the “periphery”.
14
included in Chapman’s dataset. Of course, not all banks issued new bonds every year, meaning that Figure 3 only include observations when banks were active in the underwriting market. Though the acceptance business was less concentrated; it remained in the hand of some merchant houses. I now proceed to explore the link between these two markets for a specific case: Baring and Argentina in the 1880s.
Section III. The Baring crisis of 1890: How the bank coped with the default In previous works of my own (Flores 2004, 2007a, 2007b) I already explained the microeconomic aspects of the Baring crisis of 1890. This merchant bank was a highly reputated investment bank in London and the main underwriter for Argentina since the independence of the country. The 1880s was a decade of increased competition between underwriters to enter Argentina's Market, as that country offered a new perspective as a prosperous and promising emerging market, and London investors strongly demanded high yield from exotic countries as home’s interest rates remained comparatively extremely low. Due to overexposition of Argentina's bonds, Baring needed a financial bailout organized by the Bank of England, with the participation of some of the main financial institutions of London, as well as the Bank of France. Argentina, on the other hand, suffered a triple crisis (currency, banking and debt) after an expansive period of over 15 years. It was so important due to a probable case of contagion.
How did Baring reacted to the possibility of default by Argentina? Flandreau and Flores (2009) showed that Baring actually specialized in the “speculative grade” issues. Though initially reluctant to participate in the underwriting of foreign government debt, Baring accepted to act as a distributor of Buenos Aires’ bonds in London. After the default in 1827, it constantly looked for an agreement and defended the interests of investors until the final agreement in 1857. Baring actually did actively participated in other countries’ negotiations with the bondholders as a mediator, but rarely issued any bonds on behalf of these countries. Baring pressured through the participation in investors’ committees in those countries, with correspondence and occasionally through active participation of one of its agents in the deals with foreign Governments. With Argentina, Baring had a permanent agent with which it communicated to get constant information of economic, political and social matters. Those agents also had a particular relationship with Governments providing even short15
term advances when necessary. At the end, thus, Baring did not use to recur to menaces in its availability of credit or cutting its trade finance, or to pressure the British Government to intervene (rather the opposite).
As Argentina’s situation deteriorated during the late 1880s, Baring tried to organize a bailout with the participation of other merchant houses in London (German and French houses retired from the agreement before). The issue could not take place as Baring’s declared bankruptcy and the instability of Argentina’s Government made impossible to issue the new bonds. However, what we do see is the evolution of the acceptances that Baring received from S.B. Hale, the merchant house through which Baring operated in Argentina. Table 2 resumes the proportion of total acceptances in Baring’s liabilities’ side of its balance sheet at end of year, and those received during that year on behalf of S.B. Hale. The number of acceptances coincides with the exports and capital flows booms of Britain, but this increase is still more important for Argentina, as shown in the last column by the ratio of Hale’s acceptances to the total. However, as it can also be deducted, the crisis also involved a decrease in this ratio: the effects on economic activity, and on the position of S.B. Hale cased the acceptances business to considerably shrink after 1890. In fact, S.B. Hale made bankruptcy and its name disappears fom Baring’s balance sheets since 1893.
Regarding the underwriting business, the crisis had an important effect on both, Baring’s activities (the bank was split in two) and Argentina’s access to the London market. Baring became the sole underwriter of Argentina’s bonds after two successive bailout loans (1891 and 1893, and later also the deals regarding the provinces debts) ten years after 1890. Most important, Baring reduced its holdings of Argentina’s bonds as shown in Table 3. By the late 1880s, the proportion of Argentina’s stock to total in Baring’s portfolio was roughly ten percent. Though Baring’s accountability is not clear after that year, the next year for which we have comparable figures is 1892, where this figures descend to 2%. After that year, however, the figures again increase to return to the previous ten percent. It can thus be noticed that Baring’s need for liquidity involved sales of Argentina’s stock, but not any kind of punishment to the defaulting country.
16
Section IV. Probability of turnover, bilateral trade and defaults: Some empirical evidence As the last part of our argument we can have a broad picture on the implications from the analysis advanced so far. We should bear in mind that not all countries depended to the same degree from foreign finance. Other countries, particularly in Europe, developed their own financial institutions and relied less on the British financial market. Those countries only occasionally tapped foreign resources, or relied on their domestic markets. We will thus focus on those countries who constantly recurred to London as a main source for finance, and test whether those countries who had stronger ties to the UK also defaulted less, and how the industrial organization interfered in the results.
We begin by capturing the effect of the concentration of loan issuers for specific countries. Less underwriters issuing loans for a specific countries mean that they would be more affected from a default, as they would more likely be obliged to reduce their holdings of the securities of the defaulting countries—depreciating prices and making more difficult to remarket them—and to reduce trade credits, as a form of direct penalty or because agents from that country become more risky, particularly if a default is accompanied by an economic crisis. This argument is reinforced by other already advanced in the literature. For instance, in previous works of my own (Flores, 2007), I have already mentioned how competition between underwriters can lead to overborrowing, as demonstrated by the 1880s road to the Baring crisis. An additional element is the fact that monopoly underwriters have more incentives to closely monitor the financial situation of foreign Governments (Flandreau, 2003). Finally, countries with fewer underwriters can also be more easily “punished” in the event of default, as they can control market access (Flandreau and Flores, 2009).
Figure 4 shows this point by showing the market sentiment towards countries that more frequently changed of underwriters. Investors demanded a higher risk premia (measured as the spread between a particular country’s yield and the yield of the U.K. consol). The probability of turnover is calculated as the proportion of total issues where countries changed of underwriter. Spreads over U.K. consols are spreads at issue. As we see, countries with lowest spreads, i.e. whose bonds were considered as a safer investment, were also countries whose loans were issued by few underwriters. If we 17
classify countries in two groups, defaulting and not defaulting. It demonstrates that the average probability of turnover of borrowing countries is higher for defaulting countries, and that the difference is statistically significant running a Z-test of mean differences.
We proceed next to test how trade with the U.K. intervened in the decision of defaulting. Figure 5 shows the relationship between proportion of trade with the UK and number of years in default for the period 1880-1913. Countries whose bilateral trade with the UK is more important relative to total trade also seem to default less. Notable exceptions are Portugal, not included in the sample as for the period studied it did not issue a loan in the London market. This country had more than half of its total international trade done with the U.K., but due to political instability could not reach an agreement with foreign investors until 1902.
I have thus proceeded to control for the determinants in the number of years in default taking into account other factors such as monetary and fiscal variables, as well as the typical openness measures. The data used come from a number of different sources. Total trade estimates are from Mitchell (2003). Bilateral trade between capital recipients and the UK are from Tena’s bilalteral trade database, whose original sources are the official accounts from different countries and the Statistics from Great Britain. Monetary and public finance variables are mainly from Flandreau and Zumer (2004) for Europe, Brazil and Argentina. Other Latin American countries’ data comes from Oxlad database, with the exception of Chile, where we used Braun et al. (2000) data.
We ran OLS regressions for the following equation: y = α + βx + χTradeUK + δ Pr obTO + ε
Where y is the dependent variable and measures the number of years that a country is in default. The vector x is the set of control macroeconomic variables. Trade
UK is the variable that measures the proportion of total trade that a country has with the UK. ProbTO is the probability of turnover measured for each country as shown above, and ε is a standard error term. Our null hypothesis is concentrated on the sign and
18
significance of both, the χ and the δ parameters. We expect a negative sign for the former: a country that has more to loose from trade contraction would have less incentives to default, and it defaults, to rapidly achieve an agreement to resume payments; for the latter we expect a positive sign: those countries which constantly change of underwriter have less to loose from a default and are less likely to suffer from trade credit reductions.
Results are shown in Table 4. The first two regressions take into account only the pertinent variables for our analysis, and both are significant and with the expected sign. However, if we take into account the control variables (Regressions 3-5), the significance of the variable Trade UK disappears. From the control variables, the deficit as a percentage of total fiscal revenues is the only significant, and to a less extent trade openness, measured as the ration exports to GDP. The main result is provided by the significant persistence of the probability of turnover, which remain significant even including control variables. Though I am aware that this result does not necessarily mean that this result is directly link to the disincentive of defaulting do to the shrinkage of trade finance, it does give a hint on the importance of the role of financial intermediaries when Governments default.
Section V. Conclusions In this paper, we have advanced two main arguments. First, there is a strong link between the existence of a market for foreign debt and capital flows and trade, identified in this paper within the financial intermediaries. Second, trade and trade finance were important variables that deterred Governments from defaulting. Contrary to recent previous works on the history of sovereign debt markets, this paper shows the importance of an analysis of the microstructure of capital markets. Whether trade was affected by a sovereign default was thus less caused by supersanctions but rather, whether trade with the creditor country was important and whether this trade was concentrated by few financial intermediaries who could react to the default through a rationing in trade credit.
I have presented the case study of Baring and Argentina’s default. I am aware that this was an exceptional case in the sense that Argentina’s problems caused a liquidity crisis to the bank, a fact that was rarely present in other cases of sovereign 19
defaults during the 19th century (with the exception of the 1820s). However, it does show the trend and final consequences of a proper “misbehaviour” during a capital exports’ boom. The Baring crisis is just an extreme case of a sovereign default shrinking economic activity. Due to the fall in trade credit, merchant houses in Argentina also made bankruptcy, but trade recovered afterwards with the intervention of other merchant banks in London. At the end, Baring lost market share in trade finance, but it recovered its monopoly power in underwriting Argentina’s sovereign bonds.
This
further bifurcation in trade finance, where the expansion of trade was accompanied by the emergence of more agents and underwriting, where relationships between Governments and banks or ban’ syndicates stabilized should be further researched, along with the inclusion of further variables to understand the last consequences of this trend as for instance, acceptances commissions and the profits of merchant banks derived from each activity.
References Periodicals, Archives, Reports The Banker’s Magazine, 1876. The Economist, various issues. ING Baring Archives. Ledgers and Balance Books, 1880-1894. Corporation of Foreign Bondholders, Annual reports, 1915.
General references Borchard, E., 1951, State insolvency and foreign bondholders, Vol. 1: General Principles, New Haven: Yale University Press.
Braun, J., M. Braun, I. Briones, y J. Díaz (2000): "Economía Chilena 1810-1995: Estadísticas históricas". Documento de Trabajo Nº 187, Pontificia Universidad Católica de Chile, Santiago. Bulow, J. and K. Rogoff, 1989. "A Constant Recontracting Model of Sovereign Debt," Journal of Political Economy, 97(1), pp. 155-78.
Catão, Luis,Sudden Stops and Currency Drops: A Historical Look(May 2006). IMF Working Paper, Vol. , pp. 1-61, 2006. Chapman, S., 1984, The Rise of Merchant Banking, London: George Allen and Unwin.
20
Dawson, F. G., 1990, The First Latin American Debt Crisis. The City of London and the 182225 Loan Bubble, Princeton: Princeton University Press.
Eaton, Jonathan, and Raquel Fernandez. 1995. “Sovereign Debt.” In Handbook of International Economics, Vol. 3, ed. Gene M. Grossman and Kenneth Rogoff Amsterdam: North Holland.
Eichengreen and Portes, 2000. Debt restructuring with and without the IMF, unpublished manuscript, London Business School. Eichengreen B. and R. Portes, 1989, “Settling defaults in the era of bond finance”, World Bank Economic Review, 3 (2): pp. 211-39. Esteves, R. P., 2007, “Quis custodiet quem? Sovereign debt and bondholders’ protection before 1914”, Working Paper, N° 323, Oxford University.
Feis, H. (1964). Europe, the world's banker, 1870-1914, an account of European foreign investment and the connection of world finance with diplomacy before the war. New York : Kelley for the Council on Foreign Relations. Ferns, H. S., 1952, “Beginnings of British Investment in Argentina”, Economic History Review, New Series Vol 4, N° 3., pp. 341-52 Ferns, H.S., 1960, Britain and Argentina in the nineteenth century. Oxford: Claredon press.
Flandreau, Marc, & Jobst, Clemens (2005). The Ties that Divide : a Network Analysis of the International Monetary System, 1890-1910. The Journal of Economic History, vol. 65(n°4). Flandreau, M., 2003, “Crises and Punishment. Moral Hazard and the Pre-1914 international financial architecture”, in M. Flandreau (ed.) Money doctors. The experience of international financial advice 1850-2000, London Routledge.
Flandreau, M., Flores, J (2009a) Bonds and Brands, Journal of Economic History, Forthcoming. Flandreau, M, J. Flores (2009b), The Industrial Organization of Prestige: Conditionality Lending in Theory and History, unpublished manuscript (2009). Marc Flandreau and Clemens Jobst (2008) “The Empirics of International Currencies: Network Externalities, History and Persistence” (forthcoming, Economic Journal) Esteves, R. P. “Quis custodiet quem? Sovereign Debt and Bondholders’ Protection Before 1914.” Working Paper, No. 323, Oxford University, 2007. Flandreau, M, J. H. Flores, N. Gaillard and S. Nieto-Parra (2009b), “Two Centuries of Government Bond Underwriting”, Paper prepared for the World Economic History Congress, Utrecht. Flandreau, M., Frederic Zumer (2004) The Making of Global Finance. OECD.. Flores, J., 2007, “Lending Booms, Underwriting and Competition: The Baring crisis revisited”. Working Papers in Economic History, Universidad Carlos III, 07-01.
21
Flores, J., 2007, “Information asymmetries and financial intermediation during the Baring crisis : 1880-1890”. Working Papers in Economic History, Universidad Carlos III, 07-16
Flores, J. (2004) Lorsque le leader suit la foule, une analyse micro´conomique de la crise Baring, unpublished PhD dissertation. Sciences-Po, Paris. Ford, A. G. (1971). British investment in Argentina and Long Swings. Journal of Economic History, 31, 650-63. Gille, B., 1965, Histoire de la Maison Rothschild, Vol I: Des origines à 1848. Geneva: Droz. Hidy, R. W., 1941, “The organization and functions of Anglo-American Merchant Bankers, 1815-1860” Journal of Economic History, Vol 1, pp. 53-66. Hidy, R. w., 1949, The House of Baring in American Trade and Finance: English Merchant Bankers at Work, 1763-1861, Cambridge University Press Mitchell, B.R. (2003). International historical Statistics : the Americas, 1750-2000. New York:
Palgrave Macmillan. Mitchener, Kris and Marc D. Weidenmier, 2005, “Empire, Public Goods, and the Roosevelt. Corollary.” Journal of Economic History, 65: 658-692. Mitchener, K. J. and M. Weidenmier, 2006, “Supersanctions and Sovereign Debt Repayment”, NBER Working Paper 11472. Marichal, C. (1995). Las inversiones extranjeras en América Latina, 18501930. Nuevos debates y problemas en historia económica comparada . México : Fondo de Cultura Económica. Marichal, C., 1989, A Century of Debt Crises in Latin America: from Independence to Great Depression, 1820-1930, Princeton: Princeton UP.
Mauro and Yafeh (2003). The Corporation of Foreign Bondholders, IMF Working Paper, WP/03/107. Orwell, J. (2000). The historical archives of ING Barings, Financial History review,7, pp.89-104. Panizza, U; F. Sturzenegger; J. Zettelmeyer (2008). The economics and Law of Sovereign Debt and Default, Forthcoming Journal of Economic Perspectives Platt, D. C. M., 1968, Finance, Trade and Politics in British Foreign Policy 1815-1914, Clarendon.
Regalsky, A. (2001). Mercados, Inversiones y Elites. Las inversiones francesas en la Argentina, 1880-1914. Buenos Aires : EDUNTREF, 2001, 480 p. Roberts, R. (2003) What's in a name: merchants, merchant bankers, accepting houses, issuing houses, industrial bankers and investment bankers. Business History, v.35, 3, p. 22-38.
22
Rose, Andrew K. 2005. "One Reason Countries Pay their Debts: Renegotiation and International Trade.” Journal of Development Economics, 77(1): 189–206. Rose, Andrew K., and Mark M. Spiegel. 2004. “A Gravity Model of Sovereign Lending: Trade, Default, and Credit.” IMF Staff Papers, 51 (Special Issue): 50–63. Summerhill, W., 2006, “Sovereign commitment and financial underdevelopment in Imperial Brazil”, Working Paper, UCLA.
Tomz, M. (2007), Reputation and International Cooperation: Sovereign Debt across Three Centuries. Princeton, NJ: Princeton University Press, 2007. Tomz, M. & Mark L. J. Wright, 2007. "Do countries default in “bad times”?," Working Paper Series 2007-17, Federal Reserve Bank of San Francisco Wright, M. 2002. “Reputations and Sovereign Debt.” Unpublished. Wynne, W. H., (1951), State insolvency and foreign bondholders, Vol.2: Selected case histories of governmental bond defaults and debt readjustments, New Haven: Yale University Press. Ziegler, Ph., 1988, The Sixth Great Power. Barings’, 1762-1929, London: Collins.
23
TABLES AND FIGURES 3.5
120
3
100
2.5 80
2 60 1.5
40 1
20
0.5
0
0 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 Total_trade_GB (in gold pesos)
Gross Capital Flows
Figure 1. Capital flows and trade between Britain and Argentina. Source: See text. 12000
10000
8000
6000
4000
2000
18 18 18 21 18 24 18 27 18 30 18 33 18 36 18 39 18 42 18 45 18 48 18 51 18 54 18 57 18 60 18 63 18 66 18 69 18 72 18 75 18 78 18 81 18 84 18 87 18 90 18 93 18 96 18 99 19 02 19 05 19 08 19 11
0
Total
Number of defaults
Figure 2. Herfindahl Index in Sovereign debt market issues, London. Source: Author’s calculations.
24
1.2
1
Issue market share
0.8
0.6
0.4
0.2
0 0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
Acceptances end of year
Figure 3. Market shares of top six merchant banks in Acceptances and Underwriting. Source: See text. 6
Countries included in the sample: 5
Spreads over UK consols
STO.D M
y = 1.3812e0.0106 x R2 = 0.3363
4 URU
MEX
3
GRE BRA
2
JAP
CUB
ITA ARG TUR
HUN
RUS
EGY
1
SWE
DEN
0 0
20
40
60
80
100
120
Probabiilty of Turnover
Figure 4. Probability of Turnover and Spreads over U.K. consols. Source: Author’s calculations.
25
45%
DEN
40%
35% SWE Trade with U.K./Total Trade
NOR 30%
GREE
25%
COR ARG
COL
BRA
20%
UR RUS
ECU NIC
VEN
15%
GUA 10% MEX 5%
0% 0
5
10
15
20
25
30
Years in default
Figure 5. Years in default and trade with U.K. (proportion to total trade for each country). Source: See text.
Country Argentina (Buenos Aires) Mexico Venezuela Colombia
Year of Default
Originally issued by Barings
First Agreement
Final Settlement
New Bonds’ issue by Barings after agreement
1827
Yes*
1857
1857
Yes, 1866
1828 1826 1826
No No No
1831 1862 1846
1888 1881 1862
No No No
Table 1. Barings’ involvement in negotiations between Foreign Governments and Bondholders. Sources: See text.
26
Year
Acceptances in Liabilities at end of year
To S.B. Hale
1884 1885 1886 1887 1888 1889 1890 1891
7’407 5’514 6’960 6’334 11’603 11’060 8’932 3’455
50 155,4 318,1 208,5 NA 1’465 188,3 14,2
Percentage of Hale acceptances to Total (%) 0,6 2,8 4,5 3,3 13,2 2,1 0,4
Table 2. Barings’acceptances. Source: ING Baring Archives.
Total Bills receivable Cash Stock Argentina’s Stock Argentina’s Stock / Total Stock (%)
1889 11’988 5’229 1’254 6’634 603
1892 4’172 1’232 818 1’172 13,8
1893 3’898 1’211 88 689 69
9,1
1,2
10,1
Table 3. Baring’s portfolio, Argentina’s stock. Source: ING Baring archives.
27
Variable Constant
Reg. 1 2.3*** (2.4)
Reg. 2 0.10 (0.22)
Deficit/Revenue
Reg. 3 -0.255 (-0.20) 0.011 (0.517)
Reg. 4 1.98 (1.49) -0.003 (-0.138)
0.041 (0.68)
Reg. 5 1.6 (0.9) -0.05*** (-4.1) -0.02 (-0.7) 0.24 (1.6) -0.01 (-1.06) 0.02 (1.0) -0.13** (-1.9)
1.04** (1.51)
-0.2 (-0.8)
-0.13 (0.11) -0.08 (-1.0) -0.09 (-0.31)
0.20 24
0.6 18
0.65 18
Debt Service/Revenue Exchange rate volatility Reserves/M1 Trade GB/Total Trade Exports/GDP
-0.57** (-1.96) 0.005 (0.96) 0.023*** 0.028*** (2.4) (2.4) 0.42 (0.60)
Probability of Turnover Dummy Period 1880-1895 R2 Number of Observations
0.12 32
30
0.22 24
Reg. 6 2.6 (1.4) -0.57** (-2.1) -0.027 (-0.45) 0.15 (0.6) -0.11 (0.3)
Table 4. OLS results. Source: Author’s calculations. (Note: * significant at 10%; ** significant at 5%; *** significant at 1%)
28