Jump to ContentJump to Main Navigation
Can Latin American Firms Compete?$

Robert Grosse and Luiz F. Mesquita

Print publication date: 2007

Print ISBN-13: 9780199233755

Published to Oxford Scholarship Online: January 2008

DOI: 10.1093/acprof:oso/9780199233755.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2019. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use.  Subscriber: null; date: 17 September 2019

National Financial Systems in Latin America: Attributes, Credit Allocation Practices, and Their Impact on Enterprise Development

National Financial Systems in Latin America: Attributes, Credit Allocation Practices, and Their Impact on Enterprise Development

(p.309) 14 National Financial Systems in Latin America: Attributes, Credit Allocation Practices, and Their Impact on Enterprise Development
Can Latin American Firms Compete?

John C. Edmunds

Oxford University Press

Abstract and Keywords

Most firms headquartered in Latin America are not large enough to rank highly among the top firms in the world, and do not operate in industry sectors that are classified as rapidly growing. Most explanations for this do not assign a causative role. It is argued that those financial systems have played a role in determining the kinds of firms that obtained financing, and thus share the responsibility for the small number of world class, high growth, high valued-added firms in the region. This chapter argues that the national financial systems suffered from attributes that diminished their capability. Six attributes are identified, which would hinder the growth of these sorts of firms. Some of the attributes are well known (i.e., that Latin American national financial systems are small relative to the economies in which they operate, and suffer frequent collapses), while other attributes are less well known (i.e., that those financial systems tended to channel loans to plantations, mines, and firms that produce commodities). Evidence is presented indicating that four of the attributes do describe Latin American financial systems accurately. There are also arguments why the attributes would lead to the mix of firms observed today. The evidence is not complete enough to prove decisively that Latin America's national financial systems were the main determinants of the size and sophistication of the firms that have risen in the region.

Keywords:   Latin America, financial systems, firm growth, causative role

14.1. Introduction

In Latin America very few large, globally integrated multinational firms have arisen. In its 2005 survey, the magazine América Economía lists the twenty‐five largest firms in Latin America. The list has behemoths on it but very few of the firms are integrated or global. Eleven of the region's twenty‐five largest firms (ranked by turnover) are in the primary producing sectors of oil and gas, mining, and steel. Three telecommunications companies and two electric utilities also appear among the top twenty‐five. Most of those companies operate in a single country. They are vertically integrated, but nowhere near as globally integrated as many other big companies headquartered in the industrial countries. Among the top twenty‐five companies there are only two, TELMEX and CEMEX, that have extensive operations in many countries.

This chapter puts forward an explanation for the pattern of enterprise development: national financial systems in Latin America have held back the development of firms in the region. Access to capital was a constraint that discouraged them from some of the paths of expansion that firms in the rich countries and firms in the Asian Tiger countries followed with extraordinary success. The largest Latin American firms grew large in their own home countries, but the financing they were able to obtain restricted all but a very few (p.310) of them from becoming large, globally integrated multinational firms. This explanation focusses on the attributes of Latin American financial systems, and assembles piece by piece a compelling, but not exclusive, explanation for the pattern of business enterprise development. There are many other reasons why so few of the world's large multinational companies have arisen in Latin America. Many of those other explanations are also compelling, and hopefully this explanation focussing on the national financial systems will join with the other compelling explanations to constitute a more complete explanation of the region's distinctive pattern.

14.2. Hypotheses

This research hypothesizes that six attributes of Latin America's national financial systems have diminished their capacity for financing global‐scope multinationals and have also predisposed them to channel credit to other sorts of uses. The research presents data supporting the assertion that each attribute is an accurate description of Latin American financial systems. It does not prove that each attribute has a damaging effect on the growth of globally integrated firms, and it does not quantify how much damage each attribute caused. Further research may reveal other attributes, but the six given below have been present in Latin American national financial systems. The six attributes, taken as a whole, indicate why those financial systems were not able, and were not set up, to provide the kinds and amounts of financing that a nascent multinational would have needed in order to become an integrated world‐class player.

The six attributes are that Latin American national financial systems have been:

  1. 1. Too small to provide the amount of financing needed to support the growth of large, globally integrated multinational firms;

  2. 2. Often embroiled in crises of liquidity, so they could not provide steady infusions of capital to firms that might have been able to become global if, at pivotal times in their development, they had been able to rely on raising capital;

  3. 3. Weak in providing equity capital to firms, so the firms often had to rely on their own internal profits as their primary source of equity capital, and often had to look to the debt markets for financing, when equity financing would have been more appropriate to their needs;

  4. 4. Too focussed on traditional forms of lending, i.e. financing exports and imports;

  5. 5. Too inclined to loan money to the national governments, more than was the custom in other parts of the world, instead of lending to private firms; and

  6. (p.311)
  7. 6. Designed in a way that gave preference to local, single‐stage firms engaged in commodity production and export, instead of preferring to finance the growth of firms with far‐flung, integrated operations.

For the first three of these attributes, evidence is abundant and relatively easy to gather. Also the damaging effect of these three attributes on enterprise creation and expansion is obvious. This chapter provides recent evidence showing that Latin American financial systems do have the first three attributes.

The fourth, fifth, and sixth attributes are also damaging; however, evidence is harder to gather and will not be as convincing to skeptics, because other national financial systems outside Latin America have also suffered from those attributes at times. Companies in those other parts of the world were sometimes able to become global players, so the fourth, fifth, and sixth attributes are not as assuredly damaging to the growth of world‐spanning enterprises. Consequently, it will be harder to argue that the presence of the fourth, fifth, and sixth attributes caused the underperformance in creating world‐spanning enterprises. With regard to the last three attributes, therefore, the objective in this chapter is only to show that they are accurate characterizations of Latin American national financial systems. Some readers will infer causality but others will remain skeptical.

By showing that six attributes are accurate characterizations of Latin American national financial systems, this research seeks to draw attention to the role that national financial systems have played in the pattern of business development in Latin America, and particularly in the development of large enterprises in the region. That is a part of the explanation for the chronic underperformance of Latin American economies, and merits full consideration along with the other components of the explanation. In the sections that follow, there is a discussion of the existing literature; then the existing literature is cited in providing evidence that the six attributes do accurately describe Latin American national financial systems; then there is a discussion bringing in relevant supporting topics, including the savings rate, comparisons to the Asian Tigers, and the recent rise in the region's savings rate; then there is a brief conclusion.

14.3. Review of Existing Literature

There is an extensive literature on Latin American economic development, and also an extensive literature on monetary policies and exchange rate policies in the region. Topics include the role of central banks, price controls, exchange control, and currency boards. There is not, however, such an extensive literature on how the national financial systems in Latin America (p.312) gather and allocate capital. There are entire books on capital flight, and an extensive literature on informal economies. There are not as many articles that dig deeply into the motives of people who move money out of Latin America, nor why so many people in the region choose to operate in the informal sector.

The literature is surprisingly silent on how national financial systems in Latin America gather the savings of the local wage earners. It is easy to find references to the region's low savings rate. Most of the articles tacitly assume that the savings rates really are low, but some mention that the true savings rates may be higher than the reported figures, because some savings leak out, to be deposited in financial institutions outside the region.

The literature is even more silent on how the national financial systems allocate capital. There is now a growing literature on venture capital in Latin America, and on micro‐lending, and there are also studies of the Chilean‐style private pension fund schemes that have spread through the region. There are, however, very few studies documenting which kinds of projects got financing and which ones did not get financing. A possible explanation for this gap in the literature is that the post–World War II efforts to spur economic development in Latin America emphasized industrialization and export development. Capital was an input to both, and was chronically scarce. The existing financial institutions did not provide enough financing, and often perversely directed capital to activities like cattle raising, which were seen as less dynamic and less productive than industrial projects. In consequence, there was a tendency to dismiss the existing financial institutions as part of the old, tradition‐bound economic order. The existing financial institutions were interesting to researchers who wanted to lay bare the mechanisms by which rich families controlled the levers of power in Latin American countries. But they were not interesting to many experts on economic development, who often argued for leapfrogging the existing financial institutions and setting up government entities to finance the new manufacturing enterprises that the industrialization and export promotion policies required.

For this research the need, in view of the gaps in the literature, has been to cast a wide net, extending the search to related fields, to find relevant sources. The approach we have taken is to cite the sources that we have found in the sections of the paper that discuss each attribute of Latin American financial systems. In providing evidence and citations to support the discussion of each of the six attributes, we have used the most recent sources, and the best that we could find. Further research may uncover additional sources, but we do not expect to find sources that make specific causal links between the attributes of the region's financial systems and the types of large enterprises that emerged.

(p.313) 14.3.1. Small Size as a Limiting Factor

National financial systems in Latin America have been small for the entire history of the region. They were small during the colonial period, in the decades following independence, and continue to be small during the modern era. At the end of the colonial period, national financial systems were so subordinate and insignificant that the standard histories of the region scarcely mention them at all. Those histories treat the national financial systems as part of the set of institutions that the new elites took over after freeing themselves from their Spanish and Portuguese overlords. The financial systems are mentioned only in passing in those histories, and the few comments dedicated to them indicate how little importance authors attach to them. For example, Herring (1968) states that ‘much of the chief business in banking and trading had been in the hands of peninsulares who were now banished’ (p. 284). The author points out in many passages how the underprivileged were mistreated, and how the many countries of the region enacted new constitutions, or formed new governments, but makes no further comment about national financial systems until he describes the high inflation rates of the post–World War II commodities boom.

At the time when Latin American countries achieved their independence from Spain and Portugal, the large financial centers were London, Paris, and Amsterdam. Madrid was still repaying the debts that the Spanish monarchy incurred during previous generations, so it was not a place where large amounts of money could be raised. Lisbon was not the center of a large enough economic zone, and much of the zone under its financial suzerainty was poor. In consequence it could not offer large amounts of financing for ventures in Brazil. In the richer Latin American countries, during the colonial period the national financial systems facilitated the pattern of development that was in fashion, and that pattern did not require the financial systems to be proactive or large. Their small‐sized and subordinate role was consistent with the mercantilist trading paradigm that assigned a conventional and delimited economic role to the Latin American colonies during those years. The focus of the colonial economies was to export commodities, and the merchants in the capital cities who exported those commodities were a key constituency for the monarchy's designated local representative. Much of the financial paperwork was done in Spain, and the creoles were prohibited from doing that kind of work.

Another function those colonial financial systems performed was to finance the imports that the commodity production system needed, and the imports of luxury goods that the elite consumed. There were some instances when manufacturing in the colonies took place, but there was no financing for it from the official financial systems. As suggested by Chapman (1938) and Harding (1947), the colonial paradigm explicitly prohibited manufacturing (p.314) enterprises in many parts of Latin America, and the financing of them was therefore also prohibited.

Data from the 1980s indicated that the region's financial systems were still small, but by that time there was more awareness of how much damage they could cause. Writers no longer dismissed them as insignificant. The policy of relying on foreign capital and foreign financial institutions appeared to aggravate cyclical downturns, causing them to spiral into fully fledged depressions. There was an outpouring of research on financial systems after the Third World debt crisis broke out in August 1982. In the Eighties, there were careful attempts to measure capital flight. Everyone knew that it existed, and had a vague notion that it was harmful, but they did not attribute to it as much causality as it deserved. Then the economic stagnation persisted for almost a full decade, and when attention turned to the financial system, researchers uncovered some striking facts. Williamson and Lessard (1987) express one illustrative fact very clearly. They state that ‘much of the accumulation of foreign assets by the domestic private sector in Latin America in recent years has coincided with external borrowing by the sovereign’ (p. 20). That finding confirmed the suspicion that Latin American governments borrowed foreign currency in London and New York, which then ‘leaked’ away into Swiss bank accounts belonging to wealthy individuals who enriched themselves. The Latin American countries found themselves with large debts and little new capacity to service the debts.

The track record of the national financial systems of Latin America has not always attracted as much attention as it did from those two researchers, but those systems, whether under scrutiny or not, have continued to underperform compared to the systems in the rich countries.

Data from the IMF's Global Financial Stability Report (2005) give a measure of how small these financial systems have remained despite their recent rapid growth. For 2004, total bonds, equities, and bank assets in Latin America were US$2,739.4 billion, an amount equal to 136 percent of the region's GDP. That percentage compares unfavorably to the proportions of the financial systems in the industrial countries, which range from 358.6 percent for Canada to 413.0 percent for Japan.

The comparison with rich countries like Canada and Japan shows how small the region's national financial systems are in comparison to the systems in the rich countries. For this discussion it is more relevant to compare the Latin American systems to the ones in other emerging countries, particularly the countries that were able to nurture companies that have become integrated global giants. This comparison will reveal whether the small size of Latin America's financial systems was a plausible explanation for the region's underperformance in spawning large multinationals.

This comparison with other emerging countries reveals that the national financial systems of Latin America have performed worse than the ones (p.315) emerging in Asia and Eastern Europe in the key tasks of gathering and allocating capital (IMF 2005). Aggregate figures for the size of the capital markets show this clearly. For example, the IMF (2004) reported that in the emerging Asian countries total bonds, stocks, and bank assets were equal to 242.2 percent of GDP at the end of 2002 and 250.6 percent of GDP at the end of 2004. That is surprising because the national financial systems in emerging Asia and Eastern Europe have had very well publicized problems that have hampered their performance.

It should have been easy to outperform the national financial systems of emerging Asia because the Asian Crisis of 1997 revealed such deep flaws in the design of those countries' financial systems. The Asian Crisis was severe and recovery from it was slow, so its lingering effects should have allowed Latin American financial systems to outdistance their counterparts in emerging Asia. The crisis caused the financial systems of the Asian Tigers to shrink, then to spend time repairing the damage. But that is not the only reason why Latin American financial systems should have been able to surpass the ones in the Asian Tiger countries. The financial systems in the emerging Asian countries continue to be bank‐dominated. During the years before the Asian Crisis, companies in emerging Asia relied on bank financing and did not issue bonds very frequently. Stock markets were small and volatile and companies did not use them very often to raise equity capital. By 2002, the real economies of the Asian Tiger countries had recovered from the crisis, but the bond and stock markets in those countries were still not the dominant providers of capital. As of the end of 2002, bank assets constituted 58.4 percent of the total of bonds, common stock, and bank assets in emerging Asia (IMF 2005). Stock markets in emerging Asia rallied in the years 2003 and 2004, but as of the end of 2004 banks still remained the largest providers of capital, with bank assets accounting for 50.8 percent or the total of bonds, stocks, and bank assets (IMF 2005).

The comparison with emerging Asian countries indicates that the financial systems in emerging Asia were larger in comparison to GDP than their counterparts in Latin America. They achieved this size advantage despite being bank‐dominated, and despite remaining excessively reliant on banks as allocators of capital longer than the Latin American systems did. The Latin American national financial systems became capital‐market‐dominated quite rapidly. That is a step forward, because capital market dominated systems are more transparent, suffer less from moral hazard, and are less vulnerable to domino‐like collapses. The transformation to capital market domination happened rapidly in Latin America. Data for 2002–4 are illustrative. Bank deposits in Latin America at the end of 2002 were 44.9 percent of the total of bonds, stocks, and bank deposits. By the end of 2004 that figure had fallen to 33.6 percent of the total (IMF 2005).

(p.316) In Eastern Europe, central planning gave way to market systems. This was a difficult transition, and the procedures for gathering and allocating capital went through a painful overhaul. The capital markets in that region are still small, and have grown slowly. Total bonds, stock, and bank assets were 81.8 percent of GDP at the end of 2002, and declined to 69.8 percent by the end of 2004 (IMF 2005). Bank assets grew more slowly, dropping from 33.4 percent of GDP at the end of 2002 to 27.1 percent by the end of 2004 (IMF 2005).

The evidence above indicates that the size of Latin American capital markets was small compared to the rich countries and to the countries of emerging Asia. The size of capital markets gives an indication of how much capital is available to be allocated during each time period. When capital is more abundant, economic growth can be higher, and large enterprises can get more financing to take advantage of growth opportunities. When capital is abundant enough, the allocation process can be inefficient and can still result in high economic growth. Also, large enterprises have a better chance of getting enough capital to achieve global standing.

In assessing these data, it is clear that Latin America's financial systems performed better in 2002–4 than they did in the Eighties. But they did not grow fast enough to outpace the Asian Tigers, despite the huge stumble that the Asian Tiger financial systems took in 1997. The Latin American financial systems did manage to stay ahead of the ones in Eastern Europe, at least in size. The comparison with the emerging countries of Eastern Europe shows that those systems have not yet been able to gather and allocate comparable amounts of capital. The Latin American financial systems also have another accomplishment to their credit. In the aggregate they were more successful than the Asian Tigers in diminishing the degree of dominance of commercial banks in the capital allocation process.

14.3.2. Crises of Liquidity as a Limiting Factor

During the past two centuries, every country, with the possible exception of Switzerland, has had financial crises. The frequency and severity of these crises have affected the amounts of financing that growing companies can obtain. Liquidity crises have also had a less obvious effect. The more frequent and severe they are, the more they bias companies against using debt financing. To see why, consider two risks of using debt financing. One is that the income stream of the business may be inadequate to repay the company's debts. This happens during the most typical type of financial crisis, in which the money supply shrinks rapidly, and prices of goods and services often fall, so the borrower's output may not bring in enough revenue to pay the debt. This also happens during hyperinflation, when the borrower has a hard time maintaining the profit margin needed to service debt. The second risk is that (p.317) the company's short‐term lenders may be unwilling to renew their loans to the company. Because of those two risks, in countries where financial crises are frequent, companies that survive are careful not to expand faster than their rate of internal cash generation permits.

In Latin America, there are many examples of family‐controlled businesses that choose not to expand as much as the market will allow. They grow by reinvesting some of the profits, but do not seek debt or equity financing to expand faster. A sociological explanation is that the owners are engaging in ‘satisficing’ behavior—that is, they have reached a level of income where the marginal utility of more income falls so low that they choose to spend more time on leisure activities instead of working longer hours. An explanation from the theory of finance is that their business is vulnerable to downturns, so they are wary of using debt financing, because sooner or later debt turns out to be very expensive or very risky.

If Latin American businesses are indeed vulnerable to cyclical downturns and crises in the national financial systems, the observed aversion to using debt could be the result of survivorship bias. This bias distorts many statistical studies of financial performance because investigators inadvertently ignore the firms that did not survive, and so put too much weight on the behavior and performance of the ones that did survive. In the Latin American context, it could have happened that firms with the most conservative financial policies had a higher rate of survival. If the ones that shunned debt survived and the ones that used too much debt perished, an observer would, at any point in time, see more companies that shun debt. If that description corresponds to reality, an implication would be that older Latin American business executives would display a visceral aversion to debt, because they would have seen many good businesses fail as a result of owing too much money and being unable to repay it.

This possibility suggests an explanation for the observed pattern of enterprise development in Latin America. A worldwide macro‐level version of survivorship bias may have been in operation. Suppose that one group of countries found a way of designing its financial systems so that they are more stable, and suppose that another group of countries does not improve the design of its financial systems, so that in that group of countries frequent financial crises continue to occur. Also suppose that in both groups of countries at some starting date in the past there were family businesses that potentially could have become globally integrated multinational firms. In the countries with stable financial systems, the firms that expanded are more likely to be the survivors, and at some point in their development they may have gained an advantage by expanding overseas. In the countries with unstable financial systems, the firms that were more cautious about using debt financing are the ones that were more likely to survive. Those cautious firms expand only using internally generated funds, and do not venture abroad. An (p.318) observer who arrives several decades after the starting date would find that, with several striking exceptions, globally integrated multinationals arose in the countries that developed stable financial systems sooner.

There would be two sorts of exceptions to this pattern. In the natural resources sector there would be large quasi‐government enterprises, which arose in the countries where the natural resources are located, and used their political leverage to obtain the capital they needed. The other sort would be the globally integrated multinationals from South Korea, Taiwan, India, and China. These multinationals arose despite being headquartered in countries with national financial systems organized along traditional lines—bank‐dominated, or centrally planned, and vulnerable to crises. The Asian countries where they arose put high national priority on supporting the growth of large multinational firms, and succeeded by overriding every obstacle, including the tradition‐bound lending practices of private banks. The evidence indicates that these Asian countries steered capital to the companies that became large, world‐class players. Describing South Korea, Martin Hart‐Landsberg (1988) wrote, ‘South Korea went through an extended period of domestic industrialization. … It was a forced march, directed by military dictatorship, but it did raise living standards’. The Latin American countries, with the exception of Brazil, did not have such centrally directed policies of industrial development.

For the Asian Tiger countries, the regional financial crisis of 1997 was a severe blow, which called into question the feasibility of building an economic miracle in the goods producing sector without remedying the inadequacies of the national financial systems. Many high‐ranking people in the Asian Tiger countries felt that the severity of the crisis was out of proportion to the sins that some borrowers and lenders in the Asian Tiger economies had committed.1 The financial systems in the Asian Tigers had been handmaidens of the industrialization process. Their sudden collapse was doing so much damage that for several months skeptics were questioning the viability of the Asian Tiger economic growth strategy.

The response to the crisis varied. India and China escaped financial contagion because their financial systems were closed. Malaysia instituted capital controls. Thailand's and Indonesia's financial systems crashed, and their recovery was slow and painful. South Korea was a long way from the epicenter of the crisis, and many South Koreans felt that their country should not have been engulfed in it. South Koreans understood their financial system, and at first believed that the world financial market was overreacting. Later they acknowledged that the chaebol had expanded too rapidly, and had borrowed too much. South Koreans grudgingly accepted the need for consolidation in the banking sector and better disclosure regulations. They did not fully accept the need for the subsidiaries of a chaebol to cease guaranteeing the debts of the other subsidiaries.2

(p.319) The South Korean response to the financial crisis was instructive and contrasts sharply with the Latin American response to that same crisis. Several of the South Korean chaebol were in financial distress after the Asian Crisis broke out. If South Korea had adopted a pure laissez‐faire policy, the damage to the country's carefully constructed industrial powerhouse might have been profound. The South Korean response was to circle the wagons and preserve as much of the existing structure of corporate entities as they could. The de facto national economic policy was to negotiate debts while ramping up exports. The country also subjected its financial system to careful scrutiny, first to understand what caused the vulnerability, and then to reform it. Meanwhile, the top priority for financial policy was to direct the export revenues to paying debt. In that fashion South Korea weathered the Asian Crisis (Lodge and St. George 1998).

The Asian Crisis spread across the Pacific and affected Latin America. It is revealing to compare how Chile, Brazil, and Argentina responded to the financial contagion from that crisis. Chile found that it had become a dumping ground for Asian manufactured goods, particularly South Korean goods. Chile's response was to allow its currency to weaken 10 percent versus the US dollar. That was not the only response. For the preceding decade Chile had regulated inflows of foreign portfolio investment by imposing a penalty if the foreign portfolio investor withdrew the money before one full year had elapsed. This regulation was called the encaje. It had the effect of raising the real interest rate in Chile, and is one of the reasons why that country's famous retirement system delivered such strong returns.

Toward the end of 1997, Chile lowered the encaje to zero, in effect eliminating it. Short‐term portfolio investment then flowed in, and quickly lowered the real rate of interest in Chile. This might sound like a boon to the economy, but it was not. Capital became cheaper to obtain in Chile, but the companies that had access to it did not take the opportunity to expand. Instead the economy stagnated. Large Chilean companies, which had expanded into other South American countries in the early Nineties, did not continue expanding abroad. Manufacturing companies that had survived the Seventies and Eighties, when Chile repealed tariffs on imports, suffered, and many disappeared. There were 60,000 manufacturing jobs in the shoe industry, and they disappeared, with the last ones going away after 1997.3

The real rate of interest remained low in Chile from 1998–2002, but did not have the hoped‐for effect of stimulating economic growth. Borrowers were cautious, and government incentives to invest in new export‐oriented businesses did not coax many companies to undertake new investments. According to the Central Bank of Chile (2003), returns on pension accounts fell, and the country's real growth rate was between 0 and 2 percent per year from 1998–2002. The Chilean economy did not resume its 6 percent real (p.320) growth rate until 2003. By that time, South Korea's exports were 42 percent above their 1997 level (ISI Emerging Markets Database 2003).

Brazil and Argentina both suffered following the Asian Crisis. They each were trying to impose monetary discipline and each had set a fixed parity with the US dollar. Each country succeeded in maintaining the fixed parity after the Asian Crisis broke out, but Brazil wavered after the Russian default of August 1998. Brazil struggled to stem the outflows of foreign exchange and the speculative pressure on the fixed parity in the last four months of 1998, and then abandoned the fixed parity in January 1999. The Brazilian real, which was at par with the US dollar in mid‐1996, had weakened to 1.20 per US dollar by the beginning of 1999. Then, in the middle of January 1999, it shot up and reached 2 reales per US dollar by the end of February (Federal Reserve Bank of St. Louis 2000).

Brazil was powerless to prevent the sudden weakening of its currency. The effects on its large companies were damaging because they owed foreign currency. The effects on Argentina were delayed but ultimately more serious. Argentina's famous fixed exchange rate ultimately could not survive the devaluation in Brazil.

Chile, Brazil, and Argentina each responded to the one‐two punch of the Asian Crisis and the Russian default in different ways. Their responses, however, were similar in that they all allowed local firms, including large ones that operated internationally, to suffer. The South Korean response was strikingly different. In South Korea the priority was to preserve the large manufacturing companies that were the country's leading exporters, and to keep the whole structure of large firms as intact as possible. In Latin America, the priority was not as focussed on maintaining the health of the large firms. In Chile, many of the large companies were controlled by Spanish companies, so there was not a strong sense of ownership or national interest when the large companies stopped expanding internationally or withdrew their listings from the Santiago Stock Exchange. At the time when these large Chilean companies were suffering from the financial aftermath of the Asian Crisis, other priorities in Chile were more pressing. In Argentina, the collapse of the convertibility plan left the big public utilities with a mismatch. Their revenues in devalued pesos were no longer adequate to service their debts. The effect was to make the common shares of those companies almost worthless. For example, the market price of Telefónica de Argentina shares in New York reached US$50.00 on March 3, 2000, and then crashed to US$2.00 on October 9, 2002.4 For comparison, shares of Korea Electric Power traded in New York at US$21.37 in February 1997, and fell to US$6.27 on June 16, 1998.5 That is a 96 percent drop for the Argentinian stock and a 61 percent drop for the South Korean stock. The severity of the price declines is only one indication of each country's response to the financial crises, but the comparison does measure the losses that shareholders suffered.

(p.321) Crises of liquidity, in summary, have had a deleterious effect on the growth of large companies headquartered in Latin American countries. The comparison with the effect of crises on the large companies in the Asian Tiger countries, and how those countries tried to preserve their large companies, is noteworthy, and deserves to be an important part of the conventional explanation for the paucity of large integrated Latin American multinationals.

14.3.3. Equity Offerings as a Limiting Factor

From 1988 to 1993 Latin America was experiencing a stock market boom. During that time there was a wave of initial public offerings, and Latin American stock markets rallied strongly. Then, following the Tequila Crisis of December 1994, those equity markets crashed, trading volume declined, and initial public offerings came almost to a complete halt. From 1995 to 2002 many companies went private because the paperwork associated with maintaining the public listing was more trouble than it was worth (Chong and Lopez‐de‐Silanes 2003). Data from Chile show this pattern very clearly. There was a burst of initial public offerings in 1990–2 in response to a tax incentive, and by 1994 there were 335 companies listed on the Santiago Stock Exchange. This number gradually declined to 288 by 2001. The volume of trading fell by 45 percent from 1995 to 2002. Stock market capitalization stagnated, rising from the post‐crash low of US$27.3 billion at the end of 1994 to only US$34.3 billion at the end of 2002.6

Figures for initial public offerings in Chile reveal that the stock market had become insignificant as a source of capital. The number of initial public offerings in 1999 was three. There were four in 2000 and three in 2001. Then in 2002 there was only one! In 2003, the number rose to two (Abuhadba 2004). Finally, in 2004 the market took off and there were ten that year and several high profile ones in 2005. Stock market capitalization and trading volume boomed. But by then many Chilean companies which had been on track to become multinationals had pulled in their horns. Chilean companies had been expanding internationally, particularly to other South American countries in the early Nineties. They pulled back, and shelved their international aspirations, except in the retail sector, where Chilean companies were expanding aggressively in Peru and laying plans for expansion into Colombia.

ENDESA Chile, a large electric utility, was becoming a multinational in the early 1990s, and then pulled back. Its American Depository Receipts trade on the New York Stock Exchange. In 1994, they traded as high as US$14.58 on October 18, 1994. At that time ENDESA Chile was actively bidding for electric power plants that were being privatized in Argentina, Brazil, and Colombia. When ENDESA Chile won the bid, the company would send Chilean personnel to manage the power plants.

(p.322) Then three events intervened. First, the Tequila Crisis hit, and ENDESA Chile's stock price stagnated. For most of 1995 and 1996 it was in the US$11 to US$12 range. Second, the Asian Crisis hit. The company's stock had struggled up to the $19 level when the crisis hit, only to sag down to US$7.80 by October 1998. The company's stock struggled back up after 1998, but slid again because of a corporate governance scandal, falling as low as US$6.84 in March 2003. From that point it climbed steadily, reaching US$30 on November 17, 2005.

This lackluster stock market performance was surprising. Chile's advantages made ENDESA Chile an obvious candidate to become one of South America's dominant electric utilities. It shied away from that role. Chileans gave many reasons for the company's disturbing unwillingness to continue developing into a continent‐wide dominant player.

One reason they gave was the company's personnel policy. Mid‐career engineers took two‐year assignments managing the power plants the company had acquired in other South American countries, but then had a hard time reintegrating themselves into the management of the company when they came back from these foreign assignments. Two of them complained that when they accepted the foreign assignment they thought it would be a feather in their caps, but after they came back they felt that the foreign assignment had been a detour in their progression to top positions in the company.7

Another reason that Chileans gave why ENDESA Chile pulled in its horns relates to the control of the company. The corporate structure is complicated, but there was a master block of stock that gave the holder effective control. During the second half of the Nineties, the control block was pledged to lenders in Madrid. The owner of the control block ordered the company to direct a large part of its cash flow to dividends, to pay the lenders in Madrid. In effect, the owner of the control block turned the company into a cash cow.8

The third and most well‐publicized reason why ENDESA Chile shied away from additional expansion opportunities in South America is that there was a corporate governance scandal. This was known as the Enersis scandal in Chile. To summarize a convoluted matter, one of the company's officers arranged to transfer control of the company. Most of the shareholders received too little, because Chilean corporate law did not yet give tag‐along rights to minority shareholders. But the person who arranged the deal managed to obtain a price for his shares that was 11,000 times higher than the price the other shareholders received.9 As of 2005, litigation was still continuing about this transaction.

This history of a large Chilean company, which had a very highly qualified engineering staff, and had access to the most modern capital market in South America, is sobering. The growth opportunities were there, and the company was taking advantage of them, then it stopped. The reasons it stopped relate to the financial system, to its management development policy, and to the (p.323) defects in its corporate governance. The story is not entirely discouraging, because the company is now doing well, and anyone who bought the stock in 2003 or 2004 has done well, but the company is still not expanding abroad.

In Brazil, the pattern of equity market development was similar, with slight variations in chronology. The commercial banks competed successfully with the stock market until a change in tax treatment in 2002 made investing in mutual funds more attractive. Around the same time the Novo Mercado was created. This was a new section of the Brazilian stock market. New issues of common stock had to be floated on the Novo Mercado, except in special cases, and the rules of corporate governance in the Novo Mercado made the shares floated there more attractive.

Under the Novo Mercado rules, all shares had to have one vote each, and there could not be any shares with more than one vote. The pattern in the past had been for controlling families to keep a huge percentage of voting control. They did that by creating a class of shares with many votes and keeping those shares for themselves, and offering shares with only one vote or no vote to the public. That practice circumvented one problem but created another. It allowed the controlling families to keep control of the companies without having to subscribe to their pro rata share of each new issue of shares that their companies made. The drawback of that practice was that the shares offered to the public were unattractive. The public did not invest very much in the stock market. As recently as the year 2000, individual investors accounted for only 20.2 percent of the transactions on the Bolsa de Brazil. Institutions (including mutual funds) accounted for only 15.8 percent of the transactions. Foreigners, businesses, and financial institutions including commercial banks accounted for the rest of the transactions.10

Understandably, when the old set of rules was in place, the stock market remained small. Its total capitalization stagnated during the period 2000–2. It was R$441 billion as of the end of 2000, and R$436 billion as of the end of 2002. The amount of equity capital that firms were able to raise there was small compared to the GDP and compared to the opportunities for expansion that Brazilian firms had available to them. Meanwhile, the stock markets in the industrial countries were raising larger amounts of equity capital for firms that could then expand and become global players.

The combined effect of the Novo Mercado legislation and the rise of mutual funds had a dramatic effect from 2003 onward, and by the end of 2004, the picture was different. Stock market capitalization had risen to R$905 billion and the participation of individual investors and institutions including mutual funds in the total volume of trading had risen to 27.5 and 28.1 percent respectively.11 There were eight initial public offerings, by law all in accordance with the Novo Mercado standards of corporate governance, and those offerings raised a total of R$4.5 billion reales.12

(p.324) This impressive upsurge in stock market activity creates new possibilities for Brazilian firms. Previously, the controlling shareholders could raise equity capital only by putting their control at risk or by issuing shares at low multiples of earnings per share. Many of them chose not to expand or to expand only by using debt financing, which carries its own well‐known risks. If Brazilian firms had been able to obtain more equity financing, a larger number of Brazilian firms might have expanded and become globally integrated multinationals.

14.3.4. Focus on Traditional Forms of Lending as a Limiting Factor

In every country the national financial system develops, evolves, and reforms itself concomitant with the struggles among groups that are vying for preeminence. Beginning from the time of independence, there were long‐running rivalries in several Latin American countries. These rivalries played a role in focussing the financial systems on particular kinds of economic activity. There was a consensus that the national financial systems should perform routine functions, for example, provide currency in the form of coin and paper money to lubricate the wheels of local commerce. That consensus, however, was superficial, and masked a deeper discord. The motivations for issuing local currency included, in addition to the familiar ones of providing a means of payment and gaining seigniorage, also to assert authority over each entire country, to preempt provincial caudillos and hacendados from issuing their own scrip or coins.13 The central governments were concerned that provincial strongmen and landowners would become too powerful, and disavow fealty to the central government, so issuing national currencies was a way of asserting the authority of the capital city over the provinces.

The national currencies were legal tender, with privileged status over the local scrip that hacendados in the hinterland sometimes issued. That privileged status created a moral hazard problem. Powerful people in the capital city had an incentive to issue too much of the national currency, relative to the central government's tax revenues, as a way of extending the capital city's economic power and military dominance over the hinterland. The elite in the capital city also had an incentive to restrict the convertibility of the national currency into foreign exchange, and to keep control of the country's reserves of foreign exchange in their own hands.

The historical rivalry between ruling elites in the capital cities and local strongmen running their own fiefdoms is one possible explanation why the national currencies of Latin America often suffered from high rates of inflation and occasional massive devaluations. The monetary policies were sometimes expansionary, both in the sense that we use the term today, and in the sense that would have seemed logical one or two centuries ago: the creation of money in excess of tax revenue would have been seen as an attempt to stimulate commerce, and as an attempt to extend the geographical (p.325) range over which the national currencies ruled as the sole means of effecting transactions. The goal was not only the one that comes to the modern reader's mind, that is, seigniorage; it was more ambitious, seeking to unify each entire national economy, and bring all its far‐flung regions under the sway of the central government. The goal was also to gain profits from all trade within the country, and especially to skim profits from trade involving the country's commodity exports, and to concentrate those profits in the hands of the elite in the capital city, which before independence had to be shared with the Crown.

This view of the role of currency, and the set of institutions that it engendered, is of course a caricature, and did not exist in such manifest form in all Latin American countries for the entire postcolonial period, and might not have existed in such pure form anywhere in the Latin American region. This caricature, nevertheless, is useful because it suggests that several countries committed the error of setting up their financial systems to help win a power struggle. The caricature is also useful in this argument, because it points out how the management of the currency and the chartering of financial institutions could have influenced the path of enterprise development. In this argument, it is useful to introduce the concept of risk aversion, which plays an important part in finance theory. Risk aversion is a rational response to unstable conditions in financial markets. Applying the concept helps explain decisions that investors and managers made in Latin America during several time frames since independence. A currency regime that is predisposed to inflation can inadvertently tilt the national economies toward commodity production, by favoring the known export‐oriented businesses, whose products are priced in foreign currency. It can also raise the economic power of the capital cities, by apportioning a large share of the profits to financiers and exporters in those capital cities.

A differently conceived national financial system might have induced many Latin American economies to become diversified and entrepreneurial, and might have fostered geographical decentralization, with many regional subcenters of wealth and economic power. Financial risk, however, was always present, and access to capital was always difficult, except for projects that had strong sponsorship. Risk aversion therefore argued for investing in activities that central government favored. Brazil's industrial development is a good source of examples of how risk aversion operated. The history of Brazil's industrial development is a lengthy story, with many promising starts that fizzled, and many tantalizing moments of opportunity that did not deliver the results which seemed possible. There were many beginnings in Bahia, Minas Gerais, Rio, and Sao Paulo. If these beginnings had developed fully, they would have given Brazil a geographically decentralized manufacturing sector, and might have transformed Brazil into an exporter of manufactured goods in the early years of the twentieth century. Exports of manufactured goods would (p.326) have shifted economic power away from the landowning families, and would have diminished Brazil's extreme reliance on coffee as its main source of foreign exchange. All that might have happened many decades before the government‐sponsored industrialization policy that accelerated during the 1951–4 presidential term of Getulio Vargas and continued during the 1956–61 presidential term of Juscelino Kubitschek. It is easy to imagine alternative histories for Brazil, but the glimmering possibilities did not materialize. What happened in reality was that the country opted for the centrally directed style of industrialization that these presidents adopted. That style delivered more lasting accomplishments than the de facto laissez‐faire policy that had prevailed during earlier periods. It is tantalizing to speculate what prevented the happy developments that did not occur. The assertion here is that the financial system's focus on traditional forms of lending is partly responsible.

14.3.5. Too Inclined to Lend Money to Their National Governments

Another form of traditional lending was to lend to the government. This would also deprive nascent firms of the capital they needed to become global players. Data compiled by the IMF show that for several Latin American countries during the time frame 1957–70 the local commercial banks increased their lending to their national governments. IMF data also show that during those same years, commercial banks in other countries such as the United States and Singapore were decreasing their lending to their governments, and increasing their lending to private borrowers.

To show that the region's commercial banks were systematically depriving private firms of access to capital and giving preferential access to government borrowers, there needs to be a strong statistical pattern, particularly during a time frame when large multinationals were emerging in other parts of the world. Ideally, this time frame would be a period when there were no big distortions happening in the world economy that would influence lending decisions in Latin American countries. For the statistical evidence to be most convincing, it would be important to choose a time period when there was no pan‐Latin American financial crisis, no boom in commodity prices, no worldwide depression, and no world war. It would also be important to choose a time frame when there was no major shift in the dominant economic growth model.

Unfortunately every time frame for which data are available is contaminated to some degree by confounding influences.14 The time period chosen, 1957–70, was not a period of financial crises, world war, world depression, nor particularly high commodity prices. It was a period when the Asian Tigers were laying the groundwork for their economic rise. Lee Kwan Yew, the architect of Singapore's legendary economic transformation, became prime (p.327) minister in 1959. It was also a period when the European and Japanese economies grew rapidly, having fully recovered from World War II.

In Latin America the complicating factor in 1957–70 was the widespread adoption of import substitution policies. These policies biased lending decisions in favor of investments in manufacturing for the domestic markets. The companies that arose during this time frame were infant industries that depended on preferential access to local markets. They were not oriented toward foreign markets, nor toward foreign expansion.

The import substitution policies, however, should have induced a shift in lending toward the private sector. The IMF data that exist show an unexpectedly contrary pattern. In several Latin American countries there were shifts in lending away from the private sector during this period. Governments borrowed a growing portion of the total capital that was provided by commercial banks. In Peru from 1964–70, Brazil from 1965–70 and in Mexico from 1964–8 this pattern was particularly strong.15 The IMF database does not provide data about commercial bank lending patterns for all Latin American countries for the 1957–70 time frame. The data that the IMF does provide show a disturbing pattern. The data indicate that nascent multinationals in the region, if they existed, would have had difficulty obtaining large amounts of financing from local commercial banks.

Comparisons with commercial bank lending patterns in countries outside the region indicate that the Latin American pattern was not unique. A cross‐sectional analysis for all countries in the world would not reveal a very strong pattern because the data are incomplete. Also in every country commercial banks routinely lend money to their national governments from time to time. Indeed, many are required to hold government bonds in their portfolios. Ideally, the evidence would indicate that Latin American financial institutions loaned more to their national governments, and did it more often, depriving their private firms of access to capital, and so impeding their growth. The evidence is too incomplete and too mixed to show a decisive result, but the data for several countries outside Latin America do support the assertion.

In two rich countries during that time frame the commercial banks were reducing their loans to their governments and increasing loans to private borrowers. In the Unites States, commercial banks reduced their loans to the governments steadily from 1957 to 1969, and then increased them slightly in 1970. In Switzerland, commercial banks reduced their loans to the government from 1967 to 1970, in a steady downtrend that was interrupted each year by a slight seasonal uptick.16

To compare what was happening in another emerging country outside Latin America at that time, IMF data for Thailand were considered. In Thailand, the pattern of commercial bank lending in 1957–67 was similar to the pattern in Peru for the same time frame. Commercial banks increased their lending to the government.17 This is provocative, because Thailand grew rapidly enough (p.328) to be classified as an Asian Tiger, but Thailand's devaluation was the triggering event for the Asian Crisis of 1997. Subsequent revelations about the lending practices of Thailand's commercial banks made its financial system look like a hotbed of cronyism.18 Thailand's economic growth has been enviable, if volatile, but it is worth noting that Thailand has not given rise to many global multinationals. Perhaps its lending practices are an explanation.

The evidence given here is consistent with the assertion that commercial bank lending patterns had influence on the kinds of business enterprises that arose in each country. It would be premature, however, to state a definitive conclusion. The data are too fragmentary, too many of the time series have gaps in them, and in all the cases discussed here there may be problems of data quality and consistency. The most that can be said is that the evidence supports the hypothesis, but does not support it strongly enough to override other explanations for the observed composition of business enterprises in Latin America.

14.3.6. Design as a Limiting Factor

The centralized, commodity‐export‐focussed way of organizing and running the national financial systems had unintended consequences. These consequences were hard to see at the time but in retrospect are more obvious. The exporters in the capital cities probably did not see the damaging side‐effects of the financial systems that existed in their countries, and might not have been able to imagine any other way of organizing the national financial systems, but modern readers can appreciate what the effects were. One was to discourage small merchants and manufacturers from relying on the local financial institutions. Inflation and devaluation were too severe and too frequent to ignore. Also there was another type of moral hazard that scared away many small potential users of the financial system. The national financial systems were dominated by classic family‐controlled commercial banks. Those banks tended to lend to their affiliated businesses, basing their lending decisions on family ties or the borrower's social standing. That led to dangerous levels of what Latin Americans call préstamos vinculados (loans to related parties).19 The moral hazard was that the entire financial system was vulnerable to collapse, so relying on it would be suicidal in the long run. A national financial system dominated by banks that have too many loans of that sort on their books is unstable, and from time to time collapses, causing losses to depositors, a general contraction of credit, and a lingering distrust of financial institutions.

Lending to related parties is not the only practice that causes the instability of a bank‐centered financial system. Classic commercial banks seem legitimate because they have existed in Europe since at least 1100, but they are unstable. The managers face so many conflicts of interest that it is difficult to give a complete enumeration of them. When a classic commercial bank is the (p.329) centerpiece of a group of family businesses, in a financial system that depends on self‐regulation, the result is too often a domino‐like collapse that spreads to every province and to every sector of the entire economy.

In the past quarter century, the national financial systems in Latin America have become diverse. Some are now market‐centered with independent central banks, while others remain bank‐centered, with central banks under the direct control of the chief of state, but in the decades following independence they were strikingly similar to one another. They all had been established during the colonial period in accord with instructions coming from Spain and Portugal, and after independence they all sought to replicate features of the European‐style national financial systems. In particular, they replicated features of the financial systems of Spain, Portugal, the Italian city‐states, and Paris. They took the idea of merchant banking from Italy and the idea of the stock exchange from the Paris Bourse (Covarrubias 2004).

They did not copy every feature of European financial systems. They did not copy several institutions and practices that were already in widespread use in northern Europe. Surprisingly, they did not copy the idea of issuing government bonds in their local capital markets, which the Dutch had pioneered around 1500. They also did not copy the idea of a central bank, which the Swedish and the English had pioneered prior to 1700. Latin American countries floated bonds in European capital markets within a few decades of independence, but did not try to float bonds in their home markets until much later, and they established central banks as late as 1925 and 1926.20 As for regulation, they relied on the code of the gentleman and self‐regulation. The prevailing view was that financial activity was a private matter. It was like running a non‐financial business. Finance practitioners who acted imprudently or dishonestly would suffer loss of reputation. For a century after independence, that informal pressure was the main mechanism of social control that many countries had over their national financial systems. There were sporadic attempts to treat the financial system as a public good. These sought to fix prices, control the allocation of foreign exchange, and to channel loans to sectors or strata that were not obtaining enough financing. There was tacit recognition that financial booms and collapses have externalities, that is, secondary effects that reach far beyond the parties directly involved. But the efforts to find the right mix of regulation and laissez‐faire lurched and sputtered. In some countries, the usual set of checks and balances are still not fully in place.

The legal system always gave protection to loans secured by mortgages on land (inmuebles and inmobiliaria). Loans secured by circulating assets (enseres and muebles) did not enjoy the same level of protection, so there was a bias against lending for working capital. There were letters of credit and bills of lading, so finished goods for export could serve as collateral, but work in progress did not qualify for financing.21

(p.330) These legal biases must be considered in combination with two facts. First, there was always a shortage of capital compared to the opportunities to invest it profitably. Second, rich families that had extensive landholdings were often in control of commercial banks. The result was that new kinds of activity, different from what the rich families knew best, did not qualify for much financing. Also, there was no market for general‐obligation corporate bonds, secured only by the full faith and credit of the issuing company.

These institutional arrangements denied capital to companies that wanted to integrate vertically, diversify into new lines of business, or expand abroad. Vertical integration shifts the mix of assets, so that the company has more work in progress, and less immovable tangible collateral per unit of value added. Diversifying into new lines of business is riskier than expanding production of an export commodity, which has a cash price in Seville, Amsterdam or London. The returns to investing in new lines of business can be high on average, and still not attract capital in a financial system like the one described. The downside risk is too great, so the riskier sorts of ventures require equity financing, with upside participation.

There were no mechanisms in the national financial systems to bankroll companies that wanted to expand abroad. Overseas assets, in theory, could serve as collateral, but in practice they would not have been able to qualify for financing in competition with local fixed assets. In consequence, if there had been Latin American companies wishing to expand abroad they would have had to find innovative ways of financing their international expansion.

A comparison with financial practices that were typical in Amsterdam, Paris, and London illustrates the difference. In Amsterdam, joint‐stock companies appeared as early as 1520, to finance shipping ventures to the Spice Islands. These ventures often established trading posts in faraway places. Investing in these overseas trading ventures was riskier than investing in letters of credit or government bonds, but offered upside potential. London had its famous money market, where short‐term commercial paper was discounted.22 In Paris, there was a large market for international bonds before World War I. French companies and Latin American governments floated bonds in that market. Latin American companies raised some capital there, but did not use it to finance expansion into other countries. In London, and later in New York, Latin American railroad bonds were issued, but the Latin American railroad companies did not invest in other countries.

Two reasons for the differences in financial practices are (a) in Europe there was more capital and fewer land‐intensive or resource‐intensive opportunities to invest in; (b) in Europe the capital markets allocated a larger portion of the total amount of credit.23 The profits from risky ventures were very high and conventions for distributing the profits gave a large enough share of the gains to the people who took the risk.

(p.331) The national financial systems of Latin America, according to this simplified characterization, gave loans to borrowers who could pledge land, to finance the production and export of primary products. Other feasible investments did not conform to the design of the national financial systems and so did not obtain as much financing.

14.4. Discussion

14.4.1. Recent Success in Gathering Savings

According to Duesenberry (1949), in every society there are people who save and others who do not. He suggests that savings behavior, however, is not absolutely hardwired in the brain of every individual. Savings behavior can be induced, coerced, or discouraged. In recent history, there have been many forced‐savings schemes, and many countries that have become known for their high savings rates. There have also been many schemes that have discouraged savings. Latin America has seen an abundant number of schemes that have discouraged saving, or pushed it into the shadowy world of the parallel economy, such as in the case of Brazil in 1997 (Amann and Baer 2003).

The region's savings rate has always been cited as a reason for its slow economic growth, for its overdependence on foreign borrowing, and for its surprising inability to diversify into more promising kinds of economic activity, whether manufacturing, or skill‐intensive services, or creating globe‐spanning enterprises.

Until recently, official statistics supported this low‐savings characterization, but anyone who has lived in the region for a long time has had personal experiences that call into question the validity of the low‐savings statistics. A stroll through the downtown of many Latin American cities brings one into contact with money changers, who are always anxious to buy dollars. The ultimate buyers of the dollars include many local citizens who are seeking to protect their savings from inflation or devaluation. Another fact is that people who go from Latin America to the United States are often asked to carry envelopes full of greenbacks to relatives living in Miami, who then deposit the cash into an FDIC‐insured bank account in the United States. Those savings are not reported in the statistics. When the money is eventually deposited in a US financial institution, the savings are incorrectly attributed to North Americans. The reality has been that for many years Latin Americans have gone to the trouble of moving their savings out of their home countries, to financial systems in the rich countries, where the risk of inflation, devaluation, or outright dispossession is not as great.

To an outside observer, it appears that the objective of these laws was to force the local citizens to hold the local currency against their will. The (p.332)

Table 14.1. Gross national savings rates by region (percentage of GDP)










Industrial countries







Latin America







Southeast Asia







Source: IMF Database.

governments that imposed the laws never explained their motivations in such coercive terms. Instead, they would speak earnestly of the need to channel resources to high‐priority domestic investment projects, or the need for all citizens to be loyal to nationalist ideals. It often happened, however, that the local currencies would lose some or all of their value. The people who were compelled to hold the local currency would lose and would be resentful. They would often jump to the conclusion that the national financial system was rigged against them. They saw themselves as chumps, and believed that bigger, better‐connected players received prior warning of the devaluation and converted all their holdings of local currency just ahead of it. The victims can be forgiven for forming a negative attitude toward saving, and for swearing that they would not be victimized again.

Very recently several Latin American countries have reformed their national financial systems. These reforms have rewritten those countries' social contracts to include commitments to maintain stable currencies and to provide adequately designed and managed financial institutions where savers can put their money and expect to earn a positive return on it. The results have been dramatic. Table 14.1 illustrates how stark the difference was between Latin America's savings rate, as it was reported in the official statistics, and the famously high savings rate in Southeast Asia during the years 1970–94. Subsequent to the time periods reported in Table 14.1, official savings rates in a few key Latin American countries have risen, and so the regional average has also improved.

14.4.2. Comparisons to Other Emerging Countries

Latin American countries are diverse, and their paths of development have diverged. Nevertheless, it is sometimes helpful to view them as a group. It is also customary and convenient to consider other emerging countries in groups. The Southeast Asian Tigers are the group that has achieved the highest rates of economic growth, for example, 7.8 percent per year for 1987–96. Latin American countries achieved only 2.7 percent per year during that same period, beating Africa, which averaged only 2.2 percent, and trailing the Middle East, with 3.4 percent.24

(p.333) The national financial systems of the Southeast Asian Tigers and the emerging countries in the Middle East were bank‐dominated, as were the financial systems in Latin America. That is, commercial banks allocated more capital than the local stock markets and bond markets did. Despite this similarity, the Southeast Asian emerging countries have been able to foster the growth of more globally integrated multinational companies, and have consistently scored higher in export diversification, skill content of exports, and other categories of competitive success (De Gregorio and Wha Lee 2003).

Data for these groups of emerging countries show clear differences in savings rates and debt service ratios that outweigh the similarities in the design of their national financial systems.

Savings rates in Latin America are still low, but no longer so much lower than the world average. For the period 1983–90, world average savings were 22.8 percent of GDP. The Southeast Asian Tigers achieved an average savings rate of 34.6 percent, while the Middle East and Latin America achieved only 18.4 and 19.8 percent during that period. By 2004 and 2005, Latin America had achieved savings rates of 21.1 and 21.5 percent respectively, not far below the world average of 21.5 and 22.1 percent for those years, but still far below the 33.0 and 32.5 percent savings rates of the newly industrialized Asian countries for 2004 and 2005.25

There is also a striking difference in debt service ratios. The developing countries in Asia spent only 5 percent of their imports of goods and services on debt service payments during the period 1983–90, while the Middle East countries spent 3.1 percent and Latin America and the Caribbean paid 15.6 percent.26 This burdensome ratio has made exports a high priority. At the same time, it biased business decisions against projects that would take large amounts of foreign exchange, or take too many years to reach fruition, even if the projects would ultimately bring in much more foreign exchange than they cost to develop.

Latin America's lower savings rate and higher debt service ratio are possible reasons for the low number of globally integrated firms headquartered in the region. Capital has been chronically scarce and expensive, and the high dependence on foreign capital magnified the instability of the region's economies, which in turn made large capital investment projects riskier, especially if they depended on private sources of capital.

14.4.3. The Recent Rise in Savings and its Implications

The rise in savings has happened very rapidly. The most rise has produced the surprising result that since 1999, the Latin American region has been a net supplier of capital to the world financial market. World financial statistics show this clearly. At the end of 1999, Latin American countries in the aggregate owed US$807.1 billion to lenders outside the region. By the end of 2005, this (p.334) amount was US$842.4 billion. During the 1999–2005 period, the region's foreign exchange reserves increased from US$143.4 billion to US$249.5 billion.27 Subtracting foreign exchange reserves from the amounts owed shows that in 1999 the region's net debt was US$663.7 billion, and by 2005 it had fallen to US$592.9. That is a decline of US$70.8 billion.

This remarkable reduction in net foreign debt was accomplished while local savings were flowing into the national financial systems. Local borrowers, including the governments, were able to replace foreign financing with local financing. In Brazil, for example, the foreign debt has fallen from US$242 billion in March 2000 to US$178 billion in July 2006.28

The rise in savings has often been attributed to the high commodity prices since 2002. A more complete explanation includes the financial reforms that several countries have implemented. Another reason for the rise is the reaction to the Asian Crisis. When that crisis broke out in 1997, it may have prodded many Latin American countries to accelerate the financial reform process. Latin American countries had just recovered from the Tequila Crisis of 1994–5. To their dismay, they found themselves once again over indebted to foreign banks, and again had to suffer another bout of contagion, credit contraction, and economic hardship. It was especially galling that the triggering event had not even occurred in Latin America! Latin American countries were being castigated for sins that they had not committed.

The international financial system was a harsh and capricious taskmaster. It punished Latin America once again for its past sins, and was deaf to the region's protestations that it was innocent this time. The need to repay foreign banks, coming such a short time after the Tequila credit crunch, prompted several Latin American countries to reform their capital markets and to overhaul their financial institutions. The reforms gave the local institutions the capability of attracting local savings. Privatized pension schemes took off in Mexico and Peru during this period. The pension schemes fueled the growth of local bond markets, which had been tiny and had mediated mostly government bonds before then. For example, as of year‐end 2001, the total value of Latin American corporate bonds was only US$23.4 billion; by the end of 2004 that amount had risen to US$215.1 billion.29 After the capital market reforms went into effect, borrowers in many Latin American countries were able to borrow locally by issuing bonds, and thus they no longer needed to borrow abroad.

The economic stagnation that the Latin American region suffered from 1998–2002 was a consequence of the pressure from foreign lenders seeking repayment. The classic way that Latin America has repaid foreign debt has been to create a trade surplus and a surplus of local savings over local investment. Devaluing the local currency creates the trade surplus by making exports more profitable and imports more expensive. Lack of alternative uses for money accomplishes the surplus of savings. New capital investment (p.335) projects look unattractive because there is excess capacity, so citizens put their money into local financial institutions. This harsh corrective mechanism brings the foreign debt back into line with the capacity to service debt, but imposes hardship and austerity on local populations.

This explanation of Latin America's prudent financial performance is harsh and skeptical because it implies that the recent rise in the region's savings rate may be nothing more than an artifact of the imperative to repay foreign debt. If this skeptical view is correct, savings rates in the region will not stay high and will not keep rising. In that case, there would be no permanent improvement in the availability of capital, so Latin American globally integrated firms would not arise in the future.

A more optimistic view of the rapid rise of saving is that local pressures were already at work before the Asian Crisis. Contagion from Asia accelerated reforms that were already in progress. Latin American countries never liked having to rely so heavily on foreign debt, and saw a way of breaking the cycle of dependency. The way was to become more successful in gathering and allocating local savings. There was clearly room for improvement in the national financial systems, and that has been the avenue that several countries have followed with growing success. Aggregate statistics indicate that Latin American financial systems were not only bad at gathering savings. They were also bad at allocating those savings to the most promising uses. Data from the Eighties and Nineties show that for the entire Latin American region, economic growth averaged 2.9 percent per year for 1984–93 and 2.5 percent for 1994–2003. During the overall 20‐year period savings rates were in the 18–19 percent range. That implies a marginal capital/output ratio much higher than the normal acceptable range of three to five. The ratio for the Southeast Asian Tigers was around four during that time period. The implication is that in Latin America capital was being allocated inefficiently. There were probably better uses for that capital. What might have happened is that the same mature, low‐technology sectors continued to attract a large portion of the available capital, while other sectors got less than they could have used productively.

The encouraging implication of the rise in savings is that in the new environment of market‐centered capital allocation, innovative firms may obtain more capital, and may be able to expand and diversify in ways they have been unable to do until now. Reforms to the national financial systems are moving forward in a handful of countries, and soon we will be able to see if this optimistic view is valid.

Until this point we have ignored the handful of Latin American firms that managed to become world‐class multinationals despite the severe limitations of the national financial systems in the countries where they arose. CEMEX is the best‐known example. It arose in Mexico, and managed to become one of the most dominant companies in a capital‐intensive business. That is a (p.336) remarkable accomplishment, and it shows how a truly exceptional entrepreneur can overcome seemingly insurmountable obstacles. CEMEX obtained capital from international markets and found innovative ways of paying for acquisitions, using national capital markets in the countries where its targets were located.30

Another success story has been Cencosud, the Chilean retail conglomerate. It started in 1952 and became an innovator in retailing, but did not need large amounts of capital until 1982, when it expanded into Argentina. Its expansion in Argentina is consistent with the argument put forward in this chapter: Chile's national financial system has functioned better than Argentina's, and consequently the Chilean company has been able to expand into Argentina. Other Chilean retail conglomerates have expanded into Peru and (to a lesser extent) into Colombia.31

14.5. Conclusion

There are many reasons why so few large, globally integrated companies have arisen in Latin America. The assertion put forward here is that the national financial systems of those countries are partly to blame because they were not set up properly to finance the growth of those companies. Assuming that there had been Latin American companies that possessed all the other preconditions necessary to become large, global‐spanning enterprises, the small‐size and idiosyncratic practices of their national financial systems would have prevented them from attaining their potential.

This chapter has given an argument consisting of many related parts, and has provided evidence in support of all the parts. The evidence for the first three parts of the argument is strong. For the other three parts, the evidence is indicative but not absolutely definitive. There can be little doubt, however, that national financial systems played an important part in creating the mix of economic activities and the kinds and sizes of business enterprises that we observe today. The objective here has been to draw attention to the role that national financial systems played. If this research achieves its purpose, the debate on Latin American business development will henceforth include the question of how the credit allocation practices in the region have operated, and how they are being reformed, so that the Latin American region will, from now on, be able to create its share of the large, globe‐spanning companies in the world.


Bibliography references:

Abuhadba, M. (2004). IPOs en Chile: Salfacorp y el Underpricing. (Informe Área de Negocios No. # 25). Santiago de Chile, Chile: Universidad Santo Tomás.

Amann, E. and Baer, W. (2003). From the Developmental to the Regulatory State: The Transformation of the Government's Impact on the Brazilian Economy. (Session 309/ECO 002, Congress of the Latin American Studies Association). Dallas, TX.

Chapman, C. E. (1938). Colonial Hispanic America: A History. New York: Macmillan.

(p.339) Chong, A. and Lopez‐De‐Silanes, F. (2003). The Truth about Privatization in Latin America. (Research Network Working Paper No. R‐486). Washington, DC: Inter‐American Development Bank.

Covarrubias, I. (2004). La Economía Medieval y la Emergencia del Capitalismo. Texto completo en http://www.eumed.net/cursecon/libreria/index.htm

De Gregorio, J. and Wha Lee, J. (2003). Growth and Adjustment in East Asia and Latin America. (Working Paper No. 245). Santiago, Chile: Central Bank of Chile.

Duesenberry, J. (1949). Income, Saving and the Theory of Consumer Behavior. Cambridge, MA: Harvard University Press.

Harding, C. H. (1947). The Spanish Empire in America. New York: Oxford University Press.

Hart‐Landsberg, M. (1998). The Asian Crisis: Causes and Consequences. Against the Current, 73 (March/April).

Herring, H. C. (1968), A History of Latin America, from the Beginnings to the Present. Random House; 3rd edition.

International Monetary Fund Global Financial Stability Report September 2003, Statistical Appendix, p. 143.

International Monetary Fund Global Financial Stability Report September 2005 http://imf.org/External/Pubs/FT/GFSR/2005/02/index.htm.

International Monetary Fund Global Financial Stability Report, September 2005, Statistical Appendix, p. 171.

Lodge, G. and St. George, A. (1998). Collapse in Asia—1997–98. (HBS Case study No. 9‐798‐084). Cambridge, MA: Harvard University.

Williamson, J. and Lessard, D. R. (1987). Capital Flight and Third World Debt. Washington, D.C.: Institute for International Economics.


() The author received generous support for this research from the W.F. Glavin Center for Global Management at Babson College and is grateful for Mario Gutierrez (M.A., Brandeis University) for many valuable contributions to this project.

(1.) Famously, Mahathir Mohamad, prime minister of Malaysia, blamed the crisis on George Soros and other currency speculators, not on lending practices in the Asian (p.337) Tiger countries. ‘Mahathir vs. Soros: A plan to weaken the region's currencies?’ Asia Week, August 8, 1997, Web edition http://www.asiaweek.com/asiaweek/97/0808/ nat2.html

(2.) South Korean bankers expressed this view to the author during a banking seminar in Seoul, South Korea in December 2003.

(3.) Professor Francisco Arroyo of the University of Chile cited this figure in conversations with the author in 2004.

(4.) Quotes are for the American Depository Receipt. Source: Yahoo Finance Website.

(5.) Quotes are also for the American Depository Receipt. Ibid.

(6.) Data from the Superintendencia de Valores y Seguros (Chile) website, Mercado Accionario statistics.

(7.) During the summer of 1997, the author taught at the Universidad Adolfo Ibanez and these two people were students in the author's class. They were studying for their MBAs with a view toward changing careers.

(8.) In 2002, Professor Roberto Darrigrandi and the author were teaching a class session in the MBA Program at the Universidad del Desarrollo in Concepcion, Chile. This explanation was provided by a student in the class. Other students confirmed this explanation.

(9.) Professor Roberto Bonifaz of the Universidad Adolfo Ibanez explained this long‐running scandal to me in those terms. The full details were very complicated and some of them are in dispute.

(10.) Bovespa.com:br/Relatiorio Annual 2004, p. 23.

(11.) Ibid.

(12.) Bovespa, op. cit., p. 31.

(13.) Modern practices differ from the ones that were considered normal at the time of independence from Spain. The idea of the ‘money supply’ scarcely existed, and what today we call macroeconomic policy was a very minor part of the responsibilities of central government. Governments maintained peace and protected citizens from foreign invasion. Weather and crop yields were the key determinants of macroeconomic conditions. For a detailed view of what motivated heads of state to issue currency in previous centuries, see Alexander del Mar, History of Monetary Systems, (facsimile reprint of the edition published by Charles H. Carr and Company, Chicago, 1896), passim.

(14.) The author is indebted to Professor Harvey Arbelaez of the Monterey Institute of International Business for this insight.

(15.) Data collected from the IMF International Financial Statistics by Vinay Kaza.

(16.) Ibid.

(17.) Ibid.

(18.) ‘Mid‐Year Economic Review 2002,’ The Bangkok Post, Web edition, http://www. bangkokpost.net/midyear2002/banking.html

(19.) For a recent article on ‘préstamos vinculados’ see ‘La Falta de Educacion Financiera en el Ecuador’, Amalia Verdezoto Vidal, Americ Economia on‐line edition, July 5, 2002, http://www.americaeconomica.com/numeros3/168/reportajes/amalia168.htm

(20.) Bolivia and Chile.

(21.) An early reference to this distinction between classes of collateral, and the way that European financial systems dealt with them, is provided by Frederic L. Nussbaum, (p.338) ‘A History of the Economic Institutions of Modern Europe’, originally published in 1935, reprinted by Augustus M. Kelly, Publishers, New York: 1968, pp. 297–298. This distinction between mortgage financing and various mechanisms for lending secured by inventory or accounts receivable was an important feature of the system that Latin American countries imported from Spain and France.

(22.) Walter Bagehot, the first editor of the Economist, describes this market in his classic book Lombard Street, referenced in the following note.

(23.) In the classic book Lombard Street: A Description of the Money Market, (New Edition, New York: E. F. Dutton and Company, 1910) Walter Bagehot gives a description of the commercial paper discounting process which he says operated in London during the years 1870–1910 on a much larger scale than anywhere else. With regard to the developing countries he wrote, ‘there is no large sum of transferable money; there is no fund from which you can borrow, and out of which you can make immense works … it is certain that in the poor states there is no spare money for new and great undertakings, and that in most rich states the money is too scattered … ’ pp. 7–8.

(24.) International Monetary Fund, World Economic Outlook, September 2005, p. 211.

(25.) International Monetary Fund, World Economic Outlook, September 2005, p. 277.

(26.) International Monetary Fund, World Economic Outlook, September 2005, p. 273.

(27.) Data from the International Monetary Fund, World Economic Outlook, September 2005, pp. 269 and 273.

(28.) ISI emerging markets Database.

(29.) Data from the IMF World Financial Stability Report, 2002 and 2005.

(30.) The remarkable success of CEMEX and its financial policies have been profiled many times. Ravi Sarathy and David T. A. Wesley wrote one of the best accounts of the company's financial expertise, CEMEX: The Southdown Offer, 2003, case number 903M13, Richard Ivey School of Business.

(31.) The history of Cencosud has been a favorite topic for MBA students in Chile to study and debate. The ‘retail war’ that has been going on among the top Chilean chains (including Falabella, D&S, and Ripley) was front‐page news in Chilean financial newspapers during 2004 and 2005. The expansion into Peru and Colombia was portrayed as a way of gaining advantage in that war, by making forays into virgin territory. For a history of Cencosud, see its corporate website www.cencosud.cl. For a discussion of the retail wars, see El Diario Financiero and Estrategia, Chile's two daily financial newspapers.