Journal of Development Economics 19 (1985) 1-24. North-Holland GOOD-BYE FINANCIAL REPRESSION, HELLO FINANCIAL CRASH Carlos DIAZ-ALEJANDRO* Columbia University, New York, NY 10027, USA Received December 1983, final version received March 1984 Some unintended consequences of financial liberalization in Latin America are analyzed in this paper. Intrinsic imperfectious in financial markets, and the policy dilemmas they pose, are reviewed first. The stylized facts of Southern Cone experiments in financial liberization are then presented. Alternative ways of organlzing domestic capital markets under Latin American conditions are discussed, and an eclectic, reform mongering program is proposed. 1. Introduction This paper seeks to understand why financial reforms carried out in several Latin American countries during the 1970s, aimed at ending 'financial repression', as defined by Ronald McKinnon see, for example, McKinnon (1980), and generally seeking to free domestic capital markets from usury laws and other alleged government-induced distortions, yielded by 1983 domestic financial sectors characterized by widespread bankruptcies, massive government interventions or nationalizations of private institutions, and low domestic savings. The clearest example of this paradox is Chile, which has shown the world yet another road to a de facto socialized banking system. Argentina and Uruguay show similar trends, which can be detected less neatly in other developing countries, including Turkey. Indeed, even in the United States some worried observers have warned against the dangers of pell-mell deregulation of commercial banks see Kareken (1981). *A version of this paper was first presented at a conference held in Bogota on November 22 and 23, 1982 -- sponsored by the Universidad de Los Andes and the Banco de la Rep6blica. ,Visits to CINVES, in Montevideo, and CIEPLAN, in Santiago de Chile, during March 1983, were very helpful for extending that early version. I am grateful to all these institutions and to numerous individuals who shared with me their views of the events narrated and of the issues discussed in the text. Among those who are unlikely to find embarrassing my acknowledgement of their help, I may mention Jos6 Pablo Arellano, Edmar L. Bacha, Guillermo Calvo, Jonathan Eaton, Richardo Ffrench-Davis, Arminio Fraga Netto, Jeff Frieden, Eduardo Garcia d'Acufia, Jose Antonio Ocampo, Hugh Patrick, Gustav Ranis, Patricio Meller, Miguel Urrutia and Laurence M. Weiss. My gratitude is no less for the often extraordinarily generous cooperation of those not explicitly named. 2 c. Diaz-Alejandro, Consequences of financial noitazilarebil The paper will first review dilemmas posed by intrinsic imperfections of any financial market, and will look at legal prerequisities for the reasonably efficient operation of those markets. Then it will examine the stylized facts of some Latin American experiments in financial liberalization, particularly those carded out in Southern Cone countries (Argentina, Chile and Uruguay). Finally, alternative ways of organizing domestic capital markets under Latin American conditions will be discussed; policies regarding the links between domestic and international financial markets will also be considered. 2. Peculiarities and dilemmas found in domestic financial markets Are banks special, and really all that different from butcher shops? Neither type of firm is exactly like the textbook idealization of the atomistic firm operating in a perfectly competitive market, where spot prices summarize all information relevant for buyers and sellers of the product. Customers at a butcher shop will not only look at price, but will also attempt to ascertain quality; in some countries they will be aided by government-established quality categories and certifications. Breakdowns in the trust consumers have in their butchers or in government certification, say because of rumors regarding meat tainted by poisonous substances, could produce a kind of 'run' on butcher shops, and widespread failures among them. The comparison is surely, being forced. Few butcher shops will deliver meat of standard quality out of town to third parties on instruction from a customer. Few externalities could be expected from a 'run' on a butcher shop under suspicion. Furthermore, a butcher will seldom turn down a customer who wants to buy with cash everything in sight at the price announced by the butcher (he will just make sure the cash is not counterfeit); a banker will surely not lend all a customer wants to borrow at the going interest rate. The former is a spot transaction; the latter involves a promise to repay in the future which may or may not be sincere or wholly credible. Enforcing the loan contract or liquidating collateral property will involve costs, and even with speedy enforcement the bank may be unable to get all of its money back. The bank will incur costs to explore the credit-worthiness of borrowers; the butcher will not care much for the reputation of cash-carrying customers. There is no humanly possible way of devising a fail-proof system of finding out the true intentions of borrowers, so lenders are likely to end up rationing credit, i.e., putting a ceiling on what arms-length customers can borrow, regardless of their willingness to pay higher interest rates see Stiglitz and Weiss (1981). Banks share with other financial intermediaries the problem of seeking reliable borrowers, but having liabilities that are payable on demand at par puts a premium on safe and fairly liquid loans for banks. Banks supply a service that is a very close substitute to the bulls and .C Diaz-Alejandro, Consequences of financial noitazilarebil 3 coins supplied by governments; such pretension imposes special limits on banks' portfolios. Financial intermediaries, including banks, rely on borrowed funds; owners of those institutions typically invest their own capital only in amounts which are a small fraction of their total lending. A depositor in a bank will have preoccupations similar to those a lender has when eyeing a borrower: will I be able to get not only the promised yield but also my money back or delivered to a third party as specified in the legal contract? In a totally unregulated system the rational depositor will attempt to process a great deal of information regarding financial institutions, balancing expected returns versus risks of not being able to dispose of his funds as promised. One can enigami a world in which all financial intermediation, including the payments system, is private and without any direct government regulation. The information imperfections may be partly offset by investments in information gathering by lenders and depositors; with the passing of time reputations would be built up. Balance sheets of financial intermediatries would be widely available and closely analyzed. It is even conceivable that possible economies of scale involved in the processing of information may not be so large as to impede large numbers of financial agents, assuring some competition among them. Unexpected shocks and informational failures in this economy could of course lead to bankrupticies of some financial intermediaries; an informed public presumably would know the source of such events, so there would be no ripple effects contaminating sounder institutions. One would expect to find a rich risk-return menu available to savers in this world, which would also generate an unregulated stock market for those preferring equity arrangements for obtaining or investing funds. Concentration of economic power, in the form of 'economic groups' and conglomerates, would presumably exist only if there are economies of scale to yield social benefits, with free entry checking monopoly behaviour. In this world, the market rate of interest would hover around the natural rate. Note that even a purely laissez-faire financial system must count with some indirect government inputs in the form of an efficient and fair judiciary and police system to punish fraud and to handle the enforcement of contracts and the settlement of disputes and bankruptcy cases. And behind that one must presumably have a polity generating such a judicial and police system, and a body of contract, antitrust and bankruptcy laws which are regarded as fair, efficient and enforceable. Financial agents feeling a temptation to cheat on their contracts or give false information must know that such behavior will have very high costs; they must also know that even if they flee their country the long arm of the law is likely to reach them abroad. Governments in this world must be expected neither to change nor to corrupt these rules; domestic private citizens must not only believe in the stability of this system, but also believe that other citizens also have that 4 .C ,ordnajelA-zaiD Consequences of financial noitazilarebil belief. Foreign financial agents, when dealing with those of our country, must not expect to have these rules changed when things go wrong for them; foreign financial agents dealing within our economy would presumably be subject to the same rules of the game as our financial agents. Even among those who find a predominatly laissez-faire financial system not only imaginable but also a desirable goal of policy, there are differences on how to organize the underlying monetary system. Some take for granted that the government will have a monopoly in the supply of cash and coins. Aware of the macroeconomic instability which could be generated by sudden changes in the public preferences between cash and the demand deposits supplied by banks subject to fractional reserves, as during the early 1930s in the United States, these observers would impose I00 percent reserves on banks supplying those deposits, while leaving the rest of the financial system unregulated, See Friedman (1959). In this view, a significant difference exists between money (cash plus demand deposits) and other financial assets. Other observers would carry a laissez-faire policy even into cash, eliminating the government monopoly, and leaving individual agents free to choose among all potential suppliers of cash, including foreign governments. In this approach there would be no need to impose reserve requirements on banks, which would be totally free of government regulations. The money supply would presumably adjust to real needs, while price and wage flexibility would maintain macroeconomic balance see Hall (1982). The laissez-faire vision also has (at least) two variants regarding how the unregulated domestic economy should manage its exchange rate. If the country is large enough, those giving the government a monopoly over the money supply advocate a freely fluctuating exchange rate, with no restriction on international trade or on capital flows. If the country is small (with the borderline between large and small left fuzzy), even money monopolists favor some form of fixed exchange rate system. For politically turbulent nations, a variant of the liassez--faire vision in a small country would do away with the local central bank, i.e., no national could be trusted with the money monopoly, which would be turned over to a presumably benign and responsible foreign government, as in the eases of Panama and Puerto Rico. No industrial country has come close to the laissez-faire vision, at least since the 1930s. The government monopoly over cash has been maintained, while banks have been regulated and subjected to fractional reserve requirements. In the United States, an explicit federal deposit insurance for accounts below a certain size was introduced during the 1930s and maintained since then, eliminating old-fashioned 'runs' on banks. Like any other insurance scheme, deposit insurance is vulnerable to moral hazard consequences, i.e., it induces depositors to think that 'one bank is as good as another', and leads bank managers to undertake riskier loans. To avoid such insurance-induced risk-taking, supervision over bank portfolios has C. Diaz-Alejandro, Consequences of financial liberalization 5 accompanied deposit insurance. Indeed, the Federal Reserve holds impressive discretionary powers regarding the lending policies of commercial banks, and over their liquidation or merger, if banks are found to be in trouble. Those discretionary powers, however, are subject to review by the Treasury, Congress and the courts. Vigilance has been exerted by the Fedral Reserve to keep non-banking companies from owing or controlling banks. While a trend towards bank deregulation has occured in the United States during the 1970s and early 1980s, removing some archaic populist controls, the view the 'banks are special' has been reaffirmed. The rescue of the Continental Illinois during 1984 in the face of a new type of run by uninsured international depositors, a rescue which extended de-facto insurance to all depositors, showed that large banks are quite special. This view is based on the externalities generated by the fiqnidity, mobility and acceptability of bank-issued transaction accounts, characteristics which take special significance in periods of financial stress see Federal Reserve Bank of Minneapolis (1982). Only banks create and incur liabilities that are payable on demand at par and that are readily transferable to third parties. Your safe payments system is also my safe payments system; it is partly a public good, not only serving individuals, but also providing backup liquidity on which the whole of the financial system rests. The Federal Reserve Bank of Minneapolis concludes: '... the case for segregating essential banking functions into an identifiable class of institutions is every bit as powerful today as it was in the 1930s. If anything, concerns regarding financial concentration, conflicts of interest, and the fiduciary responsibilities associated with lending depositors' money may be more relevant today than they were 50 years ago' (p. 13). Other industrial countries, such as Japan and those in Western Europe, are even farther away from the laissez-faire vision for the financial system than the United States. In several of those countries interest rates are also controlled or supervised by monetary authorities, and international capital flows are regulated. None relies on just price and wage flexibility to seek macroeconomic balance. (Note that those advocating a fixed-money-growth rule are typically as skeptical of the speedy workings of price and wage flexibility to maintain macroeconomic balance as Keynesian economists.) Finally, it is difficult to find small and open countries which have given pu monetary sovereignty after tasting it. Politically sovereign countries using other countries' moneys, or 'permanently' committed to a fixed exchange rate and unrestricted capital mobility with a monetary 'big brother', usually have come to those arrangements as part of a transition from a colonial status, and it is moot whether the arrangements are regarded as really permanent or 6 .C Diaz-Alejandro, Consequences of financial liberalization only as a step in monetary learning by doing, a step which had to be accepted to placate special internal and external interests rather than taken after a thoughtful consideration of national welfare. During October 1983 there were reports that Israel was about to abandon its monetary sovereignty; the Minister responsible for the proposal resigned shortly afterwards. 3. Notes on the financial history of Latin American and Southern Cone experiments While the financial history of Latin America remains to be written it appears that by 1920s most countries had succeeded in establishing commercial banks of the (then) traditional sort; several countries carried out banking reforms during those years following the advice of Professor E.W. Kemmerer, of Princeton University, and of visitors from the Bank of England. The banking system of South American countries already included institutions owned by national and provincial governments; Argentina, for example, had an important government-owned mortgage bank and several other public banks. The late 1920s were characterized in most Latin American countries by fixed exchange rates, convertibility and price stability; domestic interest rates were closely linked to those in New York and London. Although there was no 'financial repression', critics pointed to a lack of medium- and long-term credit, particularly to finance industry and non-export agriculture. Within agriculture, those without real estate collateral also complained about non-availability of credit. Domestic stock and bond markets were small; only Argentina seems to have had a promising formal domestic financial market, dominated by mortgage paper. The 1930s brought exchange controls and the expansion of government financial institutions, which at the height of the crisis proved their usefulness in decreasing the incidence of panics and runs; the massive bank bankruptcies which occured in the United States during the early 1930s were not witnessed in the large Latin American countries, apparently thanks to the presence of state banks plus an activist policy of rescuing most private banks in trouble. By the 1940s many countries had development banks granting medium- and long-term credits to non-traditional agriculture, industry and construction. Those credits, at least during the 1930s and early 1940s, seemed to have been granted at interest rates still ahead of domestic inflation, or at least not too far behind it, and in most cases contributed to an upsurge in capital formation. Public development banks remained, often through the 1950s, relatively small and efficient, as in the case of the Brazilian BNDE. The drying-up of external sources of finance during the 1930s and 1940s encouraged these efforts to mobilize local savings. C. Diaz-Alejandro, Consequences of financial liberalization 7 By the 1950s, however, it was clear that in South American countries experiencing inflation the development banks created to solve one form of perceived market failure (lack of long-term credit for socially profitable non- traditional activities) had led to another, i.e., a segmented domestic financial market in which some obtained (rationed) credits at very negative real interest rates, while non-favored borrowers had to obtain funds in expensive and unstable informal credit markets. Public controls over the banking system typically led to negative real interest rates for depositors. 'Financial repression' became an obstacle to domestic savings and their efficient allocation, and financial intermediation languished. In inflation-prone countries, financial reforms were introduced during the 1960s in the form of indexing of some loans and deposits; those involving the housing market were a particularly popular field for these new policies. Post- 1964 Brazil is the clearest example of a sustained effort to revive the domestic financial system and domestic savings using a number of indexing devices, but maintaining close government supervision of financial institutions and of interests rates charged in formal markets. The results of the Brazilian reforms have been mixed: domestic financial savings have been encouraged relative to the pre-1904 situation, in spite of continuing inflation, and the new policies supported impressive rates of capital formation. But attempts to encourage a significant stock market have failed, and the financial market remains heavily dominated by public securities. Private agents have shown reluctance to offer indexed securities. Credit to some sectors, such as agriculture, has been heavily subsidized for long periods of time. Brazil has also retained controls over the links between domestic and international financial markets, while following a passive crawling peg exchange rate policy, with sporadic jumps and other innovations. In Central American countries, with a tradition of price stability and conservative macroeconomic management (before the 1970s), bank concentration and an association of a few private banks with dominant economic groups, historically raised concerns about monopoly power. Galbis (1979, p. 349) noted: 'The experience of El Salvador suggests that interest rate freedom cannot be expected to be a panacea and automatically produce interest rate equilibrium in countries that have a relatively concentrated and unsophisticated financial system, which is the typical case in small, less developed countries. Positive policy actions to avoid market distortions might be required.' This type of concern led to the nationalization of the banking system in Costa Rica in the late 1940s and in El Salvador in 1979. Government regulation, including selective credit allocations, over such oligopolistic structures may make matters worse, consolidating access by a few favored 8 .C ,ordnajeiA-zaiD secneuqesnoC of financial noitazilarebil businesses to subsidized credit; post-Second-World-War Greece is said to be an example of inefficient and inequitable regulation of concentrated banking sector. Southern Cone countries, coming out of sundry populist experiences around the mid-1970s, undertook financial reforms going beyond those of Brazil in a laissez-faire direction. Post-1973 Chile provides the clearest example of this type of financial liberation for a careful narrative, see Arellano (1983a,b). That experiment started with a fully nationalized banking sector; a first task was to return most banks to the private sector. This was done by auctioning them off, with generous credit arrangements, or by returning them to previous owners; apparently little effort was spent on investigating the bankiug credentials of new entrants. At an early stage interest rates were freed and 'financieras' were allowed to operate with practically no restrictions or supervision; early bankruptcies in December 1976 and January 1977 of the more adventurous and unregulated 'financieras' led to the establishment of minimum capital requirements for entry. Authorities repeatedly warned the public that deposits were not guaranteed, beyond very small deposits with banks, and that financial intermediaries, like any other private firm, could go bankrupt; it was explicitly stated that there would not be a 'bailing out' of banks and other financial intermediaries. Since 1974 multi-purpose banking was allowed, on the ground that the Chilean market is too small to sustain specialized financial institutions of efficient size. Reserve requirements were steadily reduced, reaching less than ten percent of deposits by 1980. During 1977 it became apparent that an important bank (the Banco Osorno) was in serious trouble. The authorities, fearing that its bankruptcy would tarnish external and internal confidence in Chilean financial institutions, intervened, and rescued all depositors and the institution. Apparently, the fear that external loans would decrease if the Osorno had been allowed to go bankrupt was the crucial argument for intervention. Naturally, frech warnings were issued that, from then on, financial intermediaries would not be rescued. At that stage practically no inspection or supervision of bank portfolios existed; only in 1981 were sigmificant regulatory powers given to the Superintendency of Banks. One may conjecture that after this event most depositors felt, de facto, fully insured and foreign lenders felt that their loans to the private Chilean sector were, in fact, guaranteed by the State. After 1977, banks, rather than 'financieras', became predominant in the financial system. During 1979 the Chilean economic authorities started a process expected to culminate in a pseudo-exchange-rate union Csee Corden (1972) with the United States. The nominal exhange rate between the peso and the U.S. dollar was fixed in July 1979, and restrictions over convertibility and capital movements were relaxed; by 1981 those restrictions had been considerably .C ,ordnaeiA-zaiD Consequences of financial noitazilarebil 9 weakened, and Chile witnessed a massive capital inflow. Presumably the hope was to make lending to Chile subject to no more currency risk than lending to Puerto Rico or Panama; the nominal exchange rate was supposed to last 'for many years'. Some of the economic authorities dreamed of doing away with the national currency altogether, but feared that the military might not wish to go that far. The theoretical underpinnings of these policies included a special version of the monetary approach to the balance of payments, plus the hypothesis that financial markets, domestic and international, were no different from the market for apples and meat. Voluntary financial transactions between private agents were their own business, and presumably Pareto-optimal. Indeed, the nationality of those private agents was regarded most irrelevant. The then Director of the Western Hemishere Division of the International Monetary Fund put it this way at meeting held in Santiago de Chile during January 1980 see Robichek (1981, p. 171): 'In the case of the private sector, I would argue that the difference between domestic and foreign debt is not significant -- barring govern- mental interference with the transfer of service payments or other clearly inappropriate pubfic policies -- if it exists at all. The exchange risks associated with foreign borrowing are presumably taken into account as are the other risks associated with borrowing, whether it be from domestic or foreign sources. More generally, private firms can be expected to be careful in assessing the net return to be derived from borrowing funds as compared with the net cost since their survival as enterprises is at stake.' The same author went on to argue that overborrowing by the private sector, even with official guarantees, was very unlikely, provided official guarantees were given on a selective basis; only public borrowing on international financial markets was regarded as posing more serious debt service risks see Robichek (1981, p. 172). Convergence of domestic inflation and interest rates toward international ones proved to be a slow process, during which the fixed 'permanent' nominal exchange rate yielded great incentives for private capital inflows into Chile: during 1981 the current account deficit reached an astonishing 14 percent of Chilean Gross National Product, with international reserves holding their own, while domestic savings appeared to collapse. The process of financial liberation had also led to a widely noted (by opposition economists) concentration of potential economic power in the hands of a few conglomerates or economic groups, which combined financial and non- financial corporations. Before 1981, the official view seems to have been that those economic groups must reflect some economies of scale, and could be regarded as one special type of butcher shop, disciplined by free entry and 01 .C ,ordnajelA-zaiD Consequences of financial noitazilarebil other competitive pressures. Their allocation of credit resources, often heavily loaded in favor of companies associated with the group, was presumed to be more efficient than that which government bureaucrats could achieve. As late as March 1981 international business publications were writing that 'Chile's free-enterprise banking environment' was proving to be a powerful magnet for foreign banks, and that more entrants into the thriving sector were lining up. (See Business Latin America, March ,11 1981, p. 79.) By June 1981 the same publications were noting with concern the cessation of payments on local credits by CRAV, a Chilean sugar company, as well as other blemishes on the economic miracle, but argued that 'the problem areas pose no immediate threat to growth' (Business Latin America, June 3, 1981, p. 173). Following the CRAV news, the Central Bank supported financial institutions to stem incipient 'runs'. By November 1981 the position of two important private Chilean banks and several 'financieras' became critical: they were 'intervened' by the Central Bank. Further interventions of financial intermediaries occurred during the first half of 1982; rather than harsh bankruptcy proceedings, these actions apparently involved a generous expansion of credit to the private sector. Between the end of December of 1981 and the end of June 1982, domestic credit in Chilean pesos expanded by 41 percent; of the net increase in domestic credit, 92 percent went to the private sector see International Monetary Fund (1983, pp. 118-119, lines 32 and 32d). The massive use of Central Bank credit to 'bail out' private agent raises doubts about the validity of pre-1982 analyses of the fiscal position and debt of the Chilean public sector. The recorded public sector budget deficit was nonexistent or miniscule for several years through 1981, and moderate during 1982. For evidence on the apparent Chilean fiscal performance, see McKinnon (1982). The declining importance of ostensible public debt in the national balance sheet was celebrated by some observers; indeed it was argued that public sector assets, such as remaining public corporations, exceeded its liabilities. Ex post it turned out that the public sector, including the Central Bank, had been accumulating an explosive amount of contingent liabilities to both foreign and domestic agents, who held deposits in, or made loans to the rickety domestic financial sector. This hidden public debt could be turned into cash as the financial system threatened to collapse. Eminent students of fiscal and financial systems, who were involyed in the Chilean reforms, apparently overlooked this potential debt bomb. (Present at the creation and early development of the new Chilean financial system were experts brought together by the Organization of American States Program for the Development of Capital Markets; together with the Chilean Central Bank, this program sponsored seminars on capital markets in Santiago during 1974, 1976 and 1977.) By late 1981 and early 1982 Chile was also feeling the full force of the