Multipurpose Banking: Its Nature, Scope, and Relevance for Less Developed Countries

International Monetary Fund. Research Dept.
Published Date:
January 1980
  • ShareShare
Show Summary Details

Discussion of financial systems often raises the question whether there is an optimum path for financial development in a country’s progress toward maximum economic growth. Financial development is a generic term that embraces structural, functional, instrumental, and process-related aspects of the financial system. Structural aspects are denoted by the form of organization of financial institutions constituting the system, such as the nature of the money and capital markets, the relationship of institutions with industry and trade, and the machinery designed to regulate the institutions (see Goldsmith (1969)). Functional aspects are the modes of operation of the institutions, such as the methods of financial resource generation and of granting various types of term credit (Madan (1964 a)). The instrumental aspects relate to the type of instruments used in transactions involving treasury bills, acceptance paper, bills of exchange, and substitutes for bank deposits (Bloomfield (1956) and Khatkhate and Villanueva (1979)). Finally, the process-related aspects refer to how the financial system has evolved to its present stage, including governments’ influence thereon (Patrick (1966)). Since financial systems have historically evolved under the impetus of economic, social, and institutional forces that are by no means uniform across countries, it would be no more than a coincidence if the various aspects of financial development referred to previously were to reveal common characteristics in all those countries in which well-organized financial systems have evolved.

Certain financial instruments, such as money and similar forms of debt, improve the efficiency of economic transactions, while other financial instruments affect primarily the efficiency with which resources are reallocated from savers to investors. Financial institutions of various types emerged in modern economies by developing different types of financial instruments. They have enhanced the productivity of the economy by bringing about a division of labor between the financial sector on the one hand and the enterprise and household sectors on the other. There is further specialization among the different segments of the financial system, the degree of which has varied from country to country. At one extreme are countries, such as the United Kingdom, the United States, and Canada, where demarcation of functions between financial institutions is sharp, while at the other extreme there are countries, such as the Federal Republic of Germany, Austria, Switzerland, and Belgium, whose primary financial institutions are “multipurpose” or universal.

The purpose of this paper is to delineate the role, nature, and scope of multipurpose banks and to point our their relevance for less developed countries that are undertaking deliberate programs to develop their financial systems. The rest of the paper is organized as follows. Section I discusses how universal or multipurpose banking has historically evolved and the nature of such banks. By surveying the financial systems in some leading industrialized countries, it is shown that multipurpose banking has become almost universally adopted in industrial countries, as even those countries in which financial institutions originally were specialized have begun to move in the direction of multi-purpose banking. In Section II, various criteria for evaluating multipurpose banking are stated. In Section III, the implications of multipurpose banking for less developed countries (LDCs) are analyzed, taking into account their preoccupation with economic development and the accompanying institutional change needed to promote it. Some broad conclusions are presented in Section IV.

I. Multipurpose Banking—Its Evolution and Salient Features

It is worth describing, at the outset, the salient features of multipurpose banking and in what way it is distinguished from commercial banking, as the more limited form of banking is generally known.

One of the hallmarks of multipurpose banking is the close relationship between the bank and the borrowing firm, which is maintained primarily through current accounts. The current account arrangement is similar to the overdraft or cash credit account of U.K. commercial banks, under which credit limits are sanctioned on the basis of certain securities, and borrowers then use these limits when needed or credit to their accounts any receipts that may accrue to them. The current account arrangement of German universal banks is different, in that it binds the borrower to the bank more closely and for a longer duration. The average borrower depends, to a very large extent, on these current accounts, not only to obtain short-term credit but also to obtain credit for financing fixed capital formation, which is repaid subsequently by issuing securities (Whale (1930)).

Opening an all-purpose current account only marks the beginning of a bank’s almost total involvement in the affairs of the borrowing company. A common method of promoting firms in the Federal Republic of Germany is for a bank to finance, as a promotional effort, the entire capital requirements of an entrepreneur from the beginning. When entrepreneurship has to be encouraged, this comprehensive form of promotional activity is necessary. Even after the entrepreneurship ceases to be a constraint, banks continue to adhere to this practice, first because their deep insights and close surveillance into the business network of customers lessen the dangers of business failure, and second because the close partnership that develops with industry makes diversification of assets possible, thereby attenuating risk. However, promotional activity by banks does not stop at floating the initial capital and selling stock to the public. Once the profit opportunities are discerned and contact with the industry becomes closer, other areas of opportunities open up which make it advantageous for banks to continue their ownership of stock in companies.

While banks’ holdings of their client companies’ securities may sometimes be large, banks do not necessarily hold large amounts of stock by design. Their function in regard to the bulk of the stock is essentially that of a middleman who tries to diversify and to match liquidity with profitability of invested funds. More often than not, the intention is to hold a major part of stock for a minimum period of time until the stock has appreciated and is able to be disgorged on the stock market. However, banks sometimes end up involuntarily holding some stock on their own account because of temporary or prolonged difficulties in the market. Such involuntary holdings by banks may perhaps indicate the extent of risk involved in universal banking for banks themselves as well as for the economy (Immenga (1975) and Mülhaupt (1976)).

Such a close alliance between banks and industry has in the countries concerned more often than not helped to accelerate the process of industrialization. The universal banking not only extended the range of banking functions beyond that of the British model, but also transformed the organizational structure of both banks and industry. 1 The present association in the Federal Republic of Germany of universal banks with industry does not end with the provision of credit or underwriting of capital issue; it extends all the way to the organization and management of industry. Banks, both directly through ownership of equity as well as the granting of loans, and indirectly through safekeeping of shares on depositors’ account, have stakes in the growth and prosperity of industry, so that they place their representatives on the boards of client companies. This often introduces financial expertise into nonfinancial firms and thus can reduce the friction that may arise between the borrower and the lender. Along with provision of credit and the discharge of some managerial functions, universal banks also assume the role of technical consultants and marketing advisors; in short, they are the main catalyst of the process of industrialization.

Emergence of multipurpose banking has been a financial manifestation of the kind of innovative response of economic agents to the challenges they faced. The form of banking that became prevalent after the industrial revolution in the United Kingdom of the eighteenth and early nineteenth centuries was “commercial,” with the sole accent on short-term credit to trade and industry, the underlying theory being that credit should be self-liquidating. The implications of this were that the banks refrained from explicitly financing fixed capital formation because it would lock up banks’ funds, contrary to the tenets of “sound” commercial banking as reflected in the “golden rule of banking” and also because it would reduce borrowers’ interest in their enterprises’ prosperity. However, this form of banking also reflected the particular nature of British economic development. In the United Kingdom, the process of industrialization had “proceeded without any substantial utilization of banking for long-term investment purposes. The more gradual character of the industrialization process and the more considerable accumulation of capital, first from the earnings of trade and modernized agriculture and later from industry itself, obviated the pressure for developing any special institutional devices for provision of long-term capital to industry” (Gerschenkron (1962, p. 14)). Though a major factor, this was only one among many that impelled the banks to emphasize short-term lending. Industrialization in the United Kingdom was an outgrowth of commercialization that involved high turnover of capital. Naturally, therefore, the commercial culture dominated the functions of banks, which emerged to meet the needs of expanding industry. Quite apart from this, the motivating force behind industrialization was the urge of those who had surplus funds accumulated to seek out new opportunities to make more profit. This meant that the spirit of adventure, together with the wherewithal to sustain it, was abundant, so that no support from outside was called for. In other words, industrial and entrepreneurial development took place without being crucially dependent on the existing financial institutions.

However, some change came about in the United Kingdom toward the end of the nineteenth century, when large industrial concerns evolved that required mobilization of long-term funds. This need was met from the bond market that emerged in the United Kingdom, initially to finance government debt but later also to finance private sector debt. Thus, the demand by borrowers for long-term finance from banks remained limited, and their role was secondary (Pollard (1964); Deane (1961); and Khanna (1978)).

The experience of now developed countries, mainly in Western Europe, shows that those who followed in the footsteps of the pioneers have not necessarily imitated them. The newcomers on the scene encountered fresh challenges which, in turn, triggered new responses, giving rise to new techniques and institutions. France, Germany, and other Western European countries were “backward” compared with the United Kingdom, the degree of their backwardness generally varying directly with their distance from the United Kingdom (Gerschenkron (1962) and (1968)).

These differences in the development of Germany and other European countries from the U.K. pattern explain the differences in both the financial instruments and the scope of banking institutions that emerged in those countries. Thus, the role of banks in Germany was a reflection of the different needs of its economy, among which promotion of entrepreneurial activities and mobilization of savings were the most preponderant. There was an urgency to augment physical capital through various devices. Credit creation by banks was perceived as the most powerful instrument of mobilizing financial resources for industrialization. Bank credit enabled borrowers to acquire command over real resources, thereby linking up the savings and the entrepreneurship, so that credit was considered to be “essentially a phenomenon of development” (Schumpeter (1936)). Consequently, the distinction between short-term and long-term credit is unimportant, insofar as its significance for economic development is concerned. What matters is the total bank credit as such, which thus provides the raison d’etre for universal or multipurpose banking. However, the concept of universal banking has been more pervasive, in the sense that it radicalized not only the usual functions of banks but also their organization and mode of operation. Indeed, it changed the very ethos of banking. As Gerschenkron (1962, p. 14) has put it: “A German bank … accompanied an industrial enterprise from the cradle to the grave, from extablishment to liquidation, throughout all the vicissitudes of its existence” (see also Pollard (1964)).

In view of the pioneering role of Germany in setting up a universal banking system, it may be useful to enlarge on the factors that had a decisive influence on the shape of the German financial system. These factors fall into three distinct but interrelated categories: (1) historical factors; (2) regulatory and/or economic policy influences; and (3) behavioral responses.

There are two features of the German banking system that are striking—viz., its universality and its fairly low degree of concentration. Both are to a large extent the result of historical factors.

A single, unified German nation was not founded until 1871. Before that year, Germany consisted of a large number of mostly small, independent states. Many of these states issued their own currency through licensed private banks. 2 Since the use of the various currencies was generally restricted to their respective states of issue and since there had been a myriad of tariffs across them, overall size and importance of these banks was usually as limited as the size of the states in which they operated. Only banks in larger states like Prussia and Bavaria, and some at traditional trade centers, operated outside their respective states. Apart from the private banks, which were mostly owned by individuals or families, there were a number of official “state banks,” such as the Royal Bank of Prussia, which was later to become the first German central bank. Private banks generally concentrated their activities on the economic upper and middle classes of traders, artisans, and the more important farmers. Smaller savers and investors—in particular, those in small municipalities—were generally neglected by the private commercial banks. This situation changed in the second half of the eighteenth century, when in several states savings banks were founded and administered by the municipalities. The success of these institutions prompted other municipalities and states to follow suit. 3 Orders for the foundation of the first savings banks often came from enlightened sovereigns who were concerned with the security of the savings of people of modest means. The funds so generated were mostly used locally. Private mortgage finance and absorption of bond issues of the municipalities as well as short-term and medium-term loans to smaller artisans, merchants, and traders figured prominently in the loan portfolios of these savings banks.

The beginning of the industrial revolution and the accompanying strong impulse toward economic integration led to an enormous increase in both the total amount and the average size of credit demand. A great number of new banks were founded, initially as independent, privately owned institutions; soon, however, many banks merged to form bigger units—organized mostly as joint stock companies—in order to be able to meet the large credit demands of big industrial ventures that were being undertaken at the time. As monetization of the country progressed, yet another type of financial institution, the rural or commercial cooperative, appeared on the scene. The main purpose of these institutions was initially to supply their members with the short-term credit needed to cope with seasonalities in net revenue. Funds were raised mostly via self-imposed savings and deposit schemes. However, in response to emerging demand, the business of the cooperatives was soon extended to medium-term and long-term investment finance and various other financial services.

The strong expansion of economic activity and the rapid growth of cities in the second half of the nineteenth century also led to the growth of a number of specialized financial institutions. Most important of these new institutions were the private mortgage banks, which were generally organized as joint stock companies and operated along the lines of the French Crèdit Foncier. Their foremost source of funds was the issuance of bank debentures.

The foregoing discussion focused on the historical reasons for the emergence of the subsystems of financial institutions presently existing in the Federal Republic of Germany. Though the institutions of the individual subsystems each catered to a different clientele, relied on different sources of funds, and placed different emphasis on various lending activities, the simultaneous growth of all insitutions prevented individual institutions, or even subsystems, from dominating the financial sector. The competitiveness of individual savings banks and cooperative banks was achieved through a well-balanced division of labor between them and their central institutions. The broad networks of savings banks and cooperatives mobilized sufficient funds and established the basic business relationships for the two subsystems. The central institutions could draw on both. On the other hand, the central institutions supplied various types of financial services to their respective base institutions. This institutionalized demarcation of functions culminated in a competitiveness of both individual institutions and subsystems that could not have been achieved otherwise.

Apart from this, competition between financial institutions was fostered by the freedom that the institutions in the three subsystems had as to the range of financial services they could supply. If not always in practice, they were potentially universal banks from the very beginning. Restraints on the coverage of banking business were exercised voluntarily—mostly reflecting a lack of demand—rather than enforced by regulations. As such, the system as a whole was clearly demand-oriented.

With few exceptions, there were no specific laws or regulations for the financial sector until the beginning of this century. 4 This situation changed during and after the depression of 1929-31, when all banking institutions were subjected to a unified General Banking Act. However, unlike the United States, where the collapse of hundreds of banks during that period led to a rigid legislated separation between important banking functions and thus brought about a fairly high degree of specialization of financial institutions, Germany did not interfere with the organizational structure and functional orientation of financial institutions, but concentrated instead on prudential regulations covering in particular capital adequacy and liquidity. Though it has been amended several times, this Banking Act has remained unchanged over the years in its major features; in particular, it did not interfere with the universal nature of banks.

Entry into the banking business is relatively easy in the Federal Republic of Germany. Bank licenses may be refused only on grounds of lack of adequate own resources, or untrustworthiness or lack of professional qualifications on the part of the proposed managers, partners, and proprietors. The burden of proof rests with the supervisory authority. The opening of branches has to be reported but is not subject to any legal restriction nor to proof of an existing “economic need.” While rather liberal in its handling of the licensing power, the Supervisory Office has been fairly strict in the application of its powers in the area of capital adequacy and liquidity in order to prevent abuses that might endanger the security of assets entrusted with banking institutions.

Thus, it is fair to say that, on the whole, regulations under the Banking Act are defined in a rather general way and have been applied liberally by the authorities. As a result, the banking sector in Germany enjoys a freedom of action that probably is not matched by that of banks in many other countries.

Apart from legal constraints, economic policy measures, whether monetary or fiscal, tend to exert an important influence on the scope, structure, and efficiency of a financial system. One of the most prominent types of such policy measures is the setting of deposit and lending rates. In the Federal Republic of Germany, this instrument has been used rather carefully. Until recently, maximum (but never minimum) lending rates were linked by a complex formula to the discount rate of the Deutsche Bundesbank. In practice, this arrangement never served as a constraint, since keen competition among financial institutions operating in the various financial markets always resulted in rates substantially below the prescribed maximum. Some link, but by no means a stringent one, existed also between deposit rates and the discount rate. After each change in the discount rate, representatives of the central bodies of the commercial banks, savings banks, and credit associations—the so-called Central Credit Board—would discuss possible changes in deposit rates and convey its proposals for adjustments to both the Supervisory Office and the Deutsche Bundesbank. More often than not, the authorities accepted the recommendations of the Central Credit Board and did not impose their own solutions. More recently, the decision on both deposit and lending rates has been left entirely to the discretion of the financial institutions; this development reflects, however, only a change in the attitude, rather than in the legal power, of the Supervisory Office.

The behavior of economic agents in any country is principally determined by past experiences and present constraints. In the Federal Republic of Germany, participants in the financial system have generally been subject to only a few constraints on their behavior. This implies that they could freely choose arrangements that, given the pattern of demand, suited their interests best. As a result, the financial system is highly demand-oriented and reacts very flexibly to structural and cyclical shifts in the economic setting as well as to changes in preferences. The high degree of flexibility is, however, the result not only of the virtual absence of constraints but also of the sometimes venturesome and often innovative behavior of financial institutions in the Federal Republic of Germany. It is possible to cite, in this connection, the early adoption of the “sediment theory of banking,” which contrasted with the “golden rule of banking” or “real bills doctrine.” This latter rule, which has exercised an important influence on the philosophy of “sound” banking and on the resulting behavior of banks in Anglo-Saxon countries, implies that short-term deposits should be used to finance only short-term lending, and that long-term lending should be financed only by long-term deposits and other long-term sources of funds. The “sediment theory,” on the other hand, is based on the assumption that a certain core of funds always remains with banks over longer periods of time. This latter theory tends to be confirmed in particular in fast-growing economies. The “golden rule of banking,” which is probably the ideological cornerstone of specialized banking, never gained as much importance in Germany as it did in most other countries. This fact had obvious consequences for the range of activities banks were prepared to engage in and, as such, fostered universal banking in Germany.

Universality was also approached from the opposite direction. While, as described above, banks with mostly short-term to medium-term resources ventured into long-term lending, other banks that had initially specialized in such participation transactions as the foundation of industrial enterprises and issuing activities—much along the lines of the Sociètè Gènèrale du Crédit Mobilier in France—also soon turned toward offering financial services, such as accepting deposit accounts and overdraft facilities, discounting bills, and making transfer payments for their customers. This behavior, which at the time was unconventional, was not only the reflection of banks’ response to customer demand but also the reflection of banks’ attempts to reduce their own risk through greater diversification and by obtaining a regular and inexpensive flow of information about their customers’ overall financial situation. The resulting intimate relationship between banks and customers and the banks’ interest in maintaining them often led banks to move into financial activities that had been covered only by specialized institutions. Given the virtual absence in Germany of legislation restricting the range of financial services, financial institutions could fully respond to new opportunities, which promoted universality in the German financial system.

The ethos of universal banking has been spreading over the years to a number of other countries, both developed and developing. The so-called commercial banks in these countries have acquired some of the functions of other types of specialized financial institutions, notably investment banks. In the United Kingdom, commercial banks’ lending for other than short-term purposes began to grow after the Macmillan Committee located, in 1931, the “gaps” in the flow of credit to industry. These gaps were the lack of credit facilities for all sizes of firms and the absence of any credit facilities for small and medium-sized industry. British banks tried to fill these gaps by setting up subsidiaries designed to provide intermediate-term and long-term loans, in addition to their existing practice of lending on a “roll-over” basis under which ostensibly short-term credit can be utilized for long-term purposes. In the United States, banks also have widened the range of their operations and have exerted a more pervasive influence on industry through the medium of holding companies.

In almost all countries the major commercial banks have tended to become multipurpose banks, offering a wide range of services to corporate and individual customers. However, there are still some differences between countries with regard to the extent to which deposit-taking institutions have been permitted by the authorities to move in the direction of universal banking (Inter-Bank Research Organization (1978)).

II. Evaluation of Multipurpose Banking—Suggested Criteria

Although the distinction between multipurpose banking institutions and specialized ones is beginning to be blurred, the choice between the two types is a real one for many countries. Therefore, given such a choice, it is necessary to have criteria to evaluate multipurpose banking so as to help policymakers choose between alternative models of financial systems. There are several criteria for judging the appropriateness of a particular financial institution, but for the purpose of this analysis five have been singled out as especially important—economic efficiency, mobilization of savings, promotion of entrepreneurial skills, financial stability, and conflicts of interest.

economic efficiency

Economic efficiency is often predicated on economies of scale. However, since the product of financial institutions is intangible, measurement of their efficiency on the basis of output has generally bristled with difficulties (Smith (1965), Kaufman (1966), Bell and Murphy (1967) and (1969), Greenbaum (1967), Stillson (1974), Heggestad and Mingo (1976), Richard and Villanueva (1980)). This measurement problem is exacerbated by the financial institutions not producing one single, clearly defined type of service, but a whole range of interrelated services. Any empirical evaluation of the relative efficiency of multipurpose banking systems thus becomes difficult. Hence the discussion of efficiency in what follows will be mainly of a qualitative nature.

Economies of scale in financial activities arise from the indivisibility of financial assets and factors of production, from portfolio diversification and management, and the existence of sizable information and transaction costs. Economies of scale can originate from individual types of financial services as well as the overall size of an institution. To the extent that it is difficult to acquire some financial assets in small denominations, a minimum size of a financial institution’s total resources is necessary so that it can derive price and cost advantages without running the risk of maintaining an unbalanced portfolio. Economies of scale also result from the indivisibility of factors of production, which may necessitate large-sized financial institutions and thus, in small countries in particular, a certain degree of concentration in the financial sector.

Portfolio diversification and management are two more factors contributing to the economies of scale from amalgamating a broad range of financial services. It is often observed that the production of a number of different, but related, types of output can lower costs when production takes place under one roof. This is particularly true when a factor input, such as computers or trained manpower, may be used alternately for the production of any one of the related products, so that idle capacity, which is created as a result of variable demand for individual products, is minimized. In these circumstances, multipurpose financial institutions can, ceteris paribus, maintain a higher average level of capacity utilization over time, thereby reducing the cost of financial services.

Information and transaction costs are important elements in the total overhead costs of any financial institution. Information costs include costs involved in the acquisition of knowledge and expertise in financial activities and costs related to the evaluation of the “quality” of an individual investment as well as of the portfolio of assets as a whole. The evaluation of the appropriateness of any financial transaction generally rests on a thorough assessment of alternatives. Hence, a fair amount of information is necessary on determinants other than those directly affecting the quality of a particular choice of a financial asset. Similarly, the information needed in assessing the desirability of a loan to a particular firm may overlap considerably with the information needed to judge the desirability of providing other financial services to the same firm. The related cost advantage of an individual institution in a universal banking system translates itself into a more efficient use of existing scarce resources for the benefit of the economy as a whole.

Responsiveness to changing demand is another important criterion on which to judge the efficiency of a banking system. A low degree of responsiveness of financial institutions implies additional costs to both suppliers of financial resources and to the ultimate users of funds. This, in turn, reduces the overall level of financial savings and impedes their efficient allocation for investment purposes. In many countries, financial sector legislation imposes de jure constraints on the degree of responsiveness of financial institutions by narrowly defining the range of activities they are permitted to carry out. In consequence, the demand for financial services is made to adjust to the supply pattern. By contrast, universal banking systems tend to be demand-oriented.

It seems to follow that universal banking institutions would, on average, be large-sized, reaping important economies of scale. However, this need not lead to a higher degree of concentration than under a specialized financial system. This proposition is supported by Table 1, which gives concentration ratios for selected countries. Measured by the Herfindahl index 5 the two countries with acknowledged universal banking systems, viz., the Federal Republic of Germany and Switzerland, rank among the three lowest, Switzerland being followed closely by the United Kingdom, a country with a specialized system. On the other hand, the Netherlands and Sweden, two countries with a fairly extensive network of multipurpose banks, display by far the highest degree of concentration. The picture is equally inconclusive when an alternative measure, represented by the share of the three largest banks in total bank assets, is used. The Federal Republic of Germany displays the second lowest level of concentration, after Japan. On the other hand, the Netherlands and Sweden report again by far the highest degree of concentration, which may be attributed both to historical factors and to the fact that these two countries possess perhaps the smallest financial sectors of all the countries in the sample.

Table 1.Twelve Countries: Financial Sector Concentration Ratios, 1973
Herfindahl IndexThree-Bank Concentration Ratio 1Universality Rank
Germany, Fed. Rep.0.03990.25971
United Kingdom0.07620.450511
Source: Short (1979).

Share of three largest banks in total assets of deposit-taking institutions.

Source: Short (1979).

Share of three largest banks in total assets of deposit-taking institutions.

In order to see how concentration ratios were related to both market size and degree of universality, a regression analysis was carried out using these two variables as explanatory variables for the two concentration measures. The degree of universality is based on the ranking as given in Table 1. Universality was judged after verifying the range of activities permitted to financial institutions in the countries included in this exercise. The regression results as presented in Table 2 bring out clearly that concentration, as measured both by the Herfindahl index and the three-bank concentration ratio, tends to decline as the size of the market expands and the degree of universality increases. Notice that, owing to the ranking system applied in the regressions, the size of the universality variable declines with an increase in universality. The positive sign of this variable therefore implies that concentration rises as a financial system becomes less universal. The negative relationship between the degree of universality and concentration is contrary to the general belief that concentration increases under a universal banking system.

Table 2.Relation Between Concentration in Banking, Market Size, and Degree of Universality
Measure of Concentration 1ConstantMarket Size 1Universality RankR¯2SEE
Herfindahl 2-0.132

(-3.19) *


Three-bank 2-0.538




Denotes t-probability of 95 per cent.

Denotes t-probability of 99 per cent.

Source of data: Short (1979); market size in billions of U.S. dollars.

Data in parentheses give the t-statistics.

Denotes t-probability of 95 per cent.

Denotes t-probability of 99 per cent.

Source of data: Short (1979); market size in billions of U.S. dollars.

Data in parentheses give the t-statistics.

It should be noted in this context that concentration ratios are not strictly comparable between countries. At one extreme, two equal-sized banks operating in a universal banking system will be competitors in all areas of banking services demanded by customers. At the other extreme, in a specialized system, two banks charged with providing two distinct kinds of financial services will be monopolists in their respective markets. For this reason, a particular concentration ratio may imply, for countries of about equal size, greater concentration and more important effective market power in a specialized banking system than in a universal banking system. Since the regression uses figures on overall market size rather than effective market size, the results in Table 2 may well underestimate the effect of specialization on concentration.

mobilization of savings

Efficiency of financial institutions is not only judged in terms of technical efficiency but also by the success with which they help transform savings into financial form. There are three different, but not mutually exclusive, ways of achieving this objective. One, the fiscal approach, is to have the government use taxation to acquire savings, and then channel these into investment; the second way is to rely on financial institutions of one sort or another, which issue their own liabilities to the public, to transfer resources for investment. The third way is to arrange to transfer savings directly from excess saving sectors to borrowing sectors by having the latter issue medium-term and long-term instruments, such as bonds and shares, to savers. The third way usually requires both underwriting to facilitate the original issue and the emergence of a market for secondary trading to enhance the liquidity of these longer-term instruments to make them more attractive to savers. It is the last two ways that are important in the development of financial systems.

The question then is what type of financial system is able to bring about a rapid transformation of savings into financial form. Commercial banks generally attract short-term savings, and the mobilization of long-term funds is the task left either to the capital market or to the specialized institutions. However, the specialized agencies have generally relied on the government, which raises the funds through taxation and borrowing, for their resources. Hence, the shortage of investment capital is acutely felt in a large number of economies. To some extent, commercial banks resolve this problem through term transformation. However, this function of banks is often constrained by absence of institutional safeguards against the risk of illiquidity of bank loans used for financing fixed capital. As observed earlier, the multipurpose financial institutions have a distinct advantage, in that they can marshall both short-term and long-term capital under competitive conditions for financing industrialization. Since term capital is one of the most essential ingredients in the development process, how efficiently and effectively it is provided is one of the most important criteria for choosing among alternative financial systems.

promotion of entrepreneurship

Apart from the mobilization and efficient allocation of savings, the development of entrepreneurship is another criterion by which financial institutions should be evaluated. When commercial banks first came on the scene in the United Kingdom, they were not called upon to undertake this task because of the special conditions prevalent there. Finance followed, rather than preceded, entrepreneurship, which was spawned by the thriving trade, internal and external. However, in countries that industrialized after the United Kingdom, entrepreneurship was a major desideratum, and the financial institutions had to promote it actively, not only by purveying credit but also by supplying management and technical expertise. For this reason, it is necessary to evaluate how effectively the different types of financial institutions can activate dormant entrepreneurial skills in an economy, so that its financial resources are put to optimal use.

financial stability

Another consideration in the choice between alternative patterns of financial systems is whether multipurpose financial institutions are prone to a greater degree of financial instability than specialized ones. The stability of financial institutions depends largely on the risks involved in their operations, which, in turn, are primarily a function of the amount and quality of information, the degree of term transformation, and the degree of portfolio diversification. As mentioned earlier, a broadly gauged financial institution tends, in the course of its business, to accumulate a large and continuous flow of information at a very low cost. This is related to the question of how to match the marginal cost of the additional investment required to obtain information with the marginal revenue derived from safer, higher-yielding assets. On that reasoning, the economies of scale in information gathering and processing available to a multipurpose institution should not only be larger than those of a specialized institution but should also be accompanied by smaller risk of default.

Term transformation—that is, transformation of short-term liabilities of financial institutions into longer-term assets—constitutes a key element in multipurpose banking. Term transformation involves the risk that the institution may not have sufficient primary or secondary reserves if the owners of short-term liabilities unexpectedly withdraw their funds from a financial intermediary. Similar problems of illiquidity can arise if loans are not repaid on time. Risks involved in maturity transformation will generally increase with the extent of transformation of funds from short term to medium term or long term and as the percentage of short-term funds that is transformed into longer-term lending increases. However, considerable, though not unlimited, maturity transformation can be undertaken in an economy without any great risk. For one thing, a fair degree of term transformation can be undertaken by a financial institution without endangering its liquidity, because as long as total deposits maintain a moderately steady growth, there is always a core of deposits that remains with a financial intermediary and can be invested in any way it deems proper. 6 The degree of stability of these core deposits helps to determine what portion of resources available in the current period can be relied upon to remain available over future periods. 7 Besides, much of the deposit volatility experienced by individual financial institutions merely represents shifts—often temporary—between such institutions. One bank’s loss of deposits is typically the gain of another. Such movement of deposits has given impetus to interbank deposit markets, which are often used to offset the potentially disruptive effects of such shifts. In addition, financial institutions in most countries can have recourse to the rediscounting facilities of a central bank in the event of delay in the expected realization of cash from the sale of assets.

Quite apart from this, it has always been a prime development task of financial intermediaries to meet simultaneously the time preferences of savers and those of users of funds, in view of the usual shortage of long-term investable funds. If necessary term transformation “inside” a financial institution is not permitted either by law or current policy, it will perforce have to take place “outside”—that is, in financial markets or via direct transactions among financial institutions of different types (viz., commercial banks, investment banks), unless the development objective is to be totally abandoned. It is therefore not appropriate to assert that term transformation is a special characteristic of only multipurpose banking systems. In either case, widespread deposit withdrawal or arrears in repayments can create a liquidity crisis that can only be alleviated through intervention by the central bank to provide an infusion of liquidity into the system. Where banks can have access to short-term funds in domestic or international markets, this will relieve the pressure on their liquidity. Thus the risk of term transformation is seen to depend critically on the existence and efficiency of financial markets, which facilitate the flow of funds from financial institutions and nonbanks in surplus to those in deficit.

The relative size of securities markets is often interpreted as an indicator of the ease with which long-term financial instruments, such as bonds and shares, can be sold at short notice. Furthermore, it is sometimes claimed that a universal banking system undermines such markets, thus reducing the efficiency of financial intermediation. There is no conclusive evidence to prove or disprove this presumption, though it is possible to argue, on the basis of a casual empiricism, that there need not be an antithetical relationship between universal banking systems and the operations of security and bond markets. In fact, the existence and stability of these markets is considered by financial institutions as essential for the efficient functioning of a universal banking system as it is for the specialized commercial banking system of the U. K. or U.S. type.

Financial stability depends not only on the term transformation but also on the degree of portfolio diversification. In theory, total portfolio risk can be reduced by portfolio diversification, provided that the covariance among assets is low or negative. Given the broad range of assets and liabilities universal banks are permitted to acquire and sell, it appears that overall operational risks are lower for them than for their specialized counterparts, since “diversification remains a bank’s best protection against risk in an uncertain world” (Federal Reserve Bank of New York (1978), p. 5).

conflicts of interest

It is often argued that conflict-of-interest situations arise more frequently under a universal banking system than under a specialized system, because the latter embraces a much wider range of banking services and consequently is characterized by greater portfolio and service diversification. Conflicts of interest can originate with a combination of, for example, (a) deposit and securities business; (b) securities business for the accounts of customers and for the bank’s own account; (c) lending and security business on the bank’s own account; and (d) lending and underwriting business. By contrast, a bank that is strictly confined to accepting a single type of deposit and extending a single type of loan does not face such conflict-of-interest situations.

However, it is arguable whether it is wise to opt, for that reason, for a specialized system, with its resulting fragmentation. The proper policy under these circumstances is to aim at an optimal trade-off between the benefits of universalization of banking and the problems arising out of conflict-of-interest situations. This trade-off point will, in turn, be a critical function of the degree of competition in individual markets and in the system as a whole. Where intense competition prevails, the danger of conflict-of-interest situations is sharply reduced, because banks cannot in the longer run make portfolio and other decisions at the expense of their customers. 8 Thus, in a competitive environment, every bank that aims at the maximization of long-run profitability will have to establish and maintain lasting relations with its customers—relations that are built on mutual interest and trust.

To sum up, the financial system in any country may be evaluated on the basis of any of the previous five criteria or any combination thereof. Often a financial system is judged to meet one of the criteria, such as scale efficiency or savings mobilization, but does not meet other criteria or satisfies them inadequately. It is imperative that the final judgment be made on the basis of the total effectiveness of a system in the particular context of a given country.

III. Relevance of Multipurpose Banking for Less Developed Countries

In this section, the relevance of the multipurpose banking is examined, on the basis of the same set of criteria that were given in Section II, from the viewpoint of LDCs and bearing in mind the special characteristics of their economies.

efficiency, degree of concentration, and conflicts of interest

At a lower level of economic development, the size of both an economy and its financial sector tends to be relatively small. This fact inevitably has implications for the scope for exploitation of economies of scale in the financial system. Furthermore, for a given market size, any increase in the scale of operations will tend to raise the market power of financial institutions, thereby encouraging negative oligopoly or monopoly effects on prices and output. Thus, overall “supply efficiency” of any particular financial system chosen in an LDC is given by the net result of the positive efficiency effect derived from economies of scale and the negative effect of market power. 9

Figure 1 depicts the general nature of the relations between the degree of concentration and scale efficiency on the one hand, and between the effects of market power on the price, quality, and quantity of financial services on the other. The s curves display the relationship between the degree of concentration and the economies of scale for different market sizes. Both the shape and the relative location of the s curves reflect the assumption of declining marginal productivity. For example, for a market size represented by the s4 curve, the efficiency of the average size of financial institutions that implies a competitive financial sector is equal to e’. For smaller financial sectors, the same degree of concentration implies a lower base level of technical efficiency (i.e., eʺ for a market size s1, and so on). In addition, Figure 1 demonstrates that in a market as large as s4, economies of scale are fully exploited at a level of concentration below monopoly (i.e., c′). In contrast, in a small market like s1 even a monopolistic institution is not able to exploit fully the economies of scale.

The t curve reflects the negative relation between the degree of concentration on the one hand and the price, quantity, and quality of the supply of financial services available to the community on the other hand. As the degree of concentration in the financial sector increases, financial institutions are expected to raise the price and/or reduce the quality and quantity of their services. Since the t curve reflects behavioral adjustments expected to take place with an increase in market power, neither its location nor its shape can be determined with precision. The t curve drawn in Figure 1 presumes that negative behavioral adjustments tend to become more important as market power increases. Alternatively, negative behavioral adjustments may be assumed to start once a positive level of concentration, cʺ, exists.

The relation between concentration and the net efficiency with which financial services are supplied can now be demonstrated by reference to the s and t curves. This efficiency, which may be called “supply efficiency,” is given by the distance between the s curves and the t curve. It is observed that supply efficiency—at least for smaller markets, and assuming that monopolistic behavior is reflected in the t curve—first increases with concentration and then declines. For each individual market size, the point representing optimality is reached where the distance between the t curve and the respective s curve is the largest—that is, at A1B1 for the s1 curve and at A4B4 for the s4 curve in Figure 1. Given the assumptions underlying the s and t curves, Figure 1 shows that the larger the financial sector of an economy (a) the greater is the supply efficiency and (b) the smaller is the degree of concentration that maximizes supply efficiency. For developing countries, which generally have relatively small economies with underdeveloped demand for financial services, this result has two implications. First, the overall contribution of the financial sector to economic growth is necessarily smaller in LDCs than in developed countries with a higher demand for financial services. Notice, however, that the marginal contribution is higher in LDCs than in developed economies. Second, economies of scale and the related degree of supply efficiency can be exploited only at a higher degree of concentration in the financial sector of LDCs compared with developed countries. However, it is evident from Figure 1 that as long as the t curve is shaped as it is in that figure—that is, as long as a strengthening of monopolistic behavior can be assumed with an increase in concentration—the optimal situation will, even in a small developing country, always be one below monopoly.

Within this analytical framework, the implications of the choice of a financial system facing developing countries (i.e., specialized versus universal) for the optimal degree of concentration and the related degree of supply efficiency may now be discussed.

Assume that the overall demand for financial services in a country is reflected in the financial sector size represented by the s3 curve in Figure 2. Assume, furthermore, that the country’s financial institutions are universal, in that they are permitted to supply the entire range of financial services and thus to compete with each other across the board. In Figure 2, the maximum level of supply efficiency for this situation is represented by A3B3. This degree of supply efficiency derives from economies of scale equal to e3 and the negative supply effects owing to monopolistic behavior equal to n3 at a level of concentration c3.

Assume now that the country adopts a specialized system that provides two distinct kinds of financial services, each covering half of its total demand. 10 Under this assumption, institutions are prohibited from covering both lines of services and must choose between the two. In Figure 2, the new size of each market is represented by the s2 curve. It is observed that at the old level of concentration, c3, the halving of the market size owing to specialization results in a supply efficiency of A3B2, which is significantly less than the one obtained for the initial universal banking system. Figure 2 suggests, in addition, that c3 is not the optimal degree of concentration for the given market size s2; rather, the optimal situation is reached at a higher degree of concentration, c2. Despite this increase in the optimal degree of concentration, the new situation still displays a level of supply efficiency, A2B2, that is lower than the level A3B3 attained under universal banking.

The foregoing analysis suggests that the fragmentation of the financial sector that follows from legislated specialization tends to produce two undesirable consequences: a decline in overall efficiency and an increase in the degree of concentration.

The conclusions reached so far follow from two basic assumptions. First, efficiency gains from economies of scale, as displayed by the s curves, are attributed to overall size only; and second, the “monopolistic behavior curve” (i.e., the t curve) is assumed to be the same regardless of the type of financial system. Both of these assumptions may now be dropped, to be replaced by new ones.

It is often argued that specialization increases efficiency. In banking, such efficiency gains are expected to arise primarily from specialized knowledge and information. If, indeed, the efficiency of providing financial services rises with specialization, the s curves in the figures presented so far will require modification. Instead of one single curve representing a given effective market size, there would be two curves, one for a financial system with universal banks and another for a system with specialized banks. The latter curve would be hypothesized to lie to the right of the former, since for a given market size and degree of concentration specialized institutions would be expected to exploit—in addition to the regular economies of scale—specialization advantages.

In Figure 3, it is assumed that specialization advantages grow with size. As a result, the distance between the curves representing a specialized banking system and those representing a universal banking system increases with the degree of concentration for any given market size. It is not easy to quantify the possible efficiency gains from specialization. They could be relatively small, as is assumed for the s2 curve. In this case a country would, despite existing specialization gains, be better off with a universal banking system. In Figure 3, this situation is depicted by a level of supply efficiency, A2B2, of the modified s2 curve that is lower than the level A3B3 for the universal banking case. Observe that the efficiency gains from specialization (move from s2 to s2 do not compensate for the loss of economies of scale owing to the reduction in effective market size (move from s3 to s2). Notice also that the degree of concentration required for maximum supply efficiency increases. If, by contrast, efficiency gains owing to specialization are large, as depicted by the sʺ curve, legislated specialization can raise the supply efficiency, as is depicted by A2*B2 as compared with A3B3. In this case, the specialization gains (move from s2 to s2 are larger than the loss in scale efficiency (move from s3 to s2); specialization thus should benefit the economy.

For the latter result to obtain, it must be argued that specialized and separately established financial institutions will be significantly more efficient than specialized departments of multipurpose banks. While some arguments, such as administrative clumsiness and the inertia of large size, can be advanced in favor of this proposition, it appears unlikely that they overcompensate for the efficiency gains of size and diversification arising from broader and better information and better use of resources—that is, those factors that were discussed in Section II—to not only affect economies of scale but also to constitute a considerable cost advantage for multipurpose banks over specialized banks. It is unlikely, therefore, that specialization by institutions per se would result in a major gain in efficiency of the financial system as a whole.

It could be argued that legislated specialization in the financial sector might improve the quality of services provided, avoid conflict-of-interest situations, and ensure more competitive behavior. In that case, there would be a leftward movement of the t curve toward the t′ curve, as shown in Figure 3, that could result in higher supply efficiency of a specialized system. Under the assumption of simultaneous and moderate specialization gains, the resulting supply efficiency A2B2* is seen to exceed the level observed for the universal banking system, which is represented by A3B3. However, it is hard to imagine that legislated specialization would exercise a significant influence on either the quality of services (as compared with that of a specialized department within a multipurpose bank) or on the competitive conditions. By contrast, certain gains from legislated specialization can be expected in regard to avoiding conflict-of-interest situations. However, such situations can be largely avoided in a universal banking system, as long as a competitive environment is enforced. It is unlikely, therefore, that the t curve representing the behavior of banks in a specialized system will always lie considerably to the left of the one representing the universal banking system.

In developing countries, demand for even basic financial services has often not yet been appropriately articulated. In such situations, it appears desirable to generate through official intervention such special sources of supply that can meet socially desirable, albeit partially dormant, private demand. For this purpose, developing countries have often established new specialized financial institutions to satisfy the previously unmet demand. Operations of such institutions are generally insulated from competition by appropriate legislation and are even given substantial subsidies. Such actions are often defended by arguments that resemble those employed in the infant industry advocacy. However, the efficiency gains expected from such specialized and protected institutions are unlikely to be realized, because the necessary competitive conditions are often absent. In fact, a specialized institution created by special statute often assumes a monopoly position. The establishment of a special institution can be justified only if it will expand the overall size of the financial sector, widen its spectrum of financial services, and reduce the degree of concentration. In order to accomplish these goals, the new institution needs to be broadly based and, after the infancy phase is over, needs to be exposed to competitive forces across the board.

mobilization of savings

Mobilization of savings for investment is one of the most important preconditions for accelerated economic growth. LDCs are generally characterized by a low rate of domestic savings. However, the inadequacy of the domestic savings rate itself seems rather exaggerated. With few exceptions, namely the very poor countries, the gross domestic savings rate has been found to be fairly high in a large number of LDCs. What is really lacking is the speed and efficiency in the process by which savings being accumulated in the economy are translated into savings usable for productive investment. In LDCs, savings tend to remain idle or to be dissipated owing to the inherent inadequacy and inefficiency of institutional channels. Savings therefore need to assume a financial form in which they can be transferred conveniently and efficiently to the most productive investment. The importance of this factor is thrown into bold relief by the low ratio of financial savings to total national savings in most of the developing countries (see Tun Wai (1972)). This implies that there is considerable urgency to devise an institutional network to put savings into financial assets in LDCs.

The transformation of savings into financial form is currently being accomplished in several ways. First of all, as was not the case in the now developed countries during the nineteenth and early twentieth centuries, the governments in most LDCs have been deeply involved in tax efforts to mobilize savings for investment (Chelliah, Baas, and Kelly (1975)). Apart from this, commercial banks have been the major institutions mobilizing savings—first, because of their wide network of offices; second, because the commercial banks, through normal credit operations, often activate savings lying dormant elsewhere; and third, because the banks’ liabilities, which are part of the money supply, are highly liquid and thus attract savers. As a consequence, the principal financial assets in LDCs have been currency and bank deposits (Madan (1964 a) and (1964 b) and Bhatia and Khatkhate (1975)).

Savings are also transformed into financial form through the issue of bonds, shares, etc., which are then absorbed, either directly by the savers or indirectly by the financial institutions, which, in turn, obtain their resources from private savers. Issues of bonds and securities are a rather uncommon mode of providing finance for fixed capital formation in LDCs. Commercial banks in most of these countries provide term loans up to a certain proportion of their liabilities under the active guidance of the central banks (Madan (1964 a)). In addition, special long-term industrial and agricultural funds have been set up at some central banks in order to channel term credit to industry and agriculture through the commercial banking system. In some LDCs, commercial banks have even been given options to convert loans into equity if this is considered by the authorities to be in the interests of the borrowers as well as the lending institution (Srivastava and Oza (1978)). Furthermore, commercial banks also buy shares and debt of the specialized financial institutions or development banks, whose sole business lies in providing medium-term and long-term credit to industry, agriculture, and other sectors (Crick (1965)).

Of course, the involvement of commercial banks in term credit poses certain problems for the liquidity of banks and the safety of their deposits. The liquidity of banks providing term credit is commonly ensured through (a) liberal refinancing from the central banks, subject, of course, to various conditions and procedures and (b) through formal and informal arrangements under which banks will be prevented from transgressing the prudent limits of such lending. Depositors’ interests in many LDCs are protected through deposit insurance schemes and also through appropriate legal frameworks. Above all, there is close supervision of banks’ operations by the central banks in these countries, which also act, in practice, as development agencies.

Thus, in several LDCs, the rudiments of, and preconditions for, universal banking may appear to exist. Yet there are some important differences between developed and developing countries. First, intervention by the government or the central bank is conspicuous and strong, unlike in the typical universal banking case in developed economies. Second, the financial system in LDCs generally remains fractured as a result of the creation of too many institutions that cover selected areas of financial services and that often work independently of each other. Third, the extent of competition is extremely limited, mainly because of the supply-leading pattern of financial development in LDCs. It is for these reasons that LDCs’ financial systems lack both institutional flexibility and responsiveness to autonomous changes in the preferences of savers and investors. As a result, numerous inefficiencies in the financial intermediation process (i.e., the mobilization and efficient allocation of savings) have emerged that have impeded an acceleration in the growth of their economies.

promotion of entrepreneurship

The development of entrepreneurship is another desideratum in LDCs that their financial systems strive to provide. Certain institutional constraints militate against indigenous entrepreneurs, and particularly against new ones. The major problem they face is their confrontation with comparatively high risk, which has been demonstrated to be primarily a function of market imperfections (Masson (1972)). Moreover, the difficulties of the local entrepreneurs tend to multiply, since they do not have easy access to necessary knowledge, information, and expertise. The local entrepreneur’s ability to meet even temporary losses is also limited owing to a lack of resources, which leads to an excessive emphasis on possible losses and an overcautious evaluation of possible profits. The prevalence of risk aversion precludes the selection of projects with the highest possible profit/ probable loss ratios (Awad (1971)). Therefore, entrepreneurs in LDCs are impelled to choose only those investment projects that can yield a certain, albeit moderate, return—a return that would not fall below what is considered to be a “critical income level.” A further problem of the new and small entrepreneurs in LDCs is that their cost of funds is usually very high, since for a bank the cost of administering a loan to them is generally considerably higher on average than it is for loans to larger, well-established companies.

A major responsibility for promoting the development of entrepreneurs in LDCs has devolved on their financial systems. Several measures, often of an ad hoc nature, have been adopted in many LDCs. The most important one among them is the introduction of credit guarantee schemes (Bhatia and Khatkhate (1975)) that insure, to a varying degree, loans granted by banks against possible losses. Technical and managerial assistance facilities of various kinds are also provided by commercial banks or by development banks together with commercial banks (Bhatt (1972)), as is done in most of the countries in West Africa (Bhatia and Khatkhate (1975)). Thus, in addition to their usual activities, the financial institutions also function as mangement consultants and, in that capacity, assist local entrepreneurs with technical advice and implementation strategy, undertaking at times even market research. In many of these assistance programs, the government—or, of greater importance, the central bank—plays a direct and active role.

What emerges from the preceding discussion is the conclusion that the LDCs have attempted to foster entrepreneurship through setting up certain linkages between the industrial and financial sectors. Credit of all types is supplied, promotion of entrepreneurship is undertaken, management of enterprises is supervised, technology is furnished, and training of staff is undertaken by commercial banks and specialized financial institutions or the government, both separately and as joint efforts. Thus, as far as promoting entrepreneurship is concerned, the LDCs may not obtain particular advantages by adopting a universal banking model.

responsiveness and financial stability

For a variety of reasons, developing countries are generally more subject to exogenous shocks than are industrial countries. For one thing, the political environment is often less stable. For another, agriculture, which usually constitutes the mainstay of developing economies, often entails a sizable variance in output. Finally, foreign demand for the products of individual developing economies is subject to more fluctuations than is the demand for the products of developed economies. Therefore, related exogenous changes of world market prices produce frequent, and often important, shocks to both domestic incomes and prices. While LDCs cannot prevent such exogenous shocks, they can reduce their immediate impact by developing a responsive system of financial institutions and efficient financial markets that can function as a shock absorber for the real sectors, thereby reducing the costs of adjustment to such shocks. Constraints that either limit the room for maneuver of individual financial institutions or the interactions among them will stifle the process of flexible financial intermediation. For this reason, it is clear that a universal banking system will be better suited to fulfill the shock-absorber function than a specialized system, especially in the case of developing countries with their underdeveloped or nonexistent financial markets.

At first sight, the responsiveness of financial institutions may appear to be incompatible with their financial stability. Responsiveness suggests that portfolio adjustments and behavioral changes may be frequent, which could be viewed as unlikely to promote financial stability. Nevertheless, a prerequisite for both responsiveness and financial stability is the flexibility with which institutions can adjust to changing demands and relative returns. An institution that, in the extreme case, is confined to offering one product will be greatly affected by any shift in demand. Its overall stability, and in particular its earnings performance, will depend entirely on the stability of demand for that single product. Not so with a diversified multipurpose institution. Even if it should feel the full impact of the shift in demand for one of its products, its overall stability remains largely unaffected by such a shift. Diversification that generates a high degree of responsiveness of financial institutions also enhances their stability.

In Section II, financial stability has been described as being dependent on financial institutions’ approach to portfolio management, and in particular to term transformation. It could be argued that, as a result of the higher degree of political and economic instability in developing countries, the scope for safe term transformation is significantly lower than in more stable industrialized countries. On the other hand, it appears that the need for term transformation is greater in developing countries because of the greater preference of savers there for liquid assets and the simultaneous imperative of higher capital formation for growth. In this situation, long-term lending by financial institutions usually takes the form of rollover credits, which are, however, generally inefficient from the point of view of the allocation of resources, since they favor larger, well-established borrowers over new and smaller borrowers. Given the relatively strong need for term transformation in developing countries, their overall financial stability will depend largely on the safeness of term transformation actually practiced. Regression results on the stability of demand, time, and savings deposits for a selected number of developing countries, which are reported in Table 3, suggest that a surprisingly high percentage of such nominally short-term funds as demand deposits can be relied on to remain available over more than one period. As expected, the percentage core is even higher for the average measure of future periods. In a number of cases, the core of demand deposits is larger than that of time and savings deposits combined. This may be due to the often observed fact that small savers, who are the main holders of savings deposits, display a more erratic deposit and withdrawal behavior than do firms, which are the principal holders of demand deposits. The results in Table 3 clearly indicate that an important percentage of financial resources usually considered short term can be used for longer-term lending without endangering the liquidity of the financial system.11 In the absence of efficient money and capital markets in developing countries, the necessary term transformation can to a good extent be safely performed by internal transactions of individual institutions covering the whole range of maturities. Whether this is permitted depends upon the legislative and regulatory arrangements devised to guide financial institutions. As was suggested previously, owing to their greater degree of diversification, multipurpose banking institutions are likely to display a high degree of financial stability and may therefore also be better suited than specialized institutions to cover the need for term transformation in developing countries.

Table 3.Deposit Core in Developing Countries1(In per cent, monthly data, January 1970-December 1978)
Time and Savings Deposits
Demand DepositsCommercial banksOther financial institutionsTotal
CountryFirst period 2Average 3First period 2Average 3First period 2Average 3First period 2Average 3
Lebanon 492.698.296.599.1
Sri Lanka84.396.874.994.672.193.975.194.7
Taiwan 589.097.793.798.790.298.093.998.8
Tanzania 684.096.686.997.388.197.687.897.5
Turkey 791.198.282.796.5

The core determination involves the forecasting of out-of-sample values and the calculation of confidence intervals for them. For a description of the method employed, see the Appendix.

First out-of-sample period.

Average core for future periods.

Data period is January 1970-July 1975.

Data period is January 1970-April 1978.

Data period is January 1970-June 1978.

Data period is January 1970-November 1978.

The core determination involves the forecasting of out-of-sample values and the calculation of confidence intervals for them. For a description of the method employed, see the Appendix.

First out-of-sample period.

Average core for future periods.

Data period is January 1970-July 1975.

Data period is January 1970-April 1978.

Data period is January 1970-June 1978.

Data period is January 1970-November 1978.

IV. Some Broad Conclusions

The importance of the financial sector for the growth and development of economies has only recently been fully appreciated. Much of the contribution that the financial sector can make to growth and development depends upon the legislative and regulatory arrangements devised to guide the activities of financial institutions and markets. The formulation for the financial sector of appropriate laws that will promote growth and development through innovation and competition is not an easy task. Such legislation has to reckon with the various types of tradeoffs, such as the one between a highly competitive and a financially stable financial system, and the one between the exploitation of economies of scale (and the related efficiency) and the increase in market power that often accompanies it. Generally, it can be maintained that the best strategy for rapid economic development is to permit the financial sector a high degree of flexibility, so that it may respond to the demands of both savers and investors. Where significant market power exists, such responsiveness is weakened; similarly, a lack of competition reduces the efficiency with which services are provided and consequently leads to an inferior allocation of resources. Active and keen competition requires a minimum number of financial institutions competing with each other on as equal terms as possible. On the other hand, too many institutions are not desirable either, since their ability to respond to changing and increasingly more sophisticated demand is inhibited.

The fundamental economic, historical, and political factors in developing economies differ from one to the other. Nevertheless, the basic pattern of financial behavior is not dissimilar across countries. On the basis of various evaluation criteria, such as efficiency and financial stability, it is found that the LDCs may benefit more from the adoption of a multipurpose banking system, in particular from the exploitation of economies of scale without the inefficiency that often ensues from an increase in the degree of concentration, than from adoption of a specialized system. A multipurpose banking system implies a more liberal approach to banking legislation, thereby avoiding the fragmentation of the financial system that usually accompanies strict legislated specialization. This is not to say, however, that specialized institutions should not be permitted where there are proven gains from specialization. In all cases, though, specialization should reflect comparative advantages and a free choice rather than legislated constraints.

While lending support to the universal banking system, this paper does not ignore some of its potential shortcomings. It is recognized that conflict-of-interest situations are more likely to arise in a universal banking system than in a specialized system. Besides, in some LDCs, existing banks are already very powerful, and allowing them to cover a larger range of financial services could increase their power even more. Both cases call for official intervention. However, conflict-of-interest situations and the prevalence of excessive market power require legislation, like antitrust laws, oriented toward tackling these problems directly rather than indirectly through a narrowing of the range of activities a financial institution can cover. The possibility that such unsavory practices may recur should not be taken as a pre-text to devise a straightjacket of banking legislation that would destroy the responsiveness, flexibility, versatility, and the dynamism of the financial system.

APPENDIX: Deposit Growth, Stability, and Stable Core

In order to calculate the information contained in Table 3 of Section III on the growth of various types of deposits, their stability, and their stable core, the following econometric and statistical calculations were carried out.

The first step in the investigation involved the regression of demand, savings, and time deposits individually against time in order to determine their past growth and its stability. The equation estimated for each type of deposits is given by

where ln is the natural logarithm, Xi is the financial instrument i, and t is time. The coefficient b gives the monthly rate of growth, and its t-statistics allow for inferences about the stability of the growth path.

The second step involved (a) the forecast—on the basis of the regression of equation (1)—of the first out-of-sample value for each financial instrument, (b) the calculation of the lower-end value at a one-tailed 95 per cent confidence interval, and (c) the expression of the lower-end value as a percentage of the estimated out-of-sample value.

A further calculation was carried out in order to determine the percentage of the selected range of short-term financial instruments that would on average remain available to financial institutions over future periods.

The confidence intervals for the first period and the average are calculated in the following way: the confidence interval for the first out-of-sample-period estimate is given by

where V is the value of the explanatory variable (time) in the next period (i.e., V = n + 1 with n being the number of observations in the sample); k is the value of the Student’s t-distribution for a one-tailed 95 per cent confidence interval; SEE is the standard error of the estimate of equation (1); and t¯ is the average value of the explanatory variable (time). The formula for the calculation of the average percentage is the same as for the first-period calculation, with the exception that the root value is given by

that is, without the addition of 1 in the root.


Mr. Khatkhate, Advisor in the Central Banking Department, is a graduate of Bombay and Manchester Universities.

Mr. Riechel, economist in the Central Banking Department, is a graduate of Gӧttingen University, Federal Republic of Germany, and received his doctorate from Kiel University, Federal Republic of Germany. He has taught at the Graduate Institute of International Studies in Geneva and at Kiel University, and has worked with the Organization for Economic Cooperation and Development in Paris. He is the author of Economic Effects of Exchange Rate Changes and of a number of articles in professional journals.

Banks in the United States and Canada developed in recent years the system of interlocking directorships with industry. This practice came into vogue in the aftermath of diversification of functions of commercial banks.

Before the introduction of a single currency for the entire German Reich in 1875, some 33 independent banks still possessed the right to issue currency.

In 1835, more than 250 such savings banks already existed within the territory of what later became the Reich.

Some initial restrictions, such as the limitation of savings banks’ acceptance of savings deposits, were lifted early enough (1908) to ensure that these institutions could move into areas (sight deposits, checking accounts, overdrafts, etc.) that would be of major importance to their respective customers, though they had previously been of little importance to customers.

The Herfindahl index is defined as H=Σi=1n(Ai/Σi=1nAi)2 where Ai is the amount of the ith bank’s assets. The Herfindahl index takes account of the number of banks as well as deviations from the average bank size. The higher the index, the greater is the concentration (Adelman (1969) and Short (1979)).

This theory, which is often called the “sediment theory” of deposits, stands in sharp contrast to the theory underlying the “golden rule of banking,” which implies that the term of liabilities should be matched with the corresponding term of assets.

In order to investigate this question for developing countries, an analysis of the cores of such deposit liabilities as demand and time and savings deposits—all being generally considered short-term resources in contrast to bank bonds and equities—was carried out. The results are reported in Section III. They confirm the view that a sizable proportion of present-period deposits remains available for lending over future periods.

The lack of active competition in large areas of the highly specialized U.S. financial system has been spelled out by the President’s Commission on Financial Structure and Regulation (1972) and has prompted it to recommend that most financial institutions should be allowed to offer a wider range of financial services and to compete actively and directly with one another on the basis of comparative advantage. As a result of powerful lobbying, however, no legislative changes of any consequence have so far been introduced. In other countries, however, like Sweden (1969) and the Netherlands (1973), recent legislation has reduced fragmentation and encouraged a move toward multipurpose banking in an attempt to increase competition.

Note that, as a result of the inclusion of the effects of market power, “efficiency” is defined in the sense of social efficiency rather than private efficiency.

This simplifying assumption is not indispensable for the arguments presented here but is used to facilitate the graphical presentation.

This conclusion is based on an investigation into the stability of the banks’ funding side only. Arrears in loan repayments or even defaults, both of which are not uncommon in developing countries, may cause additional, and possibly simultaneous, problems on the banks’ uses side. Quantitative information on repayment performance that would have allowed for further empirical research of the safe degree of term transformation was not available.

Other Resources Citing This Publication