Printer Friendly

Informal financial arrangements and the stability of deposit insurance in less developed countries.

I. Introduction

Deposit insurance first appeared in the State of New York in 1829 to guarantee banknotes and deposits. A number of other states quickly established similar schemes. Some of these schemes were short-lived, as their true cost became known in the panic of 1837, while others were terminated only after the Federal government instituted taxes on banknotes, thereby effectively eliminating all competitors in the currency creation market.(1) After the Knickerbocker Trust panic of 1907, there was renewed interest in deposit insurance, but only limited adoption. The great depression during the 1930s led to the current system, which is mandatory for banks that are members of the Federal Reserve System and voluntary for some state-chartered banks.(2)

A formalized deposit insurance system like that in the U.S. is found in only a small number of other countries, with varying degrees of coverage and membership requirements.(3) Significant omissions include Panama and Hong Kong, who have large banking sectors in their economies. Accordingly, the role of deposit insurance in the growth of financial intermediaries is fairly ambiguous. One reason that deposit insurance plays a smaller role in developing countries is that they typically suffer from a less diversified economic base, making them more prone to liquidity crises that deposit insurance cannot prevent. Informal financial arrangements, such as savings clubs and rotating credit associations, may be more effective in providing credit than banks or other financial intermediaries, primarily because they do not provide demand deposit accounts and other banking services that may lead to an immediate withdrawal of funds. Informal arrangements often provide for direct renegotiating between borrower and lender. The ability to renegotiate directly with a borrower, not using a bank as an intermediary, provides more liquidity for the lender than a bank offers under similar circumstances because banks must consider the claims of all other depositors as well.

Recent research suggests that banks and other financial intermediaries evolve to lower the cost of monitoring loans and of converting illiquid into liquid assets [4; 5; 20; 21; 22]. In Diamond and Dybvig's [51 model, deposit insurance may prevent a "bad" equilibrium - bank runs and bank failures - from occurring. In their model, a random event causes a change in expectations about the future value of deposits and produces a bank run, which deposit insurance prevents by fixing the value of deposits to stabilize expectations. In Williamson's [22] model, deposit insurance does not have a stabilizing (or destabilizing) role because bank failures are caused by a particular state of endowments, preferences, and technology. A bank failure is not a "bad" equilibrium caused by random events affecting expectations, but merely an equilibrium that may arise given the model's assumptions.

The model developed here uses Williamson's [21] generic intermediation model as a starting point and introduces mixed monitoring strategies into the environment. Mixed monitoring strategies permit the intermediary to monitor randomly, somewhat like the U.S. Internal Revenue Service does with the Taxpayer Compliance Program.(4) The alternative is to preconimit to a pure monitoring strategy, in which all borrowers will either be monitored or not monitored. Williamson considers only pure monitoring strategies in his model. It is shown that the strategy of always monitoring is not a Nash equilibrium; that is, intermediaries can improve profits by monitoring project outcomes a little less often.(5)

Deposit insurance has a role in the model developed here, but it is not likely to arise in less diversified economies where standard banking institutions are expected to be unstable over time; that is, banks in countries with a less diversified economic base are potentially subject to more panics and runs. This model develops conditions under which formal financial intermediaries do not operate in less diversified areas, which gives rise to many informal financial arrangements that make loan payoffs contingent on project outcomes. The model's results are consistent with McKinnon's [13] discussion of financial repression and Shaw's [16] discussion of the lack of financial deepening in less developed countries, except that government policy is not the cause of reduced financial intermediation.

Informal financial arrangements may take many forms, such as direct lending and rotating credit clubs. This paper discusses rotating credit clubs, because they mobilize deposits as well as make loans and have rules that overcome many of the drawbacks faced by banking institutions. For example, rotating credit clubs specify the holding period for deposits and the frequency of withdrawal by members. By pre-specifying the terms of withdrawal, they immunize themselves against runs and allow for an orderly period of renegotiating in the event of default. Moreover, these clubs often offer a type of deposit insurance to members in that the organizer of the club may guarantee the payment of dues from other members.

The remainder of this paper is organized as follows. In section II, the arguments for and against deposit insurance are discussed. The question of whether government should have a role in the provision of deposit insurance is addressed. One type of financial arrangement found in informal markets is described and compared to financial contracts found at standard banking institutions. In section III, a model of financial intermediation with overlapping generations is developed. This model shows that random events affecting the value of all loans may make financial intermediaries unstable. An appeal to the law of large numbers to make banks stable, as is done by Diamond [4], Waldo [18], and Williamson [21], does not apply to loans in less developed countries because their economic activity is closely related to only a few primary sectors or commodities, and thus less diversified. Section IV provides some empirical support for the claim that less diversified economies use less financial intermediary services. Finally, section V provides a few concluding remarks and suggestions for future research.

II. Deposit Insurance and informal Lending

The basic argument in favor of deposit insurance is that it discourages bank runs during periods of unusual demand for liquidity. Banks may operate more efficiently if they can weather these periods without closing their doors, calling loans, and conducting a "fire sale " of assets. On the other hand, guaranteeing deposits encourages risky lending practices, which increases the likelihood of bank runs. Unless the quality of a bank's loan portfolio is restricted and monitored, deposit insurance simply creates a moral hazard problem.

The need for monitoring is an often cited reason for government to be involved in providing deposit insurance.(6) A government may enforce the penalties imposed on a bank if its officers violate lending or capital constraints, with the ultimate penalties being the removal of officers, forced sale of the bank, or withdrawal of the bank's charter. A private insurer may cancel a bank's insurance if it violates the rules, but this action may create a run on deposits as depositors learn of the bank's insolvency or financial irresponsibility, and this action limits the usefulness of the insurance agreement with depositors. Depositors will realize the tenuous structure of private deposit insurance contracts and quickly discount its value.

In general, it is difficult for a private insurer to overcome the moral hazard problem. The private insurer is insuring depositors, not shareholders, and thus the incentive for risky lending practices still exists. In bad times, shareholders may find it in their interest to transfer losses to depositors at the expense of the private insurer by simply declaring bankruptcy. Internalizing ownership of the bank and private insurer could solve the moral hazard problem, but on the surface this would not offer meaningful insurance to depositors, who would perceive that their deposits were insured by the bank, and thus not insured at all.

A solution that is likely to arise in a competitive market is for shareholders to post sufficient collateral to guarantee that they will not act opportunistically with deposits. Depositors will only patronize banks that have offered such guarantees, and thus all banks will offer this implicit insurance. Generally, investments in specific assets, which are assets that have a higher value if the firm is in operation than otherwise, such as advertising and reputation, act as collateral.(7) Thus, it is not always necessary for government to offer deposit insurance to guarantee the safety of banks.(8)

Even though banks may be safe from shareholders acting opportunistically, they are not completely insulated from bank runs under either government deposit insurance or private collateral schemes. Under a private scheme, it is still possible for a series of loans that were appropriate ex ante to default ex post. The problems in the energy and agricultural sectors of the U.S. economy during the mid-eighties illustrate this possibility. A run may also occur with govermnent insurance if there is an appreciable time lapse before depositors have access to their money. Moreover, if the event that precipitated the bank failure is universally distributed throughout the economy, such as a stock market crash, depositors may rationally expect the government to respond by printing money to meet liquidity requests and thereby create inflation.(9) The more quickly depositors can convert their financial assets into real assets, the less they will be affected by the subsequent inflation.

In informal financial markets, deposits are mobilized and loans made using nonstandard methods of financial intermediation in which deposit insurance is not explicit. Savings clubs and rotating credit associations, such as those described by Bouman [2; 3], Geertz [8], Ghate [9], and Osuntogun and Adeyemo [15], are common in developing countries and may, in some cases, mobilize more capital than the formal financial sector. These clubs typically have a small number of members, who agree to a predefined savings and lending plan. Typically, members agree to contribute a certain amount each period to the club, with the number of periods equal to the number of members. The total contributed at each meeting is lent to one member. A rationing scheme, often in the form of an interest rate auction, is used to determine which member receives the funds.

For example, a small club may have ten members contributing $10 every week for ten weeks. At the first meeting, the funds collected are loaned to the organizer of the club, who usually bears the ultimate responsibility for paying the "dues" of any member who reneges on his obligation to the club. The organizer receives an interest-free loan, which is in fact a payment for his organizational skills and for providing the guarantee to pay the uncollected liabilities of any member, which can be thought of as a type of deposit insurance. At the second and subsequent meetings, members who have not won an auction in an earlier period bid for the funds collected. Bids are in the form of an interest rate, where a bid of 10%, say, implies that the bidder receives 90% of the total funds. In the present example, a winning bid of 10% implies that $90 is paid to the winner. Interest is thus paid immediately, not in arrears.

These clubs do not suffer from runs like banks do. The pattern of deposits and withdrawals is agreed to by members and they do not offer demand deposits as part of the arrangement. Nonetheless, they provide liquidity and a type of deposit insurance. One characteristic that distinguishes these informal institutions is that depositors may negotiate or re-negotiate directly with borrowers if there is a payment problem. In a bank, the contract with depositors is not conditional on the contract with borrowers. Thus, a run on deposits may be an appropriate response when a bank is in trouble. These clubs do have a scale weakness, however, when compared to banks. The club rules generally equate the number of loans to the number of members, thereby assuring that every member receives a loan or the return of principal and interest. In banks, there are usually many more depositors than borrowers, and thus banks offer larger loans than those offered by informal clubs.

The essential conjecture of this paper is that one of the primary reasons these informal arrangements arise is not to compete with banks and other financial intermediaries, but to provide savings and lending services in regions or countries where banking institutions do not operate, because a lack of economic diversification makes banks that offer demand deposit accounts unstable. Another compelling reason for their existence is to avoid govenrment taxation and regulation, but we are not considering this possibility in the present model.

The next section suggests why it is unprofitable for banking firms to operate in less diversified economies. The argument is that these areas are more prone to external shocks that will lower the value of all bank loans, scaring depositors into withdrawing their funds and causing the bank to collapse.

III. A Model of Financial Intermediation

An overlapping generations model is used to illustrate the behavior of financial intermediaries. The present model differs from those developed by Wallace [19] and Williamson [21] in two important ways. First, it allows strategic behavior between borrowers and financial intermediaries. Borrowers may find it in their interest to deceive intermediaries about the success of their project. As a result, intermediaries may elect to monitor borrowers to determine the truthfulness of their claims. This approach differs from Williamson's by allowing mixed monitoring and cheating strategies, which are essentially random strategies. Pure strategies, which are a form of pre-commitment, are not a Nash equilibrium in this model.

Secondly, this model introduces an element of catastrophe in the decision process. Financial intermediaries must consider the possibility that an exogenous event may affect the return on their entire loan portfolio. This feature captures the investment climate found in many developing countries or isolated areas, where national product is generated by only a few sectors and capital and labor mobility is limited. McKinnon [13] describes this situation as a "fragmented economy," one in which differential access to new technologies limits growth and government policy tends to exacerbate the differences between poor and modernizing sectors. Shaw [16], too, has noted that many developing countries lack financial "deepening" in part because their economies are more unstable and less diversified.

Lenders and Entrepreneurs

In each period t, there are a finite number (N) of individuals born. These individuals are either lenders or entrepeneurs, depending upon whether they are born with or without an endowment (e), respectively. Each individual lives for two periods. Lenders maximize expected utility over consumption in both periods; that is, each lender maximizes:

E[ln[C.sub.1] + [C.sub.1 + 1]), subject to [C.sub.1] + [S.sub.1] [greater than or equal to] [C.sub.1 + 1] [greater than or equal to] [r.sub.t][S.sub.t]

where E is the expectation operator, [C.sub.t] is consumption in period t, [C.sub.1 + 1] is consumption in period t + 1, [S.sub.t] is savings in period t, and [r.sub.t] is the gross return on savings between period t and t + 1. The lender's utility is concave in the first period - implying diminishing marginal utility of consumption - to induce some savings behavior. Lenders are risk neutral when making their second-period consumption decision because the random elements of the model only affect consumption in period t + 1. A more general specification of utility is feasible, but this detracts from the important task of modeling financial intermediation and does not add any new insights.

The solution to the lenders' problem implies that savings may be written as:

[S.sub.t] = e - 1/[r.sub.t].

Lenders will consume in both periods if e > 1/[r.sub.t]. Savings varies directly with [r.sub.t], and may take two forms: Lenders may elect to hold fiat money, whose price is [p.sub.t], in terms of the consumption good, or to invest in a risky project proposed by an entrepreneur. There are M units of fiat money circulating in each period.(10) The expected return to holding fiat money is [Ep.sub.t + 1] / [p.sub.t]. Savings is allocated between investments so as to equate expected returns. Thus, the expected rate of return to risky investments equals the return to holding money, [Ep.sub.t + 1] / [p.sub.t]. To conserve notation, [r.sub.t] will represent the gross rate of return on risky investments, and thus, in equilibrium:

[r.sub.t] = [Ep.sub.t + 1] / [p.sub.t].

Entrepreneurs maximize expected consumption in the second period of life, [EC.sub.t + 1]. In the first period, they have access to an investment project that requires K units of the consumption good to initiate. All entrepreneurs will thus consider projects of equal size. The investment of K is sufficiently large that many lenders are required to fund a single project. The payoff to an investment project is either [Z.sub.t] or zero in period t + 1. The probability of success, denoted by [pi], is specific to an entrepreneur. This implies that entrepreneurs are different only because they have a different chance of success; their final payoffs vectors do not differ.(11) The probability of success is uniformly distributed throughout the entrepreneur population over the interval [[pi,sup.l], [pi.sup.u]].

Entrepreneurs and lenders share the same information about project payoffs and the chance of success in period t. Lenders may then identify entrepreneurs by their probability of success, which introduces a form of credit rationing into the model (see below). In period t + 1, the success of the project is only known to the entrepreneur. Entrepreneurs may find it in their interest to misrepresent the outcome of the project, and thereby keep the entire return for themselves. Realizing this, lenders will only finance a project if the final outcome can be verified. Lenders may discover the result by spending [beta] units of the consumption good to monitor the returns of the project. Because a project requires more than one lender, there is a duplication of monitoring effort if all lenders decide to monitor the outcome. One way to reduce these costs is for lenders to elect a representative to monitor the entrepreneur with the costs shared across all lenders. Monitoring costs may be reduced under this scheme, but the incentive of the elected lender is questionable. The entrepreneur may effectively bribe the elected lender to misrepresent the results to the others and they both will gain from their malfeasance. It may be noted that there is no reputation effect in this model because individuals are alive for only two periods. A type of moral hazard is therefore present in this scheme.

Financial Intermediation

A solution to the moral hazard problem is for one of the lenders to act as a financial intermediary, who lends to many entrepreneurs and thereby economizes on monitoring costs. The intermediary offers a fixed return to depositors, who are the other lenders. The fixed return resolves the incentive problem faced by the lender charged with monitoring the project's outcome. This intermediary must perform up to a minimum level, which is specified by the fixed rate of return ([r.sub.t]). It can make this commitment because it finances a large number of entrepreneurs, and the law large numbers insulates it against the likelihood of many failed projects.[12]

Monitoring by intermediaries and misrepresentation by entrepreneurs is modeled as a single period bi-matrix game. This formulation recognizes the interdependence between these two parties. That is, the decision to monitor affects the entrepreneurs' decision to cheat and vice versa. A non-cooperative mixed strategy is followed by both entrepreneur and intermediary. The intermediary selects a probability of monitoring, denoted by x, and the entrepreneur selects a probability of cheating, denoted by y. The effect of these strategies on the net return to the entrepreneur in period t + 1 is as follows:[13]

[Mathematical Expression Omitted]

where [V.sub.t] is the payment due the intermediary for the loan, which is a function of [pi], and F is a non-negative fine the entrepreneur must pay when cheating is discovered. The first term in (4) represents the payoff to the entrepreneur if the principal does not monitor and there is no misrepresentation (x = O,y = O). The second term is the payoff if there is misrepresentation and no monitoring (x = O, y = 1). This is the entrepreneur's greatest return. The third term is the payoff with monitoring but without misrepresentation (x = 1, y = O). The final term is the payoff with both monitoring and misrepresentation (x = 1, y = 1). It is assumed that the act of monitoring uncovers the entrepreneur's charade completely. Therefore, we are assuming that the monitoring technology is perfect.[14]

Similarly, the expected return to an intermediary from a given loan may be expressed as follows:

[Mathematical Expression Omitted]

where each term may be interpreted in a manner analogous to the interpretation of (4), except that there is zero payoff to an intermediary if the entrepreneur cheats and there is no monitoring.

The pure strategy solutions to this game are not Nash equilibria.[15] The no monitoring and no cheating strategy works for the intermediary, but fails for the entrepreneur. The entrepreneur finds that [[Z.sub.t] - [V.sub.t])[beta] < [Z.sub.[pi]] because [V.sub.t > 0. It is therefore optimal for the entrepreneur to cheat if the intermediary does not monitor. The cheating and no-monitoring solution is fine for the entrepreneur, but fails for the intermediary. The intermediary may gain from monitoring if ([V.sub.t] + F)[pi] > [beta]; that is, if the expected payoff plus the fine exceeds the cost of monitoring. The monitor with no cheating solution works for the entrepreneur, but fails for the intermediary. The intermediary receives the fine if the agent is caught cheating, and thus it is profitable to offer some incentive to cheat if monitoring costs are to be incurred. The unlikely result that there is monitoring and cheating fails because the fine always reduces entrepreneur's wealth relative to the no-cheating alternative.

The only stable solution to this bi-matrix game is a mixed strategy equilibrium, in which the intermediary monitors with probability [x.sup.*] = [V.sub.t]/([V.sub.t] + F) and the entrepreneur cheats with probability [y.sub.*] =, [beta]/([V.sub.t] + F)[pi]. Substituting [x.sup.*] and [y.sup.*] into (5), the return of the intermediary may be written as:

E[[R.sup.i]] = [V.sub.t][pi] - [beta][x.sup.*]

which states that the intermediary may expect to receive the promised payment [V.sub.t] with probability [pi] and incur a monitoring cost of [beta] with probability [x.sup.*].

In addition to the likelihood (1 - [pi]) that a given project will fail, an intermediary faces the possibility that an external factor affects the returns on all projects. A common catastrophe, such as bad weather during the growing season or a fall in the world price of a major commodity, is captured by this possibility. The negative external factor occurs with probability (1 - [theta]) and eliminates all project returns. This factor presents an exception to the law of large numbers argument used above, because it is not possible for an intermediary to diversify away this risk.

Free entry into financial intermediation implies that intermediaries earn zero profits on the projects they finance. Intermediaries are committed to pay lenders a fixed return of [r.sub.t] on their investment. Thus, the expected payment from an entrepreneur is set to give an intermediary an expected return of [r.sub.t] K on the project loan, which is the amount paid to lenders. The loan payment, [V.sub.t] solves: ([V.sub.t][pi] - [beta][x.sup.*][theta] = [r.sub.t] K, (7) subject to the restriction that the payment may not exceed the maximum return ([Z.sub.t]) expected by the entrepreneur. Equation (7) has two roots, with the positive root being the appropriate solution.

Credit Rationing

Given that entrepreneurs have different probabilities of success, intermediaries will choose only those entrepreneurs whose likelihood of payoff offers a return sufficient to meet their obligation to depositors. With [Z.sub.t] the maximum payoff that an entrepreneur may expect, the cutoff probability, [[pi].sup.*] is defined such that those entrepreneurs with a success probability [pi] [is greater than or equal to] [[pi].sup.*] receive loans and those with a success probability [pi] < [[pi].sup.*] do not. The latter individuals consume zero units of the consumption good in period t + 1. Using (7), the cutoff probability is given by: [[pi].sup.*] = [([r.sub.t] K/[theta]) + [beta][x.sup.z]]/[Z.sub.t], (8) where [x.sup.z] = [Z.sub.t]/([Z.sup.t] + F) and the dependence of [[pi].sup.*] on t is suppressed for notational convenience. Note that [Mathematical Expression Omitted]. If the exogenous event becomes more likely, then fewer entrepreneurs will qualify for loans. In addition, if the fine for malfeasance is increased, then there is less monitoring and (8) implies that the cutoff probability decreases. Thus, fines and enforcement policy affect the degree of credit rationing in the model.

Monitoring costs have a direct effect on the cutoff probability. If [beta] increases, fewer entrepreneurs will qualify for project financing. This result has important implications for the nature of financial arrangements in informal markets. The rotating credit clubs, discussed above, use a different monitoring technology than most financial intermediaries. Weekly meetings and small membership size may lower monitoring costs relative to a banking institution. These clubs may thus finance the sub-marginal entrepreneurs, who are not offered loans because of high monitoring costs.

The amount of loans made, in terms of the consumption good, depends on the relative size of the lender and entrepreneur populations. If [alpha] represents the fraction of lenders in the total population and (1 - [alpha]) represents the fraction of entrepreneurs, then the total amount of loans made is: [L.sub.t] = (1 - [alpha]NK([[pi].sup.u] - [[pi].sup.*])/([[pi].sup.u] - [[pie].sup.1]). (9)

The balance sheet of the economy will equate total savings to the amount of loans plus holdings of fiat money. This condition may be written as: [alpha]N[S.sub.t][p.sub.t]M + [L.sub.5], (10) where [p.sub.t] M represents the holdings of fiat money expressed in terms of the consumption good and 1/[p.sub.t] implicitly measures the price level.

General Equilibrium

The description of the model is completed by specifying the intertemporal stochastic structure of [Z.sub.t] and [p.sub.t]. Following Williamson [21], assume that these variables change according to a two-state Markov process. There is a good and bad state in the economy. In a Markov process, the probability of moving between states is only dependent upon the current state of the economy. That is, if [w.sub.g] and [w.sub.b] represent the good and bad states, respectively, then

Pr[[p.sub.t] + 1 = [w.sub.g]\[w.sub.i] = [q.sub.i], (11) where i is an index of good and bad states, and [q.sub.i] represents the probability of being in the good state in period t + 1 given that the current state is [w.sub.i]. This formulation implies that price expectations are formed as follows:

E[[p.sub.t] + 1\[w.sub.t] = [w.sub.i]] = [q.sub.i][p.sub.g] + (1 - [q.sub.i])[P.sub.b]. (12)

After substituting (12) into (3) and using (3) to substitute for [r.sub.t] in equations (2) and (8), the model admits a solution for current prices, [p.sub.i], aggregate loans, [L.sub.i], the cutoff probability, [[pi].sup.*], and savings, [S.sub.i], that depend on the current state of the world, i. Equilibrium prices are generated by solving:

[Mathematical Expression Omitted]

where i = g, b. The remaining variables may be solved by substituting the solution for [p.sub.i] into equations (2), (8), and (9).

Instability and Deposit Insurance

An important parameter in less developed countries is [theta]. The lower the value of [theta], the more likely that an external shock will occur that affects all loan projects. Accordingly, less developed countries are expected to have a lower value of [theta] than developed countries, primarily because their national product is less diversified (i.e., dependent on fewer commodities) than a developed country. In this model, there is a critical value, denoted by [theta] that represents the point at which financial intermediation is unstable. The critical value is given by: [[theta].sup.*] = [r.sub.i]K/([Z.sub.i][[pi].sup.u] - [beta][x.sup.z]) (14) which is a function of the current equilibrium state. If a region or country has a risk such that [theta] < [[theta].sup.*], then the cutoff for loans in that area is [[theta].sup.*] = [[pi].sup.*], and no loans are made. In effect, financial intermediaries are unable to commit to fixed interest payments on deposits, as the effects of an external shock have offset those of the law of large numbers. Informal savings arrangements, referred to above, are likely to arise in this environment because lenders have an incentive to save and entrepreneurs desire to consume in period t + 1. These arrangements make payoffs conditional on project outcomes because unconditional payoffs cannot be a stable equilibrium. The rotating credit clubs effectively operate in this manner. The periodic meetings act as a form of monitoring members. If a member has a problem with making a payment, it is known to all other members simultaneously. With all members in attendance, negotiations may take place with the delinquent member. Thus, in a sense, a member's payment is determined by the project outcome. If no arrangement is possible, the organizer of the club is responsible for the shortfall. The club organizer thus provides deposit insurance, albeit in only a rudimentary form.

In this light informal financial arrangements act as a natural extension of formal financial intermediation. Banks, acting as formal intermediaries, find that they cannot profitably offer demand deposit services and loans to less diversified economies. Informal financial arrangements, which are organized under a set of rules that effectively make returns depend on payoffs, arise to fill the void.

IV. Empirical Evidence on Financial intermediation and Diversification

A crucial question is whether less diversified economies use formal financial intermediation to a smaller extent than more diversified economies. The United Nations' Commodity Yearbook for 1991 gives a measure of how concentrated output is in various countries f 17]. The concentration index, denoted, is defined as the share of exports for the country's three leading commodities. This measure is expected to be highly correlated with the overall output concentration in most developing countries, because exports are often the largest component of national output in these countries. The model presented here predicts that as export concentration increases, financial intermediation decreases.

To test this hypothesis, we estimated the cross-country relationship between the ratio of time and savings deposits to gross domestic product, denoted S/GDP, and country concentration. Effectively, this is a money demand relationship, so a measure of the real interest rate, equal to the annual time deposit rate minus actual inflation in the year (denoted RR), and per capita income, equal to gross national product divided by population (denoted PCGNP), are included in the estimation equation.(16) Data on all of these variables are available for twenty-nine less developed countries for 1988.(17) In logarithmic form, the estimated regression is:

[Mathematical Expression Omitted] (15)

where "i" is used to index countries and the terms in parentheses are estimated t-ratios with an asterisk indicating significance at the 10-percent level. The adjusted R-squared for this regression is .285.

Equation (15) supports the maintained hypothesis: Countries with less diversification (higher EC) tend to place less savings with formals financial intermediaries. The country concentration variable is the only one that is significant in this equation, which suggests that diversification plays a more important role in explaining the extent of financial intermediation across countries than per capita income or real interest rate differentials.

V. Concluding Remarks

The model provides two major insights into the nature of financial intermediation in developing countries. First, the model suggests that, in equilibrium, monitoring of loans requires a mixed strategy. This does not change the argument for financial intermediation, it only alters the expected costs of monitoring and the formulation of the problem. Second, the model suggests that a sufficiently high probability of a common external shock precludes intermediaries from entering less diversified areas, which are typically located in less developed countries. Informal arrangements may arise if they provide a method of tying payments to project outcomes or if they reduce monitoring costs. Rotating savings clubs are one example of such arrangements. Deposit insurance is also found in these clubs, because the organizer may guarantee payment of dues. The model also suggests that formal financial intermediaries may co-exist with informal arrangements if the two entities have different monitoring costs. The informal arrangements may service borrowers that generate high monitoring costs for formal intermediaries.

Further research should be fruitful. For example, in the informal clubs, deposit insurance does not insure against external shocks, as in the above model, but insures against delinquent members. Presumably, the latter problem may be solved by the law of large numbers if the club were big enough. We conjecture that the clubs are not bigger because they could not effectively monitor members during meetings, arrange a convenient meeting time, or renegotiate with delinquent members.


[1.] Basar, Tamer and Geert J. Olsder. Dynamic Noncooperative Game Theory. New York: Academic Press, 1982. [2.] Bouman, Frits J. A., "Indigenous Savings and Credit Societies in the Third World: A Message." Saving and Development, 1, No. 4, 1977, 181-209. [3.] _____, "The ROSCA, Financial Technology of an Informal Savings and Credit Institution in Developing Countries " Saving and Development, 3, No. 4, 1979, 253-72. [4.] Diamond, Douglas, "Financial Intermediation and Delegated Monitoring." Review of Economic Studies, July 1984, 393-414. [5.] _____ and Philip Dybvig, "Bank Runs, Deposit Insurance, and Liquidity." Journal of Political Economy, June 1983, 401-19. [6.] Federal Deposit Insurance Corporation. Annual Report. Washington, D.C.: U.S. Government Printing Office, 1950. [7.] Fishe, Raymond P. H. "Opportunistic Behavior With Imperfect Monitoring: A Game-Theoretic Approach." Monograph, University of Miami, 1985. [8.] Geertz, Clifford, "The Rotating Credit Association: A "Middle Rung" in Economic Development." Economic Development and Cultural Change, April 1962, 241-63. [9.] Ghate, P. B., "Informal Credit Markets in Asian Developing Countries." Asian Development Review, 6, No. 1, 1988,64-85. [10.] Kareken, John H., "Federal Bank Regulatory Policy: A Description and Some Observations." The Journal of Business, January 1986, 3-48. [11.] _____ and Neil Wallace, "Deposit Insurance and Bank Regulation: A Partial Equilibrium Analysis." The Journal of Business, July 1978, 413-38. [12.] Klein, Benjamin and Keith Leffler, "The Role of Market Forces in Assuring Contractual Performance." Journal Political Economy, August 1981, 615-41. [13.] McKinnon, Ronald. Money and Capital in Economic Development. Washington, D.C.: The Brookings Institution, 1973. [14.] McCarthy, Ian S., "Deposit Insurance: Theory and Practice." IMF Staff Papers, September 1980, 578-600. [15.] Osuntogun, Adeniyi and Remi Adeyemo, "Mobilization of Rural Savings and Credit Extension by Pre-Cooperative Organizations in South Western Nigeria." Saving and Development, 5, No. 4, 1981, 247-61. [16.] Shaw, Edward. Financial Deepening in Economic Development. London: Oxford University Press, 1973. [17.] United Nations. Commodity Yearbook. New York: United Nations, 1991. [18.] Waldo, Douglas G., "Bank Runs, the Deposit-Currency Ratio and the Interest Rate." Journal of Monetary Economics, May 1985, 269-78. [19.] Wallace, Neil. "The Overlapping Generations Model of Fiat Money," in Models of Monetary Economics, edited by John H. Kareken and Neil Wallace. Minneapolis: Federal Reserve Bank of Minneapolis, 1980, 49-82. [20.] Williamson, Stephen D., "Costly Monitoring, Financial Intermediation, and Equilibrium Credit Rationing." Journal of Monetary Economics, September 1986, 159-79. [21.] _____, "Recent Developments in Modeling Financial Intermediation." Federal Reserve Bank of Minneapolis Quarterly Review, Summer 1987, 19-29. [22.] _____, "Liquidity, Banking and Bank Failures." International Economic Review, February 1988, 25-44.

(1.) The Federal Deposit Insurance Corporation provides an earlier history of deposit insurance legislation [6]. (2.) State backed insurance schemes existed until the bank runs in Ohio and Maryland in 1985 illustrated their lack of adequate capital. (3.) McCarthy [14] reviews the deposit insurance schemes found in thirteen other countries. (4.) The Taxpayer Compliance Program randomly selects a small number of corporate and individual tax returns for extensive audits. These audits are used to develop screening procedures that are computerized and applied to all other tax returns. (5.) Basar and Olsder [1] provide a discussion of Nash equilibria, pure and mixed strategies, and noncooperative game theory. (6.) Kareken [10] offers a thorough discussion of bank regulation and deposit insurance including the arguments both for and against deposit insurance and a detailed legislative history. (7.) Klein and Leffler [12] develop a model in which investments in specific assets serve as collateral to consumers in a market where product quality is not observed until after a purchase is made. The role of advertising as a specific asset is discussed in this model. For example, car dealerships often have personalized signs that are of no value if the dealership closes. (8.) Kareken and Wallace [11] also argue that deposit insurance is unnecessary for bank safety. Safety may be achieved without deposit insurance if depositors know the riskiness of a bank's portfolio and bankruptcy is costly. (9.) It is interesting to note that two schemes were considered in the 1930s for insuring deposits. The first scheme offered to insure all depositors, regardless of their account size, and guarantee to repay a fraction (80%) of the amount deposited. The second scheme, which was adopted, insured all depositors up to $10,000. Depositors with amounts exceeding $10,000 became creditors for the overage. Under the second scheme, depositors could insure all of their deposits by distributing them across several banks, thus offering a greater potential for producing inflation. If depositors must wait for their funds, however, then inflation effectively may make the second scheme equivalent to the first. (10.) Fiat money is introduced into the economy in the initial period of life for third party enforcement of aff contracts. (11.) A model in which the payoff is different and the probability of success is held constant is a straightforward generalization of the present model. (12.) The law of large numbers may not always insure the intermediary against default. If there is an exogenous factor that affects all returns, then the intermediary structure may not be a stable equilibrium (see below), (13.) The entrepreneur's level of consumption in period t + 1 is solely determined by the net return to the investment project. Thus, maximizing the expected net return is the same as maximizing expected consumption. (14.) Fishe [7] provides an analysis of the case of imperfect monitoring. (15.) Williamson [21] argues that the intermediary will select a pure strategy solution and monitor when the entrepreneur announces a bad outcome. If this is the case, the entrepreneur will never misrepresent the project's final outcome. But if the agent does not misrepresent the outcome, there is no incentive to monitor. Fishe [7] shows that this approach does not generate a Nash equilibrium. (16.) Time and savings deposits, inflation rates, and gross domestic product are collected from the International Monetary Fund's annual publication, International Financial Statistics. The time deposit rate is collected from the World Development Report, 1990, published by the World Bank. Per capita gross national product denominated in U.S. dollars is collected from World Tables, 1989-90, which is also published by the World Bank. (17.) Our initial sample contained forty-five countries, but data on time deposit rates are available for only twenty-nine of these countries. The twenty-nine countries are Nigeria, Niger, Zambia, Morocco, Uganda, Liberia, Mauritius, Malawi, Mali, Cameroon, Tunisia, Chad, Cote d'Ivoire, Honduras, Nicaragua, Jamaica, Bolivia, Chile, Guatemala, Mexico, Argentina, Uruguay, Haiti, Brazil, Indonesia, Sri Lanka, Bangladesh, Malaysia, and Kenya. (18.) The real interest rate is not entered in logarithmic form because for some countries its value is negative.
COPYRIGHT 1993 Southern Economic Association
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1993, Gale Group. All rights reserved. Gale Group is a Thomson Corporation Company.

Article Details
Printer friendly Cite/link Email Feedback
Author:Fishe, Raymond P.H.
Publication:Southern Economic Journal
Date:Jul 1, 1993
Previous Article:Income, risk aversion, and the demand for insurance.
Next Article:An economic analysis of recidivism among drug offenders.

Related Articles
Statement to Congress.
Statements to Congress.
Statements to the Congress.
Statements to Congress.
Statements to the Congress.
Banking on markets.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters