4 Current theories of financial intermediation
Current theories of the economic role of financial intermediaries build on the economics of imperfect information that began to emerge during the 1970s with the seminal contributions of Akerlof (1970), Spence (1973) and Rothschild and Stiglitz (1976). Financial intermediaries exist because they can reduce information and transaction costs that arise from an information asymmetry between borrowers and lenders. Financial intermediaries thus assist the efficient functioning of markets, and any factors that affect the amount of credit channelled through financial intermediaries can have significant macroeconomic effects.
There are two strands in the literature that formally explain the existence of financial intermediaries. The first strand emphasises financial intermediaries’ provision of liquidity. The second strand focuses on financial intermediaries’ ability to transform the risk characteristics of assets. In both cases, financial intermediation can reduce the cost of channelling funds between borrowers and lenders, leading to a more efficient allocation of resources.
Diamond and Dybvig (1983) analyse the provision of liquidity (the transformation of illiquid assets into liquid liabilities) by banks. In Diamond and Dybvig’s model, ex ante identical investors (depositors) are risk averse and uncertain about the timing of their future consumption needs. Without an intermediary, all investors are locked into illiquid long-term investments that yield high payoffs only to those who consume late. Those who must consume early receive low payoffs because early consumption requires premature liquidation of long-term investments. Banks can improve on a competitive market by providing better risk sharing among agents who need to consume at different (random) times. An intermediary promising investors a higher payoff for early consumption and a lower payoff for late consumption relative to the non-intermediated case enhances risk sharing and welfare.
The optimal insurance contract in Diamond and Dybvig’s model is a demand deposit contract, but it has an undesirable equilibrium (bank run), in which all depositors panic and withdraw immediately, including even those who would prefer to leave their deposits in the bank if they were not concerned about the bank failing. Bank runs cause real economic problems because even “healthy” banks can fail, leading to a recall of loans and the termination of productive investment.
In Diamond and Dybvig (1983) the illiquidity of assets provides both the rationale for the existence of banks and for their vulnerability to runs. A bank run is caused by a shift in expectations. When normal volumes of withdrawals are known and not stochastic, suspension of convertibility of deposits will allow banks both to prevent bank runs and to provide optimal risk sharing by converting illiquid assets into liquid liabilities. In the more general case (with stochastic withdrawals), deposit insurance can rule out runs without reducing the ability of banks to transform assets.
Financial intermediaries are able to transform the risk characteristics of assets because they can overcome a market failure and resolve an information asymmetry problem. Information asymmetry in credit markets arises because borrowers generally know more about their investment projects than lenders do. The information asymmetry can occur “ex ante” or “ex post”. An ex ante information asymmetry arises when lenders cannot differentiate between borrowers with different credit risks before providing loans and leads to an adverse selection problem. Adverse selection problems arise when an increase in interest rates leaves a more risky pool of borrowers in the market for funds. Financial intermediaries are then more likely to be lending to high-risk borrowers, because those who are willing to pay high interest rates will, on average, be worse risks.
The information asymmetry problem occurs ex post when only borrowers, but not lenders, can observe actual returns after project completion. This leads to a moral hazard problem. Moral hazard arises when a borrower engages in activities that reduce the likelihood of a loan being repaid. An example of moral hazard is when firms’ owners “siphon off” funds (legally or illegally) to themselves or to associates, for example, through loss-making contracts signed with associated firms.
The problem with imperfect information is that information is a “public good”. If costly privately-produced information can subsequently be used at less cost by other agents, there will be inadequate motivation to invest in the publicly optimal quantity of information (Hirschleifer and Riley, 1979). The implication for financial intermediaries is as follows. Once banks obtain information they must be able to signal their information advantage to lenders without giving away their information advantage. One reason, financial intermediaries can obtain information at a lower cost than individual lenders is that financial intermediation avoids duplication of the production of information. Moreover, there are increasing returns to scale to financial intermediation. Financial intermediaries develop special skills in evaluating prospective borrowers and investment projects. They can also exploit cross-sectional (across customers) information and re-use information over time.
Leland and Pyle (1977) formally show that a bank can communicate information to investors about potential borrowers at a lower cost than can individual borrowers. They focus on an ex ante information asymmetry, where entrepreneurs selling shares to the market know the expected returns of their own investment, but other agents find this information costly to observe. This results in a moral hazard problem since firms with low expected returns have an incentive to claim a high expected return so as to increase their market valuation. In Leland and Pyle’s model intermediaries can solve this moral hazard problem by monitoring the actions of firms.
Retained equity can serve as a costly signal of the entrepreneur’s information about a project. The value of the firm increases with the share of the firm held by the entrepreneur, and in contrast to Modigliani and Miller (1958), the financial structure of a firm is therefore related to project or firm value.
One problem with the Leland and Pyle analysis is that it assumes the existence of an incentive signalling equilibrium. However, as Campbell and Kracaw (1980) note, if a signalling equilibrium exists, then firms will be properly valued with or without intermediaries or other information producers.
Diamond (1984) argues that diversification within the financial intermediary is the main reason financial intermediaries exist. He also develops a model in which the outcome from firms’ investment project is not known ex post to external agents, unless information is gathered to assess the outcome, i.e. there is “costly state verification” (Townsend, 1979). This leads to a moral hazard problem because it provides an incentive for borrowers to default on a loan even when the project is successful.
In Diamond’s model, intermediaries are delegated the costly task of monitoring loan contracts. A financial intermediary must choose an incentive contract such that it has incentives to monitor the information, make proper use of it, and make sufficient payments to depositors to attract deposits. Providing these incentives is costly and diversification can reduce these costs.
The optimal contract is a debt contract (an agreement by the borrower to pay the lender a fixed amount) with a non-pecuniary bankruptcy penalty. The intermediary need not be monitored because it bears all penalties for any shortfall of payments. This is because the diversification of the intermediary’s portfolio makes the probability of incurring these penalties very small. The optimal size for a financial intermediary is infinite; costs are lowered indefinitely by diversification, as long as the returns to entrepreneurs are not perfectly correlated.
Adverse selection increases the likelihood that loans will be made to bad credit risks, while moral hazard lowers the probability that a loan will be repaid. As a result, lenders may decide in some circumstances that they would rather not make a loan and credit rationing may occur. There are two forms of credit rationing: (i) some loan applicants may receive a smaller loan than they applied for at the given interest rate, or (ii) they may not receive a loan at all, even if they offered to pay a higher interest rate.
Jaffee and Russell (1976) develop a theoretical model in which imperfect information and uncertainty can lead to rationing in loan markets, where some agents do not receive the loan they applied for. Their paper analyses the behaviour of a loan market in which borrowers have more information than lenders about the likelihood of default. The key feature in the model is the relationship between default proportions and contract sizes. There is some minimum loan size at which no default is observed, beyond that, the proportion of individuals who do not default is declining with the contract size.
Since borrowers are identical ex ante, the market interest rate incorporates a premium to take account of the aggregate probability of default. Consequently, borrowers with low default probability pay a premium to support low quality borrowers and credit rationing in the form of the supply of smaller-sized loans than those demanded by the borrowers at a quoted rate may result. High quality borrowers will prefer some rationing if the smaller loan sizes lower the market average default probabilities, thus reducing the premium.
Stiglitz and Weiss (1981) develop a model of credit rationing, where some borrowers receive loans and others do not. They assume that the interest rate directly affects the quality of loans because of an adverse selection effect or moral hazard effect. Banks making loans are concerned about the interest rate they receive on a loan, and the riskiness of the loan. For a given loan rate, lenders earn a lower expected return on loans to borrowers with riskier projects than to good quality borrowers.
The interest rate a bank charges can affect the riskiness of the loans by either sorting prospective borrowers (the adverse selection effect), or by affecting the actions of borrowers (the moral hazard effect). When the price (interest rate) affects the transaction, it may not clear the market. The adverse selection effect of interest rates is a consequence of different borrowers having different probabilities of repaying their loans. The interest rate an individual is willing to pay may act as a screening device. Those who are willing to pay high interest rates may, on average, be worse risks. They are willing to borrow at high interest rates because they perceive their probability of repaying the loan to be low. As a result there exists an interest rate that maximises the expected return to the bank and beyond which the bank will be unwilling to supply funds, making the supply of loans curve bend backwards.
A change in interest rates can affect the bank’s expected return from loans through the moral hazard effect by changing the behaviour of borrowers. Higher interest rates induce firms to undertake projects with lower probabilities of success but higher payoffs when successful. Increasing the rate of interest increases the relative attractiveness of riskier projects, for which the return to the bank may be lower. As the interest rate rises, the average riskiness of those who borrow increases and the moral hazard effect reinforces the adverse selection problem. Banks therefore have an incentive, in some circumstances, to ration credit rather than to raise interest rates when there is excess demand for loanable funds.
Williamson (1986) develops a model of credit rationing where borrowers are subject to a moral hazard problem. Borrowers are identical ex ante, but some receive loans and others do not. A borrower and lender are asymmetrically informed ex post about the return on the borrower’s investment project, and the borrower will have an incentive to falsely default on the loan. Costly monitoring by lenders of borrowers together with large-scale investment projects imply that there exist increasing returns to scale in lending and borrowing which can be exploited by financial intermediaries. The optimal contract between a lender and a borrower is a debt contract and the lender only monitors in the event of default.
An increase in the loan interest rate raises the expected return to the lender, but also results in an increase in the probability that the borrower defaults, thus increasing the expected cost of monitoring to the lender. This, in turn, generates an asymmetry in the borrowers’ and lenders’ payoff functions, which can lead to credit rationing. Because of the asymmetry in the payoff functions it may not be possible for the loan interest rate to adjust to clear the market, so that some borrowers do not receive a loan in equilibrium.
In summary, financial intermediaries play an important role in credit markets because they reduce the cost of channelling funds between relatively uninformed depositors to uses that are information-intensive and difficult to evaluate, leading to a more efficient allocation of resources. Intermediaries specialise in collecting information, evaluating projects, monitoring borrowers’ performance and risk sharing. Despite this specialisation, the existence of financial intermediaries does not replicate the credit market outcomes that would occur under a full information environment. The existence of imperfect, asymmetrically-held information causes frictions in the credit market. Changes to the information structure and to variables which may be used to overcome credit frictions (such as firm collateral and equity) will in turn cause the nature and degree of credit imperfections to alter.
Banks and other intermediaries are “special” where they provide credit to borrowers on terms which those borrowers would not otherwise be able to obtain. Because of the existence of economies of scale in loan markets, small firms in particular may have difficulties obtaining funding from non-bank sources and so are more reliant on bank lending than are other firms. Adverse shocks to the information structure, or to these firms’ collateral or equity levels, or to banks’ ability to lend, may all impact on firms’ access to credit and hence to investment and output.
- When markets do not operate costlessly, firms arise if they can reduce market transaction costs by organising resources more cheaply within the firm (Coase, 1937).
- The vulnerability to bank runs in the Diamond and Dybvig (1983) model has stimulated a lengthy debate in the literature on prudential regulation. See, for example, Bhattacharya, Boot and Thakor (1998), Dewatripont and Tirole (1994).
- Under the assumption that banks cannot select the risk of their loan portfolios, a central bank as a lender of last resort could provide a service similar to deposit insurance. However, when there is a trade-off between optimal risk and proper incentives for portfolio choice, the lender of last resort can no longer be as credible as deposit insurance. If the lender of last resort were always required to bail out banks with liquidity problems, there would be perverse incentives for banks to take on risk. Deposit insurance on the other hand is a binding commitment that, in theory, can be structured to retain punishment in the case of bank runs. See Demirgüç-Kunt and Kane (2002) for difficulties in implementing deposit insurance in practice.
- Leland and Pyle (1977) assume that investors cannot infer intermediaries’ information by observing their portfolios.
- When project returns are not independently distributed and instead depend on several common factors that are observable (such as economic conditions, interest rates, and input prices) the intermediary still monitors firm-specific information, but hedges out all systematic risks.
- Stiglitz and Weiss (1981) assume that heterogeneity among entrepreneurs arises from different probability distributions of returns to their projects.
- This occurs because an unobserved mean-preserving spread in a borrower’s project return distribution reduces the expected payment to lenders under default (Rothschild and Stiglitz, 1970).
- Unlike in Diamond (1984) monitoring decisions are made ex post and the probability that monitoring occurs is determined endogenously.