Asymmetric Information, Financial Intermediation and the Monetary Transmission Mechanism: A Critical Review

Claus, Iris; Grimes, Arthur Working Paper Asymmetric Information, Financial Intermediation and the Monetary Transmission Mechanism: A Critical Review New Zealand Treasury Working Paper, No. 03/19 Provided in Cooperation with: The Treasury, New Zealand Government Suggested Citation: Claus, Iris; Grimes, Arthur (2003) : Asymmetric Information, Financial Intermediation and the Monetary Transmission Mechanism: A Critical Review, New Zealand Treasury Working Paper, No. 03/19, New Zealand Government, The Treasury, Wellington This Version is available at: https://hdl.handle.net/10419/205524 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Asymmetric Information, Financial Intermediation and the Monetary Transmission Mechanism: A Critical Review Iris Claus and Arthur Grimes N EW Z EALAND T REASURY W ORKING P APER 03/19 S EPTEMBER 2003 NZ TREASURY WORKING PAPER 03/19 Asymmetric Information, Financial Intermediation and the Monetary Transmission Mechanism: A Critical Review MONTH / YEAR September 2003 AUTHORS Iris Claus The Treasury PO Box 3724 Wellington NEW ZEALAND Email Telephone [email protected] (64) (4) 471 5221 Arthur Grimes Victoria University of Wellington PO Box 600 Wellington NEW ZEALAND Email Telephone [email protected] (64) (4) 463 6834 ACKNOWLEDGEMENTS We would like to thank Bob Buckle, John Creedy, Stephen Burnell, Khoon Goh, Leo Krippner, Brendon Riches and Christie Smith for valuable comments on (earlier drafts of) this paper. NZ TREASURY New Zealand Treasury PO Box 3724 Wellington 6008 NEW ZEALAND Email Telephone Website [email protected] 64-4-472 2733 www.treasury.govt.nz DISCLAIMER The views expressed in this Working Paper are those of the author(s) and do not necessarily reflect the views of the New Zealand Treasury. The paper is presented not as policy, but with a view to inform and stimulate wider debate. WP 03/19 ASYMMETRIC INFORMATION, FINANCIAL INTERMEDIATION AND THE MONETARY TRANSMISSION MECHANISM: A CRITICAL REVIEW i Abstract Macroeconomic models currently used by policy makers generally assume that the workings of financial markets can be fully summarised by financial prices, because the Modigliani and Miller (1958) theorem holds. This paper argues that these models are too limited in describing how monetary policy (and other) shocks are transmitted to the economy and points to new directions. The models are too limited because they disregard an information asymmetry between borrowers and lenders and the importance of financial intermediaries not only for individual depositors but the economy as a whole. Incorporating financial market interactions into macroeconomic models will enhance the understanding of the transmission mechanisms of monetary policy and other shocks. JEL CLASSIFICATION E32 (prices, business cycle fluctuations, and cycles) E44 (financial markets and the macroeconomy) E50 (monetary policy, central banking, and the supply of money and credit) KEYWORDS Financial intermediaries; credit channel; monetary transmission mechanism; open economies WP 03/19 ASYMMETRIC INFORMATION, FINANCIAL INTERMEDIATION AND THE MONETARY TRANSMISSION MECHANISM: A CRITICAL REVIEW ii Table of Contents Abstract ...............................................................................................................................i Table of Contents ..............................................................................................................ii 1 Introduction ..............................................................................................................1 2 Interest and exchange rate channels .....................................................................2 3 Traditional literature of financial intermediation...................................................5 4 Current theories of financial intermediation..........................................................9 5 The credit channel..................................................................................................13 6 General equilibrium models of the credit channel..............................................16 7 Summary and conclusions....................................................................................19 References .......................................................................................................................21 WP 03/19 ASYMMETRIC INFORMATION, FINANCIAL INTERMEDIATION AND THE MONETARY TRANSMISSION MECHANISM: A CRITICAL REVIEW 1 Asymmetric Information, Financial Intermediation and the Monetary Transmission Mechanism: A Critical Review 1 Introduction To conduct monetary and fiscal policies successfully, policy makers must have an accurate assessment of the timing and effects of their policies on the economy. This includes an understanding of the monetary transmission mechanisms through which monetary policy affects the decisions of firms, households, financial intermediaries and investors that alter the level of economic activity and prices. This paper argues that macroeconomic models currently used by policy makers are too limited in describing the transmission mechanisms and points to new directions. Interest and exchange rates are the prototypical channels through which monetary policy affects the economy in contemporary models used by policy makers.1 The assumption in these models is that the workings of financial markets can be fully summarized by financial prices, because the Modigliani and Miller (1958) theorem holds. Under the assumptions that financial markets are complete and information and transaction costs are non-existent, the Modigliani and Miller (1958) theorem states that the mix of debt and equity used to finance firms’ expenditures does not affect the expected profitability of the project – the same investment decisions would be made, irrespective of the mix of debt and equity finance. Fama’s (1980) extension of the Modigliani-Miller theorem to the entire financial system allows the abstraction from considerations of credit market conditions in macroeconomic models. While the complete market assumption remains important in economics, the assumption of zero information and transaction costs (or perfect information) has come under increasing criticism since Akerlof’s (1970) seminal paper, which illustrated how imperfect information between buyers and sellers can cause market malfunctioning. With imperfect information, the market price reflects buyers’ perception of the average quality of the 1 See, for example, the Reserve Bank of New Zealand’s Forecasting and Policy System (Black, Cassino, Drew, Hansen, Hunt, Rose and Scott, 1997), the New Zealand Treasury NZTM model (Szeto, 2002), the Federal Reserve Board of Governors’ FRB/US model (Brayton and Tinsley, 1996), the International Monetary Fund’s MULTIMOD model for industrial countries (Laxton, Isard, Faruqee, Prasad and Turtelboom, 1998), the Bank of Canada’s Quarterly Projection Model (Black, Laxton, Rose and Tetlow, 1994) or the Australian Treasury’s TRYM model (Commonwealth Treasury 1996a, 1996b). WP 03/19 ASYMMETRIC INFORMATION, FINANCIAL INTERMEDIATION AND THE MONETARY TRANSMISSION MECHANISM: A CRITICAL REVIEW 2 product being sold, and sellers of low quality products will receive a premium at the expense of those selling high quality goods. As a result, some high quality sellers will stay out of the market, which will lower the average quality of the product and price of the product even further, leading more high quality sellers to stay out of the market. The process will continue and may preclude the market from actually opening. Efficient markets require some mechanisms for overcoming the imperfect information problem. In financial markets, an information asymmetry arises between borrowers and lenders because borrowers generally know more about their investment projects than lenders do. Intermediaries, which specialise in collecting information, evaluating projects and borrowers, and monitoring borrowers’ performance, can help overcome the information problem. Financial intermediaries thus exist because there are information and transactions costs that arise from imperfect information between borrowers and lenders. This implies that the assumptions upon which the Modigliani-Miller theorem is based, and thus the macroeconomic models used by policy makers, do not hold. Conditions in financial and credit markets can affect the real economy; and interest and exchange rates are an incomplete description of the monetary transmission mechanism. The remainder of the paper proceeds in six further sections. Section 2 describes the standard interest and exchange rate channels of the monetary transmission mechanism as they are typically incorporated in macroeconmic models. These channels are too limited as they abstract from credit market interactions. The idea that the credit creation process can have real economic effects is not new and section 3 reviews traditional theories of the role of credit markets including Wicksell’s early writings on monetary dynamics and Fisher’s (1933) “debt-deflation theory of great depressions”. Current theories are discussed in section 4 and the credit channel of the monetary transmission mechanism as it is currently characterised in the literature is described in section 5. A number of dynamic general equilibrium models that account for an explicit role of credit market frictions in business cycle fluctuations have been developed recently. These are discussed in section 6. None of the models is complete and section 7 summarises and concludes with a brief assessment of avenues for future research. 2 Interest and exchange rate channels 2 A central bank derives the power to influence wholesale money market interest rates from the fact that it is the monopoly supplier of high-powered money (also known as outside money or the monetary base). Although the institutional details differ from country to country, the operating procedure of central banks is generally similar. Central banks choose the price at which they lend high-powered money to the inter-bank market. The quantitative effect of a change in the official central bank lending rate on other interest rates, and on financial market conditions more generally, depends on the extent to which a policy change is anticipated, how the change affects expectations of future policy, interest rates and inflation, and the degree of nominal price rigidities. Real (inflation expectations adjusted) interest rates reflect the opportunity cost of current expenditure relative to expenditure in some future period, since earning a return in the interim enables greater consumption or investment at a later date. Changes in real interest rates alter the incentives to consume and invest in the present versus consuming and investing in future. 2 The channels described in this section reflect those typically incorporated in macroeconomic models of the type listed in footnote 1. Specific references to these and other sources are therefore not noted separately in the discussion that follows. WP 03/19 ASYMMETRIC INFORMATION, FINANCIAL INTERMEDIATION AND THE MONETARY TRANSMISSION MECHANISM: A CRITICAL REVIEW 3 Movements in short-term interest rates affect longer-term interest rates, at which financial institutions typically lend to and borrow from businesses and consumers, as illustrated by the expectations hypothesis of the term structure of interest rates. The expectations hypothesis states that, for any choice of holding period, the expected return is the same for any combination of bonds of different maturities. For example, the rate of return from holding a one-year note should be the same as holding two successive six-month notes.3 Real interest rates more closely reflect the costs and benefits of deferring expenditure than nominal interest rates as they account for the loss of purchasing power due to inflation. The link between nominal rates, set by the central bank and financial intermediaries, and real interest rates is through the Fisher equation. The Fisher equation decomposes the observed nominal interest rate into an expected inflation component and an expected real rate and can be written as 1(1)(1 )   ,+ ,+ ,+  +=+ ⋅+Π tt k tt k t tt k irE (1) where all values are in discrete time, ,+tt k i denotes the nominal rate of return on lending (or the cost of borrowing) from time t to time +tk, ,+tt k r is the ex ante real lending (borrowing) rate from time t to time +tk and     ,+   Π tttk E refers to the expected rate at time t of inflation between time t and +tk.4 An increase in short-term nominal interest rates will lead to higher short-term ex ante real rates if inflation expectations do not adjust upwards by the full increase in short-term interest rates and will result in higher short-term ex post real rates if there is price stickiness.5 Inflation inertia and price stickiness can arise from slow adjustment of expectations, the existence of explicit or implicit contracts that are not indexed to the rate of inflation, menu costs of adjusting prices, real rigidities, and price and wage staggering. Higher real rates directly reduce the profitability of investment projects because of higher financing costs, and indirectly because of the prospect of a slowdown in consumption. A rise in interest rates tends to encourage households to reduce current consumption because the return on saving and the cost of borrowing to finance consumption both increase. Monetary policy may have an additional effect on current consumption by lowering the disposable income of households who are borrowers. A portion of households’ spending is on servicing debt interest. Increases (decreases) in interest rates will raise (lower) the amount of debt servicing required and lower (increase) disposable income. The opposite holds true for households who are savers and the overall net effect on consumption will depend on how much the change in consumption of borrowers is offset by that of savers.6 In an open economy, monetary policy also affects output and inflation through the influence that interest rates have on the exchange rate. Interest parity arguments imply that movements in domestic interest rates will induce movements in the exchange rate. Uncovered interest parity, for example, states that the expected return on a bond 3 Uncertainty regarding future short-term interest rates coupled with risk aversion may cause the term structure to deviate from the shape implied by the risk-neutral expectations hypothesis. This deviation is captured by a term premium. 4 This formulation of the Fisher equation abstracts from taxes and, as Svensson (1985) notes, the simple Fisher relation does not hold under uncertainty. 5 It is not frequently recognised that an increase in short-term nominal interest rates will, ceteris paribus, have a neutral effect on the real economy only if prices adjust downwards in a stepwise manner, enabling subsequent price inflation to occur over the period relevant to that interest rate. 6 In the case of fixed term interest rates, changes in interest rates will affect consumption if households believe that they eventually will have to face the new interest rate and alter their consumption behaviour in anticipation. WP 03/19 ASYMMETRIC INFORMATION, FINANCIAL INTERMEDIATION AND THE MONETARY TRANSMISSION MECHANISM: A CRITICAL REVIEW 4 denominated in a foreign currency should be the same as the expected return from holding an otherwise identical domestic currency bond. A differential between domestic and foreign interest rates reflects expected movements in the exchange rate. A change in monetary policy that leads to movements in the differential between domestic and foreign interest rates therefore implies (assuming uncovered interest rate parity) that either the exchange rate or the expected exchange rate or both must change. Exchange rate movements have a direct impact on the cost of imports and domestic inflation.7 Exchange rate movements also have an indirect impact on inflation through their impact on the demand and supply of tradeable and non-tradeable goods and services. An appreciation of the exchange rate, for example, will lower the rate of growth of the domestic price level, which is a weighted average of tradeables and non-tradeables prices. This is because an appreciation of the exchange rate decreases the domestic price of tradeables (if tradeables prices are determined in foreign currency units by world markets). A decrease in the price of tradeables relative to non-tradeables increases the domestic demand for tradeables (and lowers the domestic supply of tradeable goods and services).8 Moreover, an appreciation of the exchange rate (or decline in the relative price of tradeables to non-tradeables) reduces the domestic demand for non-tradeables and increases the domestic supply of non-tradeables, causing excess supply of nontradeables and a reduction in the price of non-tradeables. Monetary policy affects asset prices more generally. For example, monetary policy directly affects the market value of future cash flows through its effect on the discount factor. The relationship between the market price of a unit future cash flow and interest rates is exp( ) ,+ =−⋅ tt k t t PV i k (2) where ,+tt k PV is the present value at time t of a cash flow that matures at time +tk, t i denotes the continuously compounding interest rate at annual rates at time t and t k is the number of years to maturity at time t. The present value (or market price) of the cash flow is inversely related to the yield, i.e. a rise in t i lowers the present value of the cash flow and a decline in t i raises it. This result generalises to all assets (bonds, equities, property, etc.) because the value of all assets can be defined as a combination of expected future cash flows. Tobin’s q theory provides a mechanism through which monetary policy affects the economy through its effects on the valuation of equities. Tobin (1969) defines q as the market value of firms divided by the replacement cost of capital. The market price of firms will increase with an easing in monetary policy. Tobin’s q rises if this market price of firms increases relative to the replacement cost of capital, i.e. if the cost of new plant and equipment capital declines relative to the market value of firms. Investment spending will rise because firms can purchase new investment goods, which will be valued in the equity market at greater than their purchase cost.9 On the other hand, when q is low, firms will not purchase new investment goods because the market value of firms is low relative to the cost of capital. If companies want to acquire capital when q is low, they can buy another firm cheaply and acquire old capital instead. As a result, investment spending will be low. 7 See, for example, Adolfson (2002) and Smets and Wouters (2002). 8 The increase in demand for tradeables is reflected in a deterioration in the current account deficit of the balance of payments (matched by a rise in the net foreign demand for domestic assets). 9 Here we are referring to marginal q rather than average q since it is the valuation of the marginal investment project which is relevant (Hayashi, 1982). WP 03/19 ASYMMETRIC INFORMATION, FINANCIAL INTERMEDIATION AND THE MONETARY TRANSMISSION MECHANISM: A CRITICAL REVIEW 11 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.20 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. 20 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. WP 03/19 ASYMMETRIC INFORMATION, FINANCIAL INTERMEDIATION AND THE MONETARY TRANSMISSION MECHANISM: A CRITICAL REVIEW 12 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.21 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.22 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.23 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 21 Stiglitz and Weiss (1981) assume that heterogeneity among entrepreneurs arises from different probability distributions of returns to their projects. 22 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). 23 Unlike in Diamond (1984) monitoring decisions are made ex post and the probability that monitoring occurs is determined endogenously. WP 03/19 ASYMMETRIC INFORMATION, FINANCIAL INTERMEDIATION AND THE MONETARY TRANSMISSION MECHANISM: A CRITICAL REVIEW 13 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. 5 The credit channel The credit channel literature examines the impact of asymmetric information and other credit market frictions on real spending and economic activity, with resulting implications for monetary policy.24 The bank lending channel analyses the impact of monetary policy on the supply of loans by depository institutions, and the balance sheet (or financial accelerator) effect focuses on the potential impact of monetary policy on firms’ balance sheets and their ability to borrow. The credit channel also operates when shifts in monetary policy alter either the efficiency of financial markets in matching borrowers and lenders or the extent to which borrowers face rationing in credit markets. With credit rationing, monetary policy may have real effects without changing interest rates in lending markets. Monetary policy can have an impact on the supply of intermediated credit, which in most countries is predominantly provided by banks. A bank is a financial intermediary that participates in the payment system and finances entities in financial deficit, generally the public sector, firms and some households, using the funds of entities in financial surplus, typically households. The reliance on bank credit is probably declining overall as corporations and, in particular, large businesses turn to the securities markets to meet their funding needs. However, an important fraction of firms, mainly small firms, are likely to remain bank-dependent at least in the near future (Trautwein, 2000). Moreover, banks are a critical source of liquidity even for large firms during times of economic stress (Saidenberg and Strahan, 1999). The asset side of banks’ balance sheets consists of loans to the public sector, firms and households. The liability side includes deposits by households and firms plus banks’ equity. Banks’ equity consists of stock issues and retained earnings. Banks typically “borrow short” and “lend long”: they take deposits that can be withdrawn on demand or in a matter of months and make loans that often are only re-paid over periods of years. As a result, banks’ assets tend to have longer maturity than their liabilities. Monetary policy can have an impact on banks’ balance sheets because of this maturity mismatch of assets 24 See Bernanke and Gertler (1995), Mishkin (1995) and Trautwein (2000) among others. WP 03/19 ASYMMETRIC INFORMATION, FINANCIAL INTERMEDIATION AND THE MONETARY TRANSMISSION MECHANISM: A CRITICAL REVIEW 14 and liabilities. In the absence of perfect interest rate hedging, a tightening in monetary policy, i.e. an increase in interest rates, tends to cause a larger decline in the present value of the assets with longer maturity than the liabilities with shorter maturity. Conversely, a decline in interest rates causes a larger gain in the present value of the assets than the liabilities. For example, the decline in the market price of a unit future cash flow (see equation 2) due to an increase in t i is a function of t k, the number of years to maturity exp( ) ,+ ∂=− ⋅ − ⋅ ∂ tt k ttt t PV kik i (3) By construction, because the asset side of a balance sheet equals the liability side, a tightening in monetary policy that leads to a larger decline in the value of loans than the value of deposits, implies a contraction in banks’ value of equity. If banks are required by regulators or depositors to retain some minimum capital ratio (defined as the value of banks’ equity as a percent of the value of loans outstanding), they will have to either reduce their supply of loanable funds or raise new equity. However, because equity takes time to raise and also because the cost of new capital has increased due to higher interest rates and a lower market value of banks, the typical initial response is a contraction in lending.25 A reduction in the supply of bank loans increases the financing cost, or reduces the financing, of firms that are dependent on banks for credit. Bank-dependent firms are typically smaller in size (Gertler and Gilchrist, 1994). These firms tend to be bankdependent because their access to (non-bank) capital markets is poor, because of reduced economies of scale with respect to intermediaries acquiring information about small firms. Moreover, the spread between the interest rate on loans paid by bankdependent (small) firms and the interest rate paid by (large) firms, who use public debt markets, tends to increase during monetary contractions (Kashyap, Stein and Wilcox, 1993). A reduction in the supply of bank credit reduces real activity. During the Asian crisis in the second half of the 1990s, for example, the disruption in the supply of credit was a major factor in the recessions experienced by the affected countries. Banks in these countries were unable or unwilling to establish credit facilities required for importers to provide overseas suppliers assurance of payment. The duration of a credit squeeze depends on how long it takes to establish new or revive old channels of credit after a disruption. In Asia it lasted several months in some countries and over two years in the case of Indonesia (Grimes, 1998). The bank lending channel is likely to be more important in small open economies than large closed countries. This is because the proportion of small, typically bank-dependent, firms tends to be higher than in larger economies. Financial innovations and deregulation are unlikely to have materially improved bank-dependent firms’ ability to borrow from the capital markets. In the presence of open capital markets the information problem is augmented by additional informational asymmetries between foreign and domestic borrowers and lenders. Foreign investors may be less willing to lend to small firms than domestic lenders. If banks are special in providing credit to a large fraction of firms in the economy, the amount of credit channelled through the banking system may have significant 25 Sofianos, Wachtel and Melnik (1990) and Bernanke and Blinder (1992) provide empirical evidence of the response of bank loans to fluctuations in interest rates. WP 03/19 ASYMMETRIC INFORMATION, FINANCIAL INTERMEDIATION AND THE MONETARY TRANSMISSION MECHANISM: A CRITICAL REVIEW 15 macroeconomic effects. Most macroeconomic models currently used by policy makers do not explicitly incorporate a banking sector and hence do not capture these bank lending effects of the monetary transmission mechanism. Information asymmetries and the inability of lenders to monitor borrowers costlessly lead to “agency costs”, which create a wedge between the costs of internal and external financing for a firm (Bernanke and Gertler, 1989). Cash flow and net worth are important determinants of agency costs and hence the cost and availability of finance, and ultimately the level of investment (Walsh, 1998). For instance, a firm with high net worth (equity) thereby signals its credit-worthiness to banks which can in turn lend to the firm without having to incur the same expected monitoring costs, so lowering the cost of borrowing for the firm. If net worth is depleted, the result is a rise in the firm’s borrowing costs. The second channel, the balance sheet or financial accelerator effect, focuses on the potential impact of monetary policy on firms’ financial positions and arises from the presence of agency costs. Agency costs in credit markets occur whenever lenders delegate control over resources to borrowers, leading to adverse selection, moral hazard and monitoring costs because of the inability to monitor borrowers or share in borrowers’ information costlessly. As discussed in section 2, a tightening in monetary policy that raises real interest rates reduces the profitability of firms’ investment projects because of higher financing costs (and indirectly because of the prospect of a slowdown in consumption).26 Higher interest expenses lower firms’ cash flow and internal funds, and could increase firms’ short-term borrowing (and interest expenses) if firms need to borrow to finance an inventory build-up as a result of slowing demand. Interest expenses may thus remain high for some time even after short-term interest rates have started coming down because of the rise in shortterm debt outstanding. The effects of a corporate cash squeeze on real activity depend largely on firms’ ability to smooth the decline in cash flows by borrowing. Gertler and Gilchrist (1994) find that larger firms, which are more likely to have recourse to commercial paper markets and other sources of short-term credit, typically respond to an unanticipated decline in cash flows by increasing their short-term borrowing. These firms are, at least temporarily, able to maintain their level of production and employment during periods of rising interest rate costs and declining revenues. As a result, inventories of large firms tend to grow following a tightening of monetary policy. In contrast, small firms, which in most cases have more limited access to short-term credit, respond to the cash squeeze by decumulating inventories and cutting work hours and production. In the presence of agency costs, the effects of a monetary tightening will be amplified further via the balance sheet or financial accelerator effect. This is because the deterioration in firms’ balance sheets due to lower cash flow and internal funds worsens the agency problem and so increases the costs of external financing. Adding to the higher cost of external financing is the fact that higher interest rates lower the market value (or net worth) of firms and hence lower the value of assets that firms can use as collateral. As a result, banks may be less willing to lend to firms because the reduction in collateral increases banks’ potential losses from adverse selection: owners will have a lower equity stake in their firms, which gives them more incentive to engage in risky investment 26 The rise in borrowers’ interest expenses is, of course, only a redistribution from borrowers to lenders. However, a redistribution between borrowers and lenders is not neutral if, for example, lenders and borrowers do not have access to the same investment and spending opportunities. WP 03/19 ASYMMETRIC INFORMATION, FINANCIAL INTERMEDIATION AND THE MONETARY TRANSMISSION MECHANISM: A CRITICAL REVIEW 16 projects. In some cases, the decline in collateral may lead to loans not being extended upon maturity or even being recalled, i.e. forms of credit rationing. A tightening in monetary policy that increases the external financing premium (or wedge between the costs of internal and external funding) will have additional contractionary effects on investment, output and spending. Higher interest rates may have a much stronger contractionary impact on the economy if balance sheets are already weak, introducing the possibility that non-linearities in the impact of monetary policy may be important.27 6 General equilibrium models of the credit channel Financial intermediaries exist because of imperfect information between borrowers and lenders. Modelling asymmetric information is difficult and probably the main reason why macroeconomic models currently used by policy makers do not incorporate a fully developed credit channel. Credit channel effects are not completely ignored by these models, but tend to be incorporated in an ad hoc manner as borrowing or cash flow constraints. For instance, in the Federal Reserve Board of Governors’ FRB/US model constraints are imposed to try to mimic the behaviour of credit-constrained households and firms. Additionally, the investment equation in the FRB/US model is augmented with cash flow influences (Brayton and Tinsley, 1996). Credit constraints are also a feature of the Reserve Bank of New Zealand’s macroeconomic model to some degree. In this model, the stock of household assets and the future path of labour income determine the sustainable, longrun flow of consumption. Some consumers, however, consume all their current period income and do not accumulate assets. This can be interpreted as a credit constraint where some individuals do not have access to credit markets and hence are unable to smooth consumption by borrowing against their future labour income (see Black, Cassino, Drew, Hansen, Hunt, Rose and Scott, 1997; Claus and Smith, 1999). Dynamic general equilibrium models that formally account for an explicit role of credit market frictions in business cycle fluctuations have begun to be developed in the literature. In these models, monetary policy can have a significant impact on the real economy because of asymmetric information and agency costs. None of the models is complete. In particular, elements that are important in an open economy are still not being captured. Moreover, these models often do not explicitly account for the bank lending channel. Typically these models incorporate a financial accelerator mechanism. Bernanke, Gertler and Gilchrist (1999) develop a dynamic general equilibrium model that incorporates a financial accelerator. The key mechanism in this model is the link between an external finance premium and the net worth of prospective borrowers. The external finance premium is the difference between the cost of funds raised externally and the 27 Credit market frictions that affect firms should also be relevant to the borrowing and spending decisions made by households, particularly spending on costly durable items such as automobiles and houses. In Carroll’s (1997) buffer stock model of consumption, for example, balance sheet effects impact on consumers’ willingness to spend. If consumers expect a higher likelihood of finding themselves in financial distress, they would rather be holding fewer illiquid assets like consumer durables or housing and more liquid financial assets. As discussed by Bernanke and Gertler (1995), this channel is still comparatively underdeveloped in the literature. WP 03/19 ASYMMETRIC INFORMATION, FINANCIAL INTERMEDIATION AND THE MONETARY TRANSMISSION MECHANISM: A CRITICAL REVIEW 17 opportunity costs of internal funds. An increase in entrepreneurs’ wealth or net worth lowers the external finance premium. A decline in net worth raises it. The basic structure of the model is as follows. There are three agents in the economy: households, entrepreneurs, and retailers. Households work, consume, and save. Entrepreneurs produce output by hiring labour and using capital, which they purchased in the previous period. Acquisitions of capital are financed out of entrepreneurs’ net worth and by borrowing. Entrepreneurs’ net worth arises from two sources: profits accumulated from previous capital investment and entrepreneurs’ income from supplying labour. Entrepreneurs produce wholesale goods in competitive markets, and sell their output to retailers who are monopolistic competitors. Retailers buy goods from entrepreneurs, differentiate them (costlessly) and then re-sell these goods to households. The monopoly power of retailers allows modelling nominal rigidities in the economy; otherwise, retailers play no role. Entrepreneurs’ net worth is an important determinant of their cost of external finance. An increase in net worth lowers the external finance premium of entrepreneurs leading to increased borrowing, and thus higher investment, spending and production. A decline in net worth has the opposite effects. The main source of variation in net worth is entrepreneurs’ equity, which in turn is sensitive to unexpected shifts in asset prices and unanticipated changes in the ex post return to capital. The financial accelerator magnifies the impact of monetary policy on the real economy and smaller countercyclical movements in interest rates are therefore required to dampen output movements. The greater the extent to which monetary policy is able to stabilise output, the smaller is the effect of the financial accelerator in amplifying and propagating business cycle fluctuations. Calibrated for the United States, the model is able to replicate observed cyclical movements in macroeconomic variables when allowing for price stickiness, decision lags in investment and limited access to credit for some entrepreneurs. Bernanke, Gertler and Gilchrist (1999) do not incorporate credit rationing in the sense of Stiglitz and Weiss (1981), where some borrowers simply do not receive loans, but assume that the price of capital can differ across entrepreneurs. Kiyotaki and Moore (1997) develop a model with Stiglitz and Weiss type credit constraints. Credit constraints arise because lenders cannot force borrowers to repay their debts unless the debts are secured. Assets such as land, buildings and machinery serve as collateral for loans and borrowers’ credit limits are affected by the value of collateralised assets. The basic structure of their model is as follows. There are two goods, a durable asset (land) and a nondurable commodity (fruit) and two types of agent: farmers and gatherers. Both farmers and gatherers produce and eat fruit. In each period, land is exchanged for fruit at a given price, and fruit is exchanged for a claim to some fruit in the next period. Farmers’ technology is idiosyncratic in the sense that, once their production has started at date t with some land, then only they possess the skill necessary to cultivate the land to bear fruit at date 1+t. Moreover, it is assumed that farmers can always withdraw their labour. This implies that if farmers have a lot of debt, they may try to threaten their creditors by withdrawing their labour and repudiating their debt contract. Lenders protect themselves from the threat of repudiation by collateralising farmers’ land; that is, farmers must make a down payment in order to purchase land. At date t farmers can borrow up to WP 03/19 ASYMMETRIC INFORMATION, FINANCIAL INTERMEDIATION AND THE MONETARY TRANSMISSION MECHANISM: A CRITICAL REVIEW 18 an amount at which the repayment does not exceed the (expected) market value of their land at date 1 +t. Gatherers’ production does not require any specific skills and gatherers, unlike farmers, are not credit constrained. Because gatherers are not credit constrained their demand for land is determined where the present value of the marginal product of land equals the opportunity cost of holding land. The amount of land is fixed, so if farmers’ demand for land increases, then in order for the land market to clear, gatherers’ demand has to decline. The dynamic interaction between credit limits and asset prices is an important transmission mechanism by which the effects of shocks persist, amplify and propagate business cycle fluctuations. For example, a temporary productivity shock that reduces the net worth of credit constrained farmers forces farmers to cut back their demand for land. For the land market to clear, the demand for land by the gatherers has to increase, which requires that their opportunity costs of holding land must fall. The land price drops by the same amount as the opportunity costs of holding land, which lowers the value of farmers’ existing landholdings, and reduces their net worth still further. Small temporary productivity shocks thereby generate large and persistent fluctuations in output and asset prices. The fluctuations in output become even more persistent when capital (trees) is reproducible, i.e. investment is introduced, and when in each period only a fraction of the farmers are able to invest. Carlstrom and Fuerst (1997) develop a general equilibrium model that explicitly incorporates financial intermediaries. There are three agents in this model: entrepreneurs, consumers and financial intermediaries (capital mutual funds). Entrepreneurs receive external financing from households through the capital mutual funds. The creation of new capital (including entrepreneurial capital) is subject to agency costs, which in turn are a function of entrepreneurs’ net worth. An increase in net worth lowers agency costs and thus lowers the cost of new capital. A decline in net worth increases agency costs and the cost of capital. Calibrated for the United States, the model is able to replicate observed movements in output. Following a temporary positive productivity shock, for example, output rises gradually. This is because the shock causes households to delay their investment decisions until agency costs are at their lowest level – several periods after the initial shock. Agency costs fall over time because the shock increases the return to internal funds and net worth.28 Carlstrom and Fuerst show how agency costs arising from the difference between the costs of external and internal funds can alter business cycle dynamics following a supplyside shock. However, the model abstracts from demand-side shocks and does not explicitly model a monetary authority. Edwards and Végh (1997) develop a theoretical model of a small open economy with privately-owned banks that allows for monetary policy shocks. They have four agents in the economy: households, entrepreneurs, banks and a government/monetary authority. Households are subject to a deposit-in-advance constraint and must use demand deposits to purchase consumption. Entrepreneurs produce output by hiring labour from households. They face a credit-in-advance constraint and must borrow from banks to pay households’ wages. Banks lend to entrepreneurs and hold households’ demand deposits. 28 Carlstrom and Fuerst (1997) assume that entrepreneurs are long-lived. In Fuerst (1995) entrepreneurs only live for a single period and the model is unable to replicate the positive serial correlation observed in output. WP 03/19 ASYMMETRIC INFORMATION, FINANCIAL INTERMEDIATION AND THE MONETARY TRANSMISSION MECHANISM: A CRITICAL REVIEW 19 Banks finance their lending to entrepreneurs through deposits and borrowing internationally. Banks’ operations are costly, which introduces a wedge between the lending rate and the deposit rate, the interest spread. Changes in the interest spread and bank credit, resulting from shocks to the banking system and world business cycle, lead to fluctuations in output and employment. Edwards and Végh (1997) explicitly model financial intermediaries and allow for a foreign sector. However, the theoretical model, which is not empirically tested, only incorporates the bank lending channel, although it appears that the bank lending channel by itself cannot explain the macroeconomic fluctuations that are observed following a monetary policy shock (Fisher, 1999). Moreover, the economy in the model does not operate under a flexible exchange rate and so does not provide information on interactions between financial intermediaries' actions and the exchange rate channel. The models discussed in this section are important contributions to understanding the effects of imperfect information in credit markets. But each is incomplete and further work is needed. General equilibrium models of the credit channel, where agents’ decisions are derived from optimising behaviour, have mainly focused on the closed economy. The credit channel has yet to be incorporated in a model of an open economy with a floating exchange rate. Moreover, models to date generally only allow for either the bank lending channel or the balance sheet channel, but not both. 7 Summary and conclusions Macroeconomic models currently used by policy makers generally assume that the Modigliani and Miller (1958) theorem holds. The theorem implies that, under the assumptions of complete markets and the absence of information and transaction costs, the workings of financial markets can be fully summarised by financial prices and allows one to abstract from considerations of credit market conditions. However, imperfect information and theories of financial intermediation suggest that information and transactions costs are important and the assumptions upon which the Modigliani-Miller theorem is based, and thus macroeconomic models used by policy makers, do not hold in practice. These models hence disregard the importance of credit markets and financial intermediaries not only for individual depositors but the economy as a whole. Financial intermediaries play an important role because they reduce the cost of channelling funds between relatively uninformed depositors to uses that are informationintensive and difficult to evaluate. If banks and other intermediaries provide credit to a large fraction of firms, who otherwise would not be able to borrow, the amount of credit channelled through the banking system can have significant macroeconomic effects, highlighting the importance of public policy in designing policies that ensure the soundness of the banking system. These policies may include deposit insurance, a lender of last resort function and prudential regulation, including imposition of a minimum capital ratio. However, such policies may themselves contribute to the impact of the credit channel. For instance, consider a bank that is required by regulation to hold at least a 4 percent tier 1 equity to risk-adjusted assets ratio (as under the Basle regime) and which initially has a 5 percent equity ratio. The bank then experiences a shock to its capital base as a result of a negative economic shock that reduces its equity ratio to 3 percent. If it cannot access new equity immediately, it has no choice but to reduce the size of its loan portfolio by a combination of extending fewer new loans, reducing loan rollovers and calling in existing WP 03/19 ASYMMETRIC INFORMATION, FINANCIAL INTERMEDIATION AND THE MONETARY TRANSMISSION MECHANISM: A CRITICAL REVIEW 20 (callable) loans. The initial negative economic shock thereby is magnified as the credit supply is diminshed. By contrast, without the 4 percent minimum, the bank may be able to undertake less drastic short-term adjustment. In some circumstances, therefore, there may be a trade-off between soundness (minimum capital) and monetary policy (credit channel) concerns. Nevertheless, policies which promote bank soundness – for instance, through a requirement for directors’ attestations regarding adequacy of risk management systems – in general, reduce the potential for shocks to impact on bank capital in the first place, so helping to mitigate both soundness concerns and credit market imperfections. The credit channel literature has made great strides in recent years, but significant issues remain unresolved. Much of the literature to date has focused on the United States, which can be adequately modelled as a large closed economy. The credit channel has yet to be incorporated in a model of a small open economy with a floating exchange rate. Moreover, existing models generally only allow for either the bank lending channel or the balance sheet channel, but not both. To account for the bank lending channel, a model would need to explicitly incorporate financial intermediaries and bank lending, while balance sheet or financial accelerator effects could be captured by making firms’ acquisitions of capital depend on their net worth. The foreign sector could be accounted for by allowing at least some agents in the model to borrow from abroad. The development of such a model based on an amalgam of recent advances of the literature presents an agenda for a programme of future work. This programme, by incorporating financial market interactions into macroeconomic models, will enhance the understanding of the transmission mechanism of monetary policy and other shocks to the economy.