The Treasury

Global Navigation

Personal tools

Treasury
Publication

Asymmetric Information, Financial Intermediation and the Monetary Transmission Mechanism: A Critical Review - WP 03/19

3  Traditional literature on financial intermediation

The credit creation process is “the process by which, in exchange for paper claims, the savings of specific individuals or firms are made available for the use of other individuals or firms” (Bernanke, 1992-93, p. 50). The idea that the credit creation process can have real economic effects is not new. The role of bank lending in the propagation of cyclical fluctuations has been examined since at least Wicksell’s early writings on monetary dynamics and Fisher’s “debt-deflation theory of great depressions”.

Wicksell’s theory focuses on the tendency for bank lending and the money supply to expand during periods of boom and strong demand for loanable funds. To explain these credit expansions during booms, Wicksell introduces the concept of the “natural rate of interest”. The natural rate is defined as the “rate of interest at which the demand for loan capital and the supply of savings exactly agree” (Wicksell, 1906, p. 193). It is determined by the demand for loans, which depends on the expected profitability of investment.

The trouble as Wicksell saw it was that the market rate of interest tends to be sticky, responding only slowly and with a delay to changes in the demand for funds. Because of this stickiness, if the natural rate rises, because of changes in domestic or foreign demand, the market interest rate will be below the natural rate for some time, increasing the demand for loans (and lowering the supply of saving).[10] The increase in borrowing and investment in turn raises spending and prices, leading to higher expected profitability of investment, further investment, spending and increased prices. This “cumulative” process continues until banks raise market interest rates (when running out of excess reserves).

Fisher (1933), who examined the Great Depression, argues that economic crises result from “over-indebtedness” and can be exacerbated by falling prices. The “great cause of over-borrowing” is “easy money” (Fisher, 1933, p. 348).

During the 1920s, households and unincorporated businesses had greatly increased their levels of debt and, in Fisher’s view, the high leverage of borrowers was the prime cause of over-indebtedness. Over-indebtedness in turn led to “over-production” (or excess supply) and prices fell to restore equilibrium – a period of over-production (excess supply) was followed by a period of under-production (excess demand), causing large cyclical fluctuations in output.

The direct effect of the business downturn that led to the Great Depression was an increase in bankruptcies, which still further contributed to the slowing of the economy. The indirect impact, and possibly the more important channel in the propagation mechanism, was the price deflation caused by the downturn. Given that debt contracts were written in nominal terms, the protracted fall in prices and money incomes greatly increased real debt burdens. The deterioration in borrowers’ net worth due to higher debt induced borrowers to reduce spending, sending the economy into further decline, and thus continuing the spiral of falling output and deflation.[11]

In Wicksell’s and Fisher’s view, business cycles are caused by the expansion and contraction of credit. In Keynes’ (1936) General Theory of Employment, Interest and Money, the financial variable most relevant to aggregate economic activity is money rather than credit. The financial system plays a less explicit role in Keynes’ theory of output determination, although credit market conditions are important for investment behaviour. A key factor in the determination of investment is the “state of confidence” and Keynes (1936) distinguishes two basic determinants of this state. The first is the borrowers’ beliefs about “prospective yields” from investment projects. The second is the “state of credit”, which is determined by the confidence lenders have in financing borrowers. A collapse in confidence of either borrowers or lenders is sufficient to induce a downturn.

The macroeconomics literature following the General Theory largely ignored any links between movements in output and credit market conditions (Gertler, 1988). Some attention was redirected toward the interaction between financial structure and real activity with Gurley and Shaw (1955), who emphasised the role of financial intermediaries in improving the efficiency of intertemporal trade and economic growth. Their argument is based on an observed correlation between economic development and the system of financial intermediation.[12] In developed countries there generally exists a highly organised and broad system of financial intermediation to facilitate the flow of loanable funds between borrowers and lenders, while in developing countries the financial system is much less evolved.

According to Gurley and Shaw (1955), the economy’s overall “financial capacity” is more relevant to macroeconomic behaviour than money. Financial capacity is the borrowers’ ability to absorb debt, without having to reduce spending in order to avoid default. Financial capacity is an important determinant of aggregate demand, and balance sheets (a key determinant of financial capacity) can enhance the movements in spending and magnify business cycles. Financial intermediaries are important because they extend borrowers’ financial capacity.

However, any potential momentum provided by Gurley and Shaw was soon deflected by Modigliani and Miller’s (1958) derivation of the formal proposition that real economic decisions are independent of financial structure in the general equilibrium framework of Arrow and Debreu. Modigliani and Miller (1958) showed that, when capital markets are perfectly competitive and information and transactions are costless, the value of a firm and its investment decisions are independent of the source of finance, i.e. it is irrelevant whether a firm finances its investment decisions through debt or equity.

The Modigliani-Miller theorem allowed economists to abstract from consideration of credit market conditions. In part, credit markets fell by the theoretical wayside because they lacked microfoundations, in contrast to the Modigliani-Miller theorem. It was not until the 1970s with the rise of the economics of imperfect information that foundations began to be developed that posit an important role for credit markets (discussed in the next section). Until that time, the emphasis was on the role of money in the monetary transmission mechanism.

Keynes’ (1936) discussion of “liquidity preference” had shifted the focus to the interaction of money supply and money demand as the determinants of the real interest rate. The study by Friedman and Schwartz (1963) of the historical relationship between money and output became the cornerstone for the monetarist theory, which opposes activist government policy intervention aimed at stabilising aggregate demand. “Monetarists” were given their name to reflect the emphasis they place on steady growth in the money supply as the basic tenet of stabilisation policy. Friedman and Schwartz’s (1963) work was an alternative explanation of the role of financial markets in the Great Depression, with the emphasis on the central importance of money.

Considerable debate arose over the empirical significance of the mechanism linking money and economic activity. One of the earliest empirical studies to try to assess the impact of money on economic activity was Friedman and Meiselman (1963). Friedman and Meiselman, who tested whether monetary or fiscal policy was more important in the determination of nominal income, using single-equation estimation, find a much more stable and statistically significant relationship between output and money than between output and their measure of government expenditure. Andersen and Jordon (1968) also reported a strong empirical relationship between money and nominal income.[13]

One important problem with these single-equation regressions is that they are mis-specified if money is endogenous, i.e. if the central bank adjusts monetary policy in response to shocks to output.[14] To overcome this endogeneity problem, subsequent empirical studies of the link between monetary policy and (real) economic activity adopt a vector autoregression (VAR) framework, pioneered by Sims’ (1984) reduced-form bivariate model of money and output. The common result, Sims and others find, is that lagged values of money can help explain variations in output. This general statistical pattern provided the motivation for developing more structural models of movements in the money supply and output fluctuations.[15]

Two key assumptions underlay the emphasis on money (and neglect of credit). First, the central role of bank deposits (and currency) as transactions media led to the assumption that money performs a special service to businesses and consumers not performed by other assets, so that other assets are not close substitutes for money. Second, with well-functioning capital markets, bank credit is simply one alternative source of investment finance for businesses. The assumption is that if the supply of bank credit is reduced, firms seeking to fund working capital or an investment project can always issue tradable bonds or equity shares in the capital market at no extra cost. Given this assumption that other sources of finance are perfect substitutes for bank credit, any disturbance to the quantity or price of liquidity impacts on the wider economy only through the demand and supply of money (see Bernanke, 1992-93; Garretsen and Swank, 1998). From the policy makers’ perspective this meant that attempts to control credit would be negated by changes in the supply of substitutes, but attempts to control money would be effective given the absence of substitutes. This argument provided the rationale for using monetary targets in many countries during the 1970s and 1980s to conduct monetary policy.

However, with the advent of financial deregulation during the 1980s, the validity of the focus on money only and the effectiveness of monetary targets to implement policy were called into question. Widespread financial innovation, especially the increased use of credit cards and electronic banking, contributed to disrupt what was previously observed as a stable medium- to long-run relation between the stock of money and aggregate nominal income. The instability between the money stock and nominal income eventually led countries to abandon monetary targeting.

New empirical work and developments in theory (discussed in the next section) rekindled interest in the role of credit in the business cycle. The empirical work, beginning with Mishkin (1978), involves a reconsideration of an earlier issue – the role of financial factors in the Great Depression. Analysing data from the Great Depression to determine whether financial factors affected consumer spending, Mishkin finds that the behaviour of households’ net financial positions had a significant influence on consumer demand. These results provide evidence that financial market conditions played an important role in the business cycle propagation mechanism, reminiscent of the one in Fisher’s (1933) debt-deflation theory. The rise in consumer real indebtedness resulting from declining incomes and deflation induced consumers to lower spending on durables and housing, which in turn magnified the decline, contributing yet further to the economic downturn.

Analysing the relative importance of monetary versus financial factors in the Great Depression, Bernanke (1983) concludes that the collapse of the financial system was an important determinant of the depression’s depth and persistence. The basic premise in Bernanke’s (1983) analysis is that, because markets for financial claims are incomplete, intermediation between borrowers and lenders requires nontrivial market-making and information-gathering services. The disruptions of 1930-33 reduced the effectiveness of the financial sector in performing these services. As the real costs of intermediation increased, some borrowers found credit to be expensive and difficult to obtain. The effects of this credit squeeze on aggregate demand contributed to turn the severe but not unprecedented downturn of 1929-30 into a protracted depression. The two major components of the financial collapse were the loss of confidence in financial institutions, primarily commercial banks, and the widespread insolvency of debtors.

The alternative explanation of the correlation between the conditions of the financial sector and the general economy is that due to Friedman and Schwartz (1963), who stress the effects of the banking crises on the supply of money. To test these two competing propositions, Bernanke’s (1983) empirical approach is to estimate output equations using monetary variables, and then to evaluate whether or not adding (non-monetary) proxies for the financial crisis improves the performance of these equations. He finds that adding proxies for the financial crisis, such as suspended bank deposits, failing business liabilities, differentials between BAA corporate bond yields and yields on U.S. government bonds, generally improves the purely monetary explanation of short-run output movements.

Notes

  • [10]Investment, which equaled saving before the opening of the gap between the natural and market rates, now exceeds saving.
  • [11]Fisher (1933) calculates that by March 1933, “debt as measured in terms of commodities”, i.e. real debt, had increased by about 40 percent from its previous peak reached in 1929 (p. 346).
  • [12]The link between the financial system and economic growth has also been investigated more recently. See, for example, King and Levine (1993) and Khan and Senhadji (2000).
  • [13]Similar single-equation regressions were estimated to investigate the link between money and real output.
  • [14]The first study to formally model the idea that a positive correlation between money and output may not indicate the true causal role of money was Tobin (1970).
  • [15]For a summary of the literature see Leeper, Sims and Zha (1996).
Page top