Nonbank Financial Intermediaries, Credit Expansion, and Monetary Policy (2024)

NONBANK FINANCIAL INTERMEDIARIES (henceforth in this paper referred to as NFI) have grown and continue to grow in importance in many countries of the world. Their development has led to serious consideration of the question whether the efficient conduct of monetary policy requires closer regulation of these institutions. At present, in most countries they are regulated in much the same manner as commercial corporations. This paper is an attempt to clarify their role in the expansion of credit and to examine the manner in which they differ from commercial banks; the results of this investigation are then used to assess some of the implications of the growth of NFI for monetary policy.¹

Credit Creation by Nonbank Financial Intermediaries

The first question to be considered is whether the operations of NFI increase the amount of credit that can be extended on the basis of any given level of central bank liabilities to the public, and if so, by how much. The introduction of NFI into the financial system can in fact be shown to increase the potential amount of credit that may be extended for any given increase in the liabilities of the central bank, and this conclusion can be extended without difficulty to a financial system where NFI that are already operating grow relatively to the banking system. One of the principal conclusions of this paper is that any threat which the NFI may present to the successful implementation of monetary policy lies not so much in their long-run secular growth as in their ability in the short run to expand credit, immune from direct central bank control.

It is well known that, if banks maintain some minimum reserve ratio against deposits, any given increase in bank reserves (i.e., in the liabilities of the central bank) makes it possible for the banking system to increase its deposit liabilities and, simultaneously, its loans and investments by some multiple of the increase in bank reserves.² If there were no NFI, these bank operations would account for all of the credit that could be extended by the financial system.

How far will this situation be changed by the introduction into the credit system of an NFI, e.g., a building society which accepts deposits on the same conditions as the banks and which makes only mortgage loans? If a private individual then shifts a deposit from a bank to an NFI, say in response to an interest rate differential, and the NFI deposits this sum in its own bank account, the total deposits of the banking system will remain unchanged.³ All that has happened is a change in the ownership of these deposits; the deposits of private individuals have fallen and the deposits of intermediaries have risen by the same sum.⁴ However, the total deposits of the financial system have increased since the building society has accepted a deposit liability which did not exist before. The corresponding asset held by the intermediary at this initial stage is a deposit with the banking system. Provided the NFI hold their reserves in the form of deposits with the banking system, the use of the funds which they have newly acquired for the expansion of mortgage loans will leave the reserves of the banking system unchanged. Since the loans and investments of the banks are unaffected by the shift in deposits, loans by the NFI represent a net addition to the credit extended to the public. If the customary reserve ratio of the building society is r’, then the society may lend, in the form of mortgage loans, an amount equal to (1-r’) times the increase in its deposits. If some of this increase in credit extended by the NFI results in a further rise in the deposits of the NFI, a further increase in credit will be possible in a manner analogous to that in which the bank credit multiplier operates. It is important to note, however, that the initial shift in deposits from the banks to the NFI is by itself sufficient to permit a net addition to the total credit outstanding to the public. The multiplier effect of any return flow to the NFI increases only the amount of the NFI’s net addition to total credit.

Influence of NFI on Bank Credit Multiplier and Combined Bank and Nonbank Multiplier

In the analysis thus far, deposits have been treated as hom*ogeneous, whereas, in fact, there are important distinctions between demand deposits and time deposits. Of course, one obvious difference is that the former are part of the money supply, as customarily defined, while the latter are not. The mechanism described in the previous section involved the transfer of a bank deposit from the account of an individual to the account of an NFI, which left the total deposits of the banking system unchanged. However, when it receives a deposit, an NFI will normally place it in a demand deposit account with a bank until, as a result of being loaned out or invested, the deposit is transferred to a demand or time deposit account of some other individual in a bank. If the initial shift of deposits is from an individual bank time deposit to an NFI deposit (and this is likely to occur frequently since NFI deposits are in many respects closer substitutes for time deposits than for demand deposits of banks), then the ratio of the demand deposits to the time deposits of the banking system will rise. This rise may have important secondary effects which will tend to diminish the amount of potential credit expansion indicated by the previous analysis.

For most banks, the reserves held against demand deposits are higher than those against time deposits. Either the banks are forced by legal requirement to maintain the higher ratio, or they do so voluntarily on the basis of the banking principle that a more liquid asset should be held against a more liquid liability. A shift in time deposits from the banking system to the NFI, which results in a higher ratio of demand to time deposits in the banking system, raises the average required or customary ratio of bank reserves to deposits. This rise in the reserve ratio lowers the value of the bank credit multiplier and will partially offset the potential increase in credit extended by the NFI.⁵

The net amount of potential credit expansion permitted by a shift in deposits from banks to the NFI depends on the reserve ratios maintained by banks against time and demand deposits, the percentage of the shift represented by time deposits, and the nonbank multiplier. If it is assumed that the banks are initially “loaned up,” an exact expression may be derived, as is shown in the Appendix, for the total amount of credit potential added to the financial system by such a shift. In general, the smaller the differential between the reserve ratios against time and demand deposits, the higher the general level of bank reserve ratios; and the lower the percentage of the shift represented by time deposits, the greater will be the amount of potential credit expansion.⁶

To obtain some idea of the quantitative importance of this effect, a numerical example will be constructed using quasi-realistic values for the United States. It will be assumed that $1 billion of bank time deposits (approximately 2 per cent of total bank time deposits) is shifted to the NFI. The value of the NFI credit multiplier is computed on the assumption that the reserve ratio of the NFI is between 3 per cent and 10 per cent and the return flow to NFI, between 5 per cent and 10 per cent (slightly wider limits for these variables will produce approximately the same results). The average ratio of reserves held by banks against demand deposits is 18 per cent; that against time deposits is 5 per cent. If the banks are “loaned up,” the following results are obtained: (1) The net increase in the credit potential of all financial intermediaries is $0.41 billion; (2) the increase in the credit potential of the NFI is $1 billion; and (3) the reduction of credit extended by the banks is $0.59 billion.

The difference between the reserve ratios of the banks against time deposits and the ratios against demand deposits is thus sufficient to offset almost three fifths of the increase in the credit potential of the NFI. However, to the extent that some of the shift might consist of demand deposits rather than of time deposits, the smaller the offset would be and the larger the total increase in potential credit expansion.⁷

Difference Between Banks and Nonbank Financial Intermediaries

If banks create credit and the NFI create credit, what is the difference between them? One difference, of course, is obvious: the banks create demand deposits and the NFI do not. In fact, in the creation of demand deposits, the commercial banks have an institutional monopoly, which is often established by custom rather than by law. Because of their ready substitutability for currency, these demand deposits are considered as part of the stock of money. The banks, therefore, may create monetary liabilities while the NFI may not.⁸

The NFI may be ranked according to the closeness of the substitute for money that is provided by their liabilities. If this closeness to money could be depicted on a linear scale, it would be found that the distance between adjacent financial intermediaries would vary greatly, with many wide gaps. At the end nearest to the money-creating institutions there would be savings and loan associations, building societies, etc., all of which produce close substitutes for bank time deposits. At the far end of the scale are insurance companies, pension funds, etc. While the financial liabilities of the financial intermediaries at the end of the scale closest to money may be readily substitutable for one another at the margin in response to small differences in the rate of return, the liabilities of the NFI at the opposite end may be less influenced by differences in the rate of return, because it is difficult to find substitutes for the services, such as life insurance claims and pension claims, which are attached to the liabilities. The liabilities of some financial intermediaries may be closer substitutes for durable goods than for money.

On the asset side, the NFI may be classified according to the liquidity of their assets or the specialization of their asset portfolio. In general, the NFI will be less liquid than commercial banks (with notable exceptions), and their asset portfolios will generally be more specialized.

It is also logical to ask, What is the difference, from a financial point of view, between an NFI and any ordinary commercial corporation that issues securities? The difference is largely one of degree. On the liabilities side, the assets of a commercial corporation are less perfect substitutes for money, particularly with reference to the stability of their money value (again with notable exceptions); and on the asset side, there is a high degree of specialization—a commercial corporation will normally consist of only a single company with perhaps a few subsidiaries. Finally, there is a very close link between the issuance of the corporate liabilities and physical investment. Although there are many exceptions to these generalizations, collectively they serve to distinguish most commercial companies from the NFI.

Some important conclusions can be drawn from these commonplace observations. Financial institutions differ from one another principally in the liquidity of their liabilities and the specialization of their asset portfolios. They influence not only the amount of credit but also the kind of credit granted.

Banks in the past have tended to be singled out for special attention by the monetary authorities because they create money. In turn, it was thought that a scarcity or plenitude of money was a potent force in affecting the economic behavior of individuals and businessmen. However, the possibility exists that other types of financial asset may be sought after without increasing the demand for money. The nonfinancial intermediary public may wish to increase its holding of other assets while keeping its stock of money constant. In this latter situation, the ability of the financial system to create credit may be partially independent of the amount of central bank liabilities to the public and of the level of legal reserve requirements for commercial banks. This is a consequence of the existence of financial intermediaries which, without requiring any net addition to the liabilities of the central bank or private banks,⁹ may create nonmoney financial assets that the public wishes to hold.

Importance of Nonbank Financial Intermediaries for the Conduct of Monetary Policy

How do the NFI influence the effectiveness of monetary policy? It was demonstrated above that the addition of the NFI to the financial system, or their growth relative to the banks, may increase the credit multiplier of the financial system. However, as shown in Table 1, the credit multiplier of the NFI will, for probable reserve ratios and return flows, lie close to 1. When the differential reserve effect discussed above is considered, the effect of the growth in the relative importance of the NFI on the expansion of credit is further reduced. The general conclusion that might be reached—although any specific statement depends on the magnitudes of the significant variables in each country—is that the growth of the NFI relative to the banking system increases the potential ability of the financial system to expand credit, i.e., the financial intermediary credit multiplier, but not at an alarming rate for reasonably expected values of the variables. In fact, the growth of the assets of the NFI represents, in part, a desire of the community to hold more of its savings in the form of financial claims as the community’s stock of wealth reaches higher levels and, therefore, in part, justifies a certain amount of credit expansion.

The threat that the growth of the NFI may present to the effectiveness of monetary policy lies not so much in the higher credit multiplier of the financial system that results from the more rapid secular growth of the NFI relative to the banks, but in the fact that this growth introduces an increasingly important element of credit expansion which in most countries, under present legislation, cannot be directly controlled by the monetary authorities. Even if, by restricting the amount of reserves available to the banks, the authorities are successful in preventing a rise in bank credit, a large increase in credit may still occur through a shift in deposits from the banks to the NFI. This shift will, in general, lead both to a net increase in the quantity of credit and to a shift in the distribution of credit in favor of the type in which the NFI specialize.¹⁰

In times of credit stringency, the NFI may be in a strong position to attract deposits from the banks to themselves by raising the rate of interest payable on their liabilities. Because of institutional and legal restrictions, the banks may not meet the higher interest rates of the NFI and as a result will lose deposits to them.¹¹ The NFI whose liabilities are close substitutes for bank liabilities will be more effective in diverting deposits from the banks than those, such as life insurance companies, which provide more remote substitutes.

On the other hand, the banks themselves may set up NFI to avoid the institutional and legal restrictions placed on the rate of interest they may charge, and as a result will indirectly obtain collectively the benefit of the reserve-economizing features of the NFI¹²

The operations of the NFI may therefore threaten to diminish the effectiveness of monetary policy when, by means of interest rate differentials or otherwise, they are able to induce a rapid shift of deposits from banks to themselves, thereby increasing in the short run, without regard to central bank monetary policy, the amount of outstanding credit. The effects of the long-run secular growth of the NFI relative to the banking system which results in a higher financial intermediary multiplier can, however, be offset, if it is desired, by lowering the rate of growth of the liabilities of the central bank. For the purposes of countercyclical monetary policy, it is changes in the rate of growth of the NFI relative to the banking system that are most relevant.¹³ This change in rates of growth has, for purposes of simplicity, been analyzed in terms of a shift in deposits from banks to NFI. If NFI growth rates are stable, the central bank can offset the effects of their operations relatively easily, although it may mean that this is done by retarding the rate of growth of the banks more than that of the NFI.

To counteract the expansion of credit resulting from short-run shifts in deposits from banks to the NFI (or the acceleration of the growth of NFI relative to the banks), the central bank can only further reduce the amount of bank credit outstanding. Under existing legislation, it cannot directly affect the credit outstanding of the NFI. This credit contraction is in addition to any credit restriction already induced by the reserve ratio differential effect of a shift of time deposits to demand deposits in the banks resulting from the shifts of deposits to the NFI. The change in the distribution of credit in favor of the types granted by the NFI may, or may not, be desirable from the point of view of the over-all economic policy being followed. The fact that the quantity of credit granted by financial intermediaries may have to be controlled by regulating the quantity of credit of the banking system alone has disturbed some individuals. However, in a growing economy this need not, in the long run, work hardship on any individual bank. It may mean that in a boom the banking system will expand less than other financial institutions; the question whether this is equitable or not raises diverse issues which cannot be discussed here. However, the larger the growth of the NFI relative to the banking system, the more burdensome will become the adjustments that the banking system will have to bear in order to achieve any given degree of monetary restraint.

The final consideration, which used to be regarded as all-important, is that the NFI do not create money and that their ability to extend credit may be limited by the reluctance of individuals to permit further transfers of bank deposits or hoards to the NFI which would reduce their holdings of money assets. Except possibly for institutions, such as savings and loan associations, that accept time deposits, the liabilities of the NFI are imperfect substitutes for the liabilities created by the banking system, and there are limits to the amount of transfers of deposits which will take place in response to moderate interest rate differentials. The upper limit on the rate of interest that the NFI may charge is ultimately determined by the rate of return on new investment.

Some of the assets associated with the operations of the NFI may be closer substitutes for durable goods than for money, e.g., life insurance may be purchased instead of an automobile on hire purchase, and the presence of the NFI may stimulate a higher rate of saving by the community. In such a situation, some credit expansion is justified, and the monetary authority need only consider whether, in view of the over-all economic policy being followed, the expansion is excessive or of an undesirable type.

Some Proposals to Regulate Nonbank Financial Intermediaries

Several countries are giving serious consideration to the advisability of prescribing some type of regulation for the control of NFI. For those institutions which accept deposits from the public, minimum reserve requirements are the most commonly considered type of regulation.

The results of a minimum reserve requirement for NFI differ considerably according to whether their reserves are allowed to be held with the private banks or are required to be maintained with the central bank. If the former alternative is adopted, the imposition of reserve requirements will not, as explained in an earlier section, cause the total deposits of the banking system to decline. However, the ratio of the banks’ time deposits to demand deposits would probably decline, though the decline might be checked if the NFI were allowed to include time deposits as part of their reserves.

The prescription of a minimum reserve ratio for the NFI, unless the ratio is set very low, would require some increase in the actual reserve ratio of the NFI since those institutions generally operate with reserves that are close to minimum transaction balances. A rise in their reserve requirements would lower their loan potential through the direct effect of sterilizing part of their deposits as reserves and lowering their credit multiplier. Loans and investments of the NFI, in general, would fall by almost the same amount as the increase in reserve requirements, since, as may readily be seen from Table 1, their credit multiplier will change by only a small amount for all except very large changes in their reserve ratio or very large return flows. The rise in reserve requirements might reduce the earnings of the NFI to a point where they would be forced to lower the rate of interest paid on their deposits, and this would reduce their ability to attract deposits.

If the reserves of the NFI were required to be deposited with the central bank, the effect on the NFI initially would be the same as if they were required to be deposited with the private banks. However, the reserves of the banking system would be reduced, with a consequent contraction of bank credit by some multiple of the contraction in reserves, and this might affect the amount of deposits shifted to the NFI and cause a reduction in both bank credit and NFI credit. In fact, it is quite possible under these circ*mstances that an increase in the reserve requirements of the NFI might compel a contraction of bank loans and investments which was greater than the contraction of the loans and investments of the NFI. The net result of requiring NFI to deposit their reserves with the central bank would then be to shift a large part of the burden of the contraction of credit to the banking system. Furthermore, there is the extreme case where, in a time of credit stringency, restrictions are placed on the rate of interest that the banks may charge. The tightness of credit might then raise the market rate of interest and enable the NFI to increase their interest rates for deposits and thus to attract new deposits. These new deposits might be enough to offset the rise in reserve requirements so that the entire burden of adjustment would be shifted to the banks.

Another proposal that has been made is to create a special central bank for the NFI in which they would be required to hold reserves. The direct effect on the lending capacity of the NFI and the banks would be the same as if the NFI were required to hold reserves with the central bank. The indirect effects would depend on whether the special central bank sterilized the reserve funds deposited with it or used them to extend credit and engage in other activities that might be assigned to it. If the special central bank performs other central bank functions, such as that of being a lender of last resort or discounting, its establishment would introduce another element of potential credit expansion into the financial system. The creation of a new lender of last resort would enhance the liquidity of all NFI and might permit them to hold smaller “excess” reserves than would otherwise have been the case. The enhanced liquidity of the NFI might also cause the public to regard the liabilities of the NFI as better substitutes for money and might increase the ease with which the NFI could attract deposits from banks and hoards. More important, however, by lending to the NFI, the special central bank would provide them with assets which would give them means to expand their credit, though the reserve ratio for the NFI would provide some control over their extension of credit. A full evaluation of this proposal would require an analysis of the types of credit that the NFI would extend if their assets and general liquidity were increased and of other questions that cannot be considered here.

Conclusions

Any reserve requirement for the NFI that would require them to maintain their reserves with the central bank or with a special bank for NFI would have important secondary effects on the banking system. If the NFI were permitted to deposit their reserves with the private banking system, they could be regulated independently of the banks. The reserve requirements imposed upon the NFI would then restrain their expansion, while the reserve requirements imposed upon the banks and other measures would restrain the expansion of the banks. It is possible, however, that the pattern of contraction or expansion that would be induced by deposits of NFI reserves with the central bank would be considered desirable, and that this type of requirement would be favored by the authorities on the ground that it would provide a means—similar to central bank open market sales of securities—of curtailing bank credit.

An important point which this analysis makes clear is that, where either the loan rate of interest of the banks differs from the rate that would be established in a free market, or the rate of interest payable by the banks on deposits is controlled, it is not sufficient merely to prescribe legal reserve requirements. The possibility of large interest differentials between NFI and banks makes the financial system vulnerable to large shifts in deposits from banks to NFI with a consequent expansion in total credit. It is this possibility which constitutes a greater threat to the effectiveness of monetary policy than the secular increase in the financial intermediary credit multiplier resulting from the relative growth of NFI.

If the monetary authorities are to obtain the same degree of control over credit expansion by the NFI as they have over the banks, it would be necessary to impose restraints similar to those placed, either by law or by custom, on the loan and deposit rates of interest of banks. In fact, in some countries, if equivalent restraints on interest rates were placed on the NFI, the danger of sudden shifts in deposits would be removed and the case for NFI reserve requirements would lose force. In addition, the narrowing of the interest rate differential might affect the secular rate of growth of the NFI. However, if the same direct control of credit that exists for the banks is desired for the NFI, reserve requirements must also be imposed.