A FEATURE of commercial banking operations is the existence of unused credit arising from the excess of commitments to lend by banks over actual advances, i.e., of credit facilities available for use but unutilized at a point of time or over a period. This situation can arise in all types of banking systems whether operating on the basis of overdrafts or on the basis of term loans. But unused credit is peculiarly characteristic of overdrafts which by their very nature and mechanics give rise more often, and for longer periods, to unused credit positions than does the term or fixed loan system. Therefore, unused overdrafts may be regarded as a leading species of the genus of unused bank credit.


A FEATURE of commercial banking operations is the existence of unused credit arising from the excess of commitments to lend by banks over actual advances, i.e., of credit facilities available for use but unutilized at a point of time or over a period. This situation can arise in all types of banking systems whether operating on the basis of overdrafts or on the basis of term loans. But unused credit is peculiarly characteristic of overdrafts which by their very nature and mechanics give rise more often, and for longer periods, to unused credit positions than does the term or fixed loan system. Therefore, unused overdrafts may be regarded as a leading species of the genus of unused bank credit.

A FEATURE of commercial banking operations is the existence of unused credit arising from the excess of commitments to lend by banks over actual advances, i.e., of credit facilities available for use but unutilized at a point of time or over a period. This situation can arise in all types of banking systems whether operating on the basis of overdrafts or on the basis of term loans. But unused credit is peculiarly characteristic of overdrafts which by their very nature and mechanics give rise more often, and for longer periods, to unused credit positions than does the term or fixed loan system. Therefore, unused overdrafts may be regarded as a leading species of the genus of unused bank credit.

Although the existence of unused overdrafts in the banking system has been often noted in economic literature,1 it is interesting to observe that unused overdrafts figured more prominently in the monetary discussions of the 1920’s and the 1930’s than in subsequent decades. Thus the Macmillan Committee and the Irish Banking Inquiry discussed them explicitly, whereas there is a virtual absence of references to them in the Radcliffe Report.2 Likewise, this aspect of commercial banking does not seem to have attracted much attention from practical bankers. Most current definitions of money supply also exclude unused overdrafts, and a large number of countries where the overdraft system prevails do not even publish separate statistics on them. An analysis of overdrafts was contained in an earlier study in the Fund by Mr. William H. White.3 Thus study, which was conducted in 1956, had to confine itself largely to theoretical discussion since statistical information on overdraft facilities and the unused portion of overdrafts was fairly scant at that time. Meanwhile, data have become available for a number of countries, so that an empirical evaluation can be attempted. But subsequent to Mr. White’s study there has been no systematic analysis of unused bank overdrafts on a comparative basis for different countries and their implications for monetary analysis and policy.4 The neglect of unused overdrafts by economists and bankers alike may even suggest that the phenomenon has perhaps no material bearing on banking operations and credit policies. But it would be misleading to draw any definitive inferences regarding the economic implications of unused overdrafts without a careful analysis of their nature, magnitude, and causes.

The existence of unused overdrafts poses several interesting and interrelated questions. First, it is essential to know how important, in quantitative terms, are unused overdrafts and what is the ratio of unused overdrafts to sanctioned overdraft limits. Also, are there any noteworthy features of the pattern and distribution of unused overdrafts? Second, what are the causal factors accounting for the emergence of unused overdrafts? Third, are unused overdraft limits as important for monetary analysis and policy as suggested by their size or are they largely economically fictitious? Do they impair the effectiveness of monetary policy, raise the cost of credit, and distort the allocation of loanable funds between individuals and different sectors of the economy? Last, is the exclusion of unused overdrafts from the conventional definition of money supply justifiable?

The present paper, which has been prompted by the use of overdraft facilities in a majority of the Asian countries, attempts to provide part of the answers to these questions by an analysis of unused overdrafts in the light of data that could be compiled for only a small group of countries comprising Australia, New Zealand, the Philippines, Sweden, and Norway. Although this group includes only one Asian country, it has still the merit of covering a fairly varied range of economic structures and banking systems. The phenomenon of unused overdrafts is obviously widespread, even if its magnitude may not be known, since it can arise in all banking systems where the overdraft or cash credit functions as the principal or an important form of credit, as, for instance, in the United Kingdom and in the Commonwealth (except Canada), South Africa, Ireland, Europe (e.g., Finland, Italy, the Netherlands), the Middle East (e.g., Libya, Jordan). Unfortunately, the lack or nonavailability of relevant statistics for most countries precludes a more extensive and detailed comparative study. But the available empirical evidence for Australia, New Zealand, the Philippines, Sweden, and Norway, even if somewhat fragmentary, does point to the existence of a sizable volume of unused overdrafts and permits some meaningful generalization about them.

This paper discusses the concept and mechanics of unused overdrafts (Section I) and then proceeds to present and analyze the empirical evidence relating to their size, fluctuations, and distribution (Section II and Appendices). Section II is followed by an analysis of the possible causal factors accounting for the emergence of unused overdrafts (Section III), their implications for monetary policy (Section IV), and their treatment in monetary statistics (Section V).

Broadly, the statistical data show that (1) the average magnitude of unused credit as measured by the ratio of unused overdrafts to outstanding credit limits is considerable, in the range of 30 per cent to 44 per cent, except in Sweden (nearly 14 per cent); (2) there is no consistent relationship between movements in the total sanctioned limits and the ratios of unused overdrafts, except in Sweden where they tend to move in the same direction; (3) seasonality is the dominant factor in the variations of ratios of unused overdrafts in countries where the data are adequate for an analysis of the components of change (Sweden and New Zealand); and (4) the urban sector accounts for a larger share of unused overdrafts in both Australia and New Zealand (the only countries that publish primary data on the sectoral distribution of unused overdrafts) than the rural sector.

Analysis of unused overdraft limits shows that a considerable portion of these are perhaps economically fictitious and therefore without any real bearing on credit policy or on the structure and level of interest rates or the optimum allocation of loanable funds among individual customers or sectors. However, the mechanics of overdrafts may complicate the implementation of an effective credit policy without really impairing it. The inclusion of unused overdrafts in the concept of money supply also does not seem to improve materially the explanatory value of monetary statistics, such as those of velocity of circulation. But it would be useful to have separate statistics on the size and pattern of unused overdrafts for policy purposes, and these may even have a limited predictive value in monetary models. The study also points to the need for further empirical testing of different aspects of unused overdrafts in each banking system in the light of its environmental peculiarities.

I. The Concept and Mechanics of Unused Overdrafts

Unused credit arises whenever a bank or any financial intermediary sanctions credit limits to a customer who may, for a variety of reasons, be unable to utilize these limits fully at a point of time or over a period. The problem of unused credit is therefore not peculiar to short-term commercial banking. It has in fact been widely observed in long-term lending institutions, too, such as development banks.5 There is, however, an important difference between unused short-term credit and unused long-term capital. Unused long-term funds are more often a temporary phase reflecting administrative and other lags in the processing, sanction, and utilization of loans. In short-term banking funds, the time lags, if any, in processing credit applications are minimal and inconsequential and, therefore, the persistence of unused credit over long periods would be more than a manifestation of institutional lags.

It is, however, important to stress that the overdraft system is not a unique cause of unused credit, since the excess of credit commitments over credit advanced can arise in other types of bank credit, like term loans. For instance, unused credit lines are not uncommon in term lending by commercial banks in the United States where the borrower pays a commitment fee on the amount not used within a certain period of time. Similarly, in Canada, where bank credit is extended largely through term loans, there is a considerable volume of unutilized limits of term loans with the chartered banks.6 Since this paper is limited to a study of unused overdrafts with commercial banks, a brief description of the mechanics of overdrafts is helpful in understanding why unused credit positions are more likely to occur under this system than, say, under term loans, which are the predominant form of bank credit in North America7 and many other parts of the world.

The overdraft system typically operates as follows: “In return for an undertaking … often reinforced by the handing over to the bank of some negotiable security of which the bank may dispose if the customer fails to meet his obligations, the bank allows the customer to overdraw his account … beyond the amount previously standing to his credit in the bank’s books, up to the limit set in the overdraft arrangement. (This limit will be fixed after the banker has considered the customer’s needs and prospects of profit.) This overdraft facility is equivalent to a bank deposit in representing part of the supply of money with which individuals can buy goods and services, and it becomes part of the visible supply of money as the right to overdraw is exercised and other people (recipients of the borrower’s cheques) acquire claims against the lending bank.”8 Thus, the essential part of the overdraft or cash credit9 agreement is the entitlement to borrow up to a specified limit. But the actual utilization of this limit is at the discretion of the borrower, who is under no financial pressure to draw up to the full limit, since interest is charged only on the actual drawing as reflected in the daily debit balance and is not related to the sanctioned limit. Consequently, actual drawings may be below sanctioned limits under the overdraft or cash credit system. But under the term or fixed loan system there is, by definition, no scope—at least from the standpoint of the lending banker—for unutilized credit, as the customer is entitled to borrow the whole of a fixed amount, which is immediately debited to his loan account and credited to his current account with interest being charged on the whole of the sanctioned loan. Therefore, the customer is more likely to utilize the loan with the minimum time lag after sanction.

The difference between the overdraft and the fixed loan is also reflected in their respective impact on a bank’s balance sheet. Under the overdraft system, it is only the actual drawing which appears as an asset in the bank’s balance sheet; the unused overdraft does not figure at all in the balance sheet. In contrast, the full agreed amount of the fixed or term loan is reflected simultaneously in the bank’s balance sheet as an asset under “Loans and Advances” and as a liability under “Deposits.” Consequently, the overdraft system does not fully reflect the extent of private claims on the banking system insofar as portions of overdraft limits are unutilized. This means that the balance sheet items of a banking system based on term loans would be larger than under the overdraft system, and more so if, as in the United States, borrowers are required to maintain a minimum compensatory balance on checking accounts during the life of the loan or of the line of credit.10

The overdraft system has obvious advantages of convenience, informality, and flexibility to the customer, since the drawings up to the approved limit can be varied without fresh approval as long as the limit is not exceeded. It is, therefore, particularly suitable when borrowings are for irregular amounts and uncertain periods. On the other hand, it appears to pose problems for the lending banker. For instance, one view is as follows: “As far as credit accounts are concerned there may be a fairly clear pattern established over the years which can be adjusted to give effect to recurrent seasonal trends. It is much more difficult to establish any clear workable pattern for anticipating withdrawals under overdraft facilities. The trend towards borrowing on overdraft rather than on loan account, while of the greatest convenience to customers, involves for the banker increased difficulty in the task of day-to-day cash control; and the extent to which it can call for increased safety margins will eat into the banker’s possible earnings.”11

Although the details of overdraft contracts may show considerable variations, the main features of the system are broadly similar in most countries. For purposes of the present analysis, however, it is more important to identify the types of overdraft arrangements that create possibilities of unused limits.

There are certain kinds of overdrafts which, by definition, do not give rise to unused limits, since the customer overdraws without a formal approved limit or else in excess of it. As there are no formal limits in these cases, all changes are in actual advances. Such situations, wherein drawings are in excess of limits, are not uncommon12 and may be said to exemplify the “lead” aspect of the overdraft mechanism.

But most overdraft arrangements that have formal limits create possibilities of a “lag” of drawings behind limits, as in the following instances.13

(1) The customer may secure a limit subject to repayment (usually in conjunction with a corresponding reduction in the limit) in regular installments. The lag of drawings is, on average, fairly short and stable, and often the full amount of the limit is drawn with little or no lag. The limit, however, vanishes with repayments, which makes this type of overdraft most like the fixed loan.

(2) The customer may obtain an overdraft limit that may remain virtually unchanged from year to year. This type of overdraft is commonly used by those with seasonal credit requirements, such as primary producers, and also by importers and others to finance fluctuating inventories. Limits of this type would rise only with the growth in number of customers and/or in the size of their average limits.

(3) The customer may obtain a new increased limit to meet a specific item of planned outlay. As the decision to go ahead with the expenditure depends on the limits sanctioned, the drawings will, therefore, usually lag behind approvals. But the limit is not necessarily reduced along with repayments. On the other hand, substantial fresh drawings are not made without renewed approval by the bank, which means that reductions will tend to be closely related to repayments. Generally the lag between limits and drawings under this type of overdraft depends on the interval between (a) the sanction of the limit and the placing of the order by the customer with his supplier and (b) the placing of the order and the presentation of the bill for payment by the supplier, at which point the actual drawing is made.

(4) The customer may secure an overdraft limit more as an insurance against future credit squeezes than with any firm or immediate intention of drawing against it. In this instance, too, there will be a considerable lag between limits and drawings.

There are also other types of overdrafts that generate positions of unused credit but not because of any lag of drawings. These are used as a base for raising further credit, and there are therefore no actual drawings. Such overdraft limits may be sought solely as security against trade credit with the knowledge of the banks concerned, or else existing unused limits may be used as backing for borrowing short-term funds from companies at a rate of interest below the bank overdraft rate. These arrangements in effect amount to transforming overdrafts from a source of credit to the status of collateral.

The above classification shows that, although the variations in the details of overdrafts can complicate relationships between formal limits and actual advances, their relevant feature for the present analysis is the fact that most of them are potential sources of unused credit through varying degrees of lag between limits and drawings.

Unused limits under the overdraft system are usually subject to commitment fees, which are related either to the limit of the facility or to the outstanding amounts at specified dates. But the detailed schedule of fees or charges varies considerably from country to country.14 The effective cost of credit under the overdraft system is measured by the sum of (1) interest on the used overdraft, (2) commitment fees on the unused limit, and (3) service charges to cover the costs of operating the overdraft account. Since commitment fees are usually a small fraction of the rate of interest on used overdrafts, unused limits are a source of “low-cost” rather than “no cost” liquidity15 to the borrower. But, as they afford convenient and flexible access to liquidity, they have a greater bearing on the availability of credit than on its cost.

II. Unused Overdrafts—The Empirical Evidence

Although the overdraft system prevails in a number of countries, only a few of them publish data on the extent of unused overdrafts, namely, Finland, New Zealand, Norway, the Philippines, and Sweden. For all these countries, the value of unused overdrafts is given on a quarterly basis, as part of the related data in the respective Monetary Surveys in the International Monetary Fund’s International Financial Statistics (IFS). Although no separate published series on unused overdrafts are available for Australia, these can be derived from the published data on outstanding overdraft limits and corresponding level of advances.

But for an adequate analysis of unused overdrafts, it is essential to have an idea not only of their absolute size but also of their magnitude relative to the outstanding limits. Such data, i.e., relating to limits outstanding and the extent of unused overdrafts, respectively, could be compiled only in respect of Australia, New Zealand, the Philippines, Norway, and Sweden. Even within this group only the data for New Zealand and Sweden were adequate for statistical analysis of the relative components of variations in unused overdrafts because of the availability of comparable series of overdraft limits and unused overdrafts on a monthly or quarterly basis over a sufficiently long period. For Finland, although the figures of unused overdrafts are published, comparable data on the limits sanctioned are not available.

Apart from the size and relative magnitude of unused overdrafts, it would be useful to analyze their distribution among different categories of borrowers by size of accounts as well as by economic sectors. This is so because the concept of liquidity, of which unused overdrafts are an integral part, is more meaningful if viewed as a “pattern” rather than merely as a quantity16 in an analysis of spending decisions and their probable impact on different sectors of the economy. In this respect, the data are grossly inadequate inasmuch as only Australia and New Zealand publish these particulars, and even for these two countries, there is only a sectoral classification and none according to the size of borrowers’ accounts.

The subsequent subsections deal, first, with the magnitude and fluctuations of unused overdrafts in New Zealand, Sweden, Australia, the Philippines, and Norway and, second, with the sectoral distribution of unused overdrafts in New Zealand and Australia. The concluding subsection brings together the main findings from the data.

The magnitude of unused overdrafts

The time series relating to unused overdrafts in New Zealand, which goes back to 1936, is the longest available among the present group of countries and is given on the basis of the average of monthly figures for each year (Table 9, column 2). It is based on the figures for “Unexercised Overdraft Authorities” given as part of the table on liabilities and assets of all trading banks (but shown separately from the other items) in the Statistical Summary for 1936-50 of the Reserve Bank of New Zealand and thereafter in its Bulletin. The total limits shown in column 1 of Table 9 are derived by adding the figures for “Advances and Discounts” published in the Statistical Summary and Bulletin to the figures for unused overdrafts. Since “Advances” also include over-drawings on accounts without or in excess of formal limits, like the seasonal drawings for tax payments, this method somewhat inflates the derived total of limits. On the other hand, it could also be maintained that it represents the effective level of limits.17 The series on “Advances and Discounts” is also subject to some limitations that, however, do not detract unduly from its utility. The inclusion of discounts is not a serious shortcoming, since discounts account for less than 5 per cent of total bank credit in New Zealand,18 which mostly is comprised of overdrafts. The series is also inclusive of interbank lending for 1936-57 and of term loans and export finance loans until April 24, 1963. Consequently, the figures since the latter date represent the real position more accurately. There is a seasonally adjusted series of advances, but this has not been used in the absence of a comparable adjusted series of limits.

Although it has been maintained that in New Zealand “allowing for very short-term deviations there is on the whole a fairly stable relationship between limits and actual advances,”19 this statement has to be construed with care, as in 10 out of 30 years (1936-66)—see Tables 9 and 11—total limits and the absolute level of unused overdrafts have moved in opposite directions. Moreover, a comparison of the relative movements of total limits and the proportion of unused overdrafts reveals a predominance of variations in opposite directions in the ratio of unused credit to total limits and those of total limits. The average ratio of unused overdrafts for this period (1936-66) works out to 39.2 per cent with a range of variation from 29.6 per cent (1952) to 45.4 per cent (1943). The monthly variations in unused overdrafts in New Zealand (Table 10) show a pronounced tendency for the troughs (i.e., the peak of advances) to cluster around March, whereas the peaks generally seem to be in December.

The statistical analysis20 (Table 1) of the variations in ratios of unused overdrafts in New Zealand shows that the seasonal component (S) is very dominant, whereas both the irregular component (I) and trend cycle (C) are comparatively small. This reflects the pattern of the New Zealand economy, where a large portion of bank credit is accounted for by such occupations as agriculture, stock raising, and meat processing, which are all subject to pronounced seasonal influences.

Table 1.

New Zealand: Variations in Ratios of Unused Overdrafts, 1959–66

article image
Source: Derived from data in Table 10.

The importance of seasonality is further corroborated by the size of seasonality (which, in any series, is (S), i.e., the average percentage change from quarter to quarter without regard to sign in the final seasonal factors) in New Zealand (3.54), which is well above the (S)-sized interval of 2.00 and above regarded as the high-seasonality group in a world-wide comparative study.21

The pattern of seasonality in unused overdrafts in New Zealand (Table 2) shows that November, December, January, and February are the months of high seasonality, whereas March, April, May, and June are months of low seasonality. The remaining months may be described as normal. The pattern of seasonal variation also indicates that an appreciable but not predominant part of the average amount of overdrafts outstanding during a year can serve only to a very limited extent as a liquidity substitute for money, being in effect “frozen” pending use at times of normal seasonal need. This point could be illustrated by assuming that unused overdrafts vary from zero to 200 during the year, with the average outstanding or seasonally adjusted amount being 100; none of that average figure of 100 would be available for use, for all of it would be frozen pending use at the time of seasonal peak demand for credit. In New Zealand, the range of seasonal variation is from 89 to 108, so that about 10 per cent of the average outstanding amount of unused overdrafts is frozen and only nine tenths of the total can be considered as a money substitute. If seasonal patterns could be derived for the unused overdrafts of each sector or industry, further amounts of the total would be found to be frozen in this way, for individual sectors or industries may well have their own seasonal pattern even if the aggregate of unused overdrafts exhibits no seasonality at all.

Table 2.

New Zealand: Seasonal Factors (Moving Average) in Ratios of Unused Overdrafts to Limits, 1960–66

article image
Source: Based on Table 10.

The data on unused overdrafts (commercial and savings banks) for Sweden22 (Tables 12, 13, and 14), available from 1955 on an annual basis and on a quarterly basis from 1957, show an average ratio of 13.9 per cent of unused overdrafts and a range of 9.5 per cent (1955) to 15.9 per cent (1966). The extent of unused credit in Sweden is the lowest in the group of countries covered by the present study. The relative contribution of seasonality (S) to variations in unused bank credit in Sweden (94.09 per cent)—see Table 3—is much higher than in New Zealand (82.86 per cent)—see Table 1—despite the fact that Sweden has a more industrialized economy than New Zealand.

Table 3.

Sweden: Variations in Ratios of Unused Overdrafts, 1957–66

article image
Sources: Based on Tables 13 and 14.

The average size of seasonality (7.96) is also consequently very much higher than in New Zealand (3.54), and the seasonality is particularly high in the fourth quarter and low in all the other quarters (Table 4).

Table 4.

Sweden: Seasonal Factors (Moving Average) in Ratios of Unused Overdrafts to Limits, 1957–66

article image
Sources: Based on Tables 13 and 14.

The movements in ratios of unused credit in Sweden during 1955–66 are in most of the years in the same direction as those of credits sanctioned; likewise, the peaks of overdraft limits and of ratios of unutilized credit generally coincide in the month of December, whereas their respective troughs diverge markedly except for the years 1957 and 1959 when the lowest points for the two series were touched in March.

The data for Australia in respect of overdraft limits for the major trading banks are available (as on the second Wednesday of the month) for July and October 1960 and January, April, and July 1961; the regular monthly series of limits, however, commences from September 1961. (See Table 15.) Since banks in Australia are requested to include only formal overdraft limits in their returns of limits outstanding and of new approvals, any advances in excess of formal limits will generally tend to be excluded. The unused overdraft limits have been derived by deducting from the sanctioned limits the corresponding level of advances, actual as well as seasonally adjusted. The ratios for unused credit in respect of actual advances are more useful for this analysis since, like the series for limits, they exclude term loans and temporary advances to credit buyers, whereas these are included in the seasonally adjusted series of advances. No seasonally adjusted series of advances that excludes term loans and temporary advances to buyers is as yet available. A monthly series of “new and increased lending commitments” (approvals) is also available since February 1962.

Australia’s average ratio of unused overdrafts (42.2 per cent during 1961-65) and range of variation in the ratios (36.4 per cent to 48.9 per cent during 1960-66) show the extent of unused bank credit in the system, which is the highest in the group of countries covered by this study. In each of the three years (1961-63) for which complete monthly series are available, the peak of limits was reached in December, whereas the lowest level was attained in January for 1962 and 1963. On the other hand, the peak of the ratio of unutilized credit coincided with those of limits in December 1961 and 1963 but not in 1962. The lowest points of the ratios of unused credit do not show any coincidence during this period (1961-63). Another significant feature is that the movements in ratios of unused overdrafts have not always been in the same direction as those of limits. While there has been a consistent rise in limits, the series for unused overdrafts has been often interrupted by declines.

The Philippines is the only country that includes unused overdrafts in the computation of money supply. The share of overdrafts in commercial bank credit, although less than that of loans and discounts, is still substantial (Table 5).

Table 5.

Philippines: Trend of Commercial Bank Credit by Types, 1949, 1954, 1959, and 1964–66

(In millions of pesos)

article image
Source: Central Bank of the Philippines, Central Bank News Digest.

The extent of unused overdrafts is shown in the published statistics separately alongside those of overdraft limits. The two series relating to total limits sanctioned and unutilized, respectively, as at the end of the year cover all commercial banks and are available for 1949-66 along with those for demand deposits and excess reserves (Table 17, columns 1 and 2). The average ratio of unused overdrafts in the Philippines for 1949-66 was 30.4 per cent with a range of 23.4 per cent (1949) to 37.7 per cent (1958). While limits have shown a virtually consistent rise over the period, the ratio of unused overdrafts has often fluctuated in an opposite direction to limits.

Overdrafts account for nearly 60 per cent of commercial bank credit in Norway and 36 per cent of savings bank credit.23 The outstanding credit limits for Norwegian banks were derived by aggregating the published total of drawings on cash credit and the figures of unused cash credits (Table 18). Norway shows a much higher average ratio of unused cash credits (37.2 per cent during 1952-66) for commercial and savings banks than Sweden, with a range of 33.3 per cent (1955) to 41.9 per cent (1960); also, in contrast to Sweden there is considerable divergence in the direction of variations in credit limits and ratios of unused credit.

The general presumption of a high average ratio of unused credit may also hold good for other countries where the overdraft system is prevalent but for which comparable data are not available. For instance, the extent of India’s unutilized credit during 1957-61 was estimated to be in the range of 34 per cent to 40 per cent (Table 6).

Table 6.

India: Annual Average of Credit Limits with Scheduled Banks, 1957–61

article image
Source: See A. G. Chandavarkar, “Liquidity in the Indian Economy (1949-62),” Reserve Bank of India Bulletin (Bombay), November 1963, p. 1403.

The distribution of ratios of unused overdrafts among categories of borrowers

The broad pattern of distribution of the ratios of unused overdraft limits by classes of borrowers for Australia and New Zealand (data for other countries are not available) shows some interesting features common to both countries.

For New Zealand a classification of limits outstanding by classes of borrowers corresponding to that for advances to the same categories is available at six-month intervals (January and July) since January 1960. The ratios of unused overdrafts by class of borrowers (Table 11) were derived as a residual by working out the relative advances/limits ratios for these groups. These data show that the highest proportions of unused overdrafts in New Zealand since 1960 were held by manufacturing (38 to 53 per cent), commerce and finance (36 to 52 per cent), construction (31 to 47 per cent), and public utilities (55 to 73 per cent). The Australian data (Table 16) relating to the second Wednesday in July and January (1962 to 1967) show that generally agriculture, grazing, and dairying held the lowest proportions of unused overdrafts (20 to 28 per cent), whereas public authorities (66 to 86 per cent), manufacturing (48 to 64 per cent), and finance (47 to 64 per cent) accounted for the highest proportions.

These data show that the urban sector tends to hold a larger share of unused overdraft facilities relative to the rural sector, which might also be the case in other primary producing countries for which similar data are not available. But these data cannot be invoked to support the view that generally the banking system in the less developed countries tends to siphon funds from the rural sector to the urban sector.24 The problem of the flow of funds between the urban and rural sectors is quite complex and would involve, among other things, an exhaustive study of the flow of total resources, including a comparison of total credit and total deposits between regions, before any conclusions could be drawn. It is therefore not possible to argue from individual items of a bank’s balance sheet or from unused limits alone that banking systems typically and deliberately divert savings from the rural to the urban sector.

There are no data on the distribution of unused overdrafts by size of accounts.25 But even assuming that most of these unused overdrafts represent the richer and more influential customers who find it easier to get large and stable overdraft limits, it does not follow that the overdraft technique by itself prejudices the allocation of credit to other customers. This distribution may represent the prudent allocation of the bank funds to customers according to normal qualitative criteria of lending.

Conclusions of the comparative survey of unused overdrafts

The foregoing survey of the size and variations of unused overdrafts and its pattern of distribution (in Australia and New Zealand) suggests the following broad conclusions.

(1) Unused overdrafts amount to a high proportion of total overdraft facilities in all countries except Sweden.

article image

(2) There is no consistent pattern of relationship between the movements in the total sanctioned limits and the ratios of unused overdrafts, except for Sweden, where they tend to move in the same direction (Table 7).

Table 7.

Selected Countries: Variations in Unused Credit, 1950–66 1

article image
Source: Derived from the respective country tables in the Statistical Appendix to this paper.

Column 1 shows year-to-year variations in overdraft limits in absolute values; column 2 shows year-to-year variations in the ratio of unused credit.

(3) Seasonality is the dominant factor accounting for the variations of unused overdrafts in Sweden and New Zealand. This may be presumed to hold good for other countries, too.

(4) The urban sector accounts for larger proportions of unused overdrafts in both Australia and New Zealand than the rural sector.

III. Nature and Causes of Unused Overdrafts

The statistical evidence for the group of countries included in the present study no doubt establishes the extent of unused overdrafts in the banking system. But it is equally important to press the analysis further to find out the real character of unused overdrafts and their underlying causal factors.

The classification of overdrafts (see Section I) points to at least one possible case where an unused limit may not really represent idle credit, namely, when it is used as security for trade credit. Similarly, debtors may create mortgage securities to cover their overdrafts.26 The foregoing types wherein there is an element of “pyramiding” are exceptional and are a small proportion of total unused limits at any given point of time. Consequently, it remains to be analyzed whether the other unused overdrafts represent really idle credit or whether the statistics of unused overdrafts include a large economically fictitious element.

The notion of an overdraft limit is not always precise insofar as there are instances where no firm formal limits are granted by a bank and where, consequently, temporary excesses may be permitted beyond the assigned limit. There is, therefore, a distinction between a “firm credit limit” and a “courtesy” line that represents an informal understanding between banker and customer to extend credit facilities to a particular amount. It is the firm limits representing commitments to be guaranteed that can be considered as legitimate contingent liabilities of banks and are presumably so regarded in financial statements to stockholders and bank examiners. It is these which are aggregated as at a given point of time in banking statistics wherever they are published. They are subject to review periodically, at least once a year, but in practice many limits may remain virtually unchanged from year to year, particularly those to the larger and more influential customers.

Insofar as overdraft limits are firm, they may influence attitudes and policies of bankers and borrowers alike, although perhaps in unequal degrees. Bankers act on the assumption that they are unlikely to be called on to provide simultaneously cash for all overdraft limits outstanding at any point of time.27 In actual fact any attempt by a bank to meet all its overdraft commitments would not be possible without running down its first and second line of reserves or else without massive borrowing from the central bank. Insofar as banks know that only a certain proportion of outstanding overdrafts would be drawn upon, unused overdraft limits can be extended in excess of the amount actually available for lending. This working rule is analogous to the bankers’ objections to aggregating outstanding deposits on the ground that the total of deposits could not be simultaneously withdrawn from the banks. Both reflect the fact that bank policies and operations are based on “averages” rather than “totals.”28

From the borrowers’ side, too, there is a certain element of “water” in the figures of unused overdraft limits inasmuch as customers may not necessarily regard their limits as the likely maximum of funds needed. Rather, they may treat them both as a precautionary balance to be resorted to in an emergency and as a substitute for other financial assets in demonstrating creditworthiness.

Thus the statistics of total overdraft commitments give an exaggerated impression of the “loaned up” position of banks, because although this volume of credit is potentially capable of being utilized, in fact it is not because of established practice and convention. Nevertheless, all unused overdraft limits are not of a purely nominal character, since the availability of unused overdrafts does often determine plans and decisions of borrowers. To that extent a prudent banker will also have to take account of them. While an aggregate of all unused overdrafts in the banking system may not be very meaningful, individual unused positions may have considerable influence on the behavior of individuals or groups of borrowers.

Even allowing for the fictitious element in the over-all total of unused overdrafts, the pertinent questions are why do genuine positions of unused credit emerge at all, and are the underlying causal factors operative on both the lender’s side and the borrower’s.

From the banker’s side the practice of lending “all or nothing”29 in response to a customer’s request for overdraft limits instead of graduating the amount of credit more finely creates an element of “lumpiness” in the credit market. It is, however, quite reasonable for a banker either to grant fully a creditworthy request by a borrower or to refuse it completely. If a banker pares down a borrower’s request it will only assure that the customer obtains inadequate credit, which may lead to the failure of the project and to default on the loan. The sanction of large limits may also be one means of retaining the larger and more influential customers.

The question then arises whether there is any, even approximate, rule of thumb determining the level of overdraft limits. If banks normally adhere to a defined ratio between reserves and actual advances, the same working rules would suggest some upper limits to potential advances (unutilized overdrafts) by banks. These limits could be defined with reference to the level of excess reserves (among other factors, such as total liquid assets) so that one could postulate a twin set of operating rules with actual advances based on minimum reserves and formal overdraft limits related to excess reserves. Total credit limits would then correspond to the over-all level of reserves (the minimum statutory or conventional reserves plus any excess reserves). But the fact that the overdraft system is consistent with a position where credit limits are considerably higher than excess reserves, as in the Philippines, or else with a position of no excess reserves, as in New Zealand, indicates either that unused overdrafts can be counted upon not to result in excess credit or that the reserve position may be less important than supposed as a determinant of overdraft limits.30

The persistence of unused overdrafts even at the peak of economic activity when there is much greater demand for credit, and also during periods of credit squeeze, suggests that they cannot be attributed wholly to lags between approvals and drawings. Equally intriguing is the fact that bankers do not tend to scale down limits even with experience of high and persistent ratios of unused credit. The question as to why drawings are less than sanctioned limits could be answered somewhat along the following lines.

First, borrowers may not always estimate correctly their credit requirements and may therefore tend to ask for higher limits than are strictly necessary. On transactions account, there is always the problem of matching the time pattern of receipts and payments, which consequently necessitates maintenance of a higher average balance. In addition, experience of past credit squeezes and expectations of future ones invariably induce borrowers to ask for higher limits. Thus, purely temporary restrictions on credit may have the effect of increasing the permanent or long-term level of liquidity preference.

The practice of finance companies in Great Britain is a good example of the sort of factors that move borrowers to seek credit limits well above their normal expected requirements both in borrowing directly from the banks and in arranging acceptance credits under which they may draw bills.31 This enables them to switch their borrowing from one source (bank overdrafts, bills discounted, deposits) to another according to their relative costs from time to time and also avoids the necessity of curtailing lending operations by allowing them to call more on one source (e.g., banks) when another source (e.g., deposits) shows signs of thinning out. This is a case where the nature of the borrowers’ business makes it peculiarly advantageous to arrange for higher than normal credit requirements.

The tendency to ask for excessively high limits is perhaps reinforced by the absence of any commitment charges on unused limits, or else because such charges cost the borrower much less than the full rate of interest on bank advances. The fact that unused overdrafts cost less reduces incentives of borrowers to economize on demand for limits.

Second, if the element of seasonality accounts for a large part of the variation in the ratios of unused overdrafts, it also suggests that borrowers tend to ask for limits approximating to the estimated peak of their busy season requirements rather than their average requirements over the year. Also, as bankers follow the practice of annual review of limits, it is advantageous to fix these limits at the highest possible estimate for the year.

The net effect of the factors described above may be to create at least some positions of genuine idle credit, although it is not possible to measure their magnitude. But far from representing loss of potential income, these unused overdrafts may enable the banker to ensure that new earning assets are available to replace payoffs of advances and securities at maturity. A certain volume of unused credit gives the banker a margin of maneuverability in the periodic reshuffle of his liquid and earning assets even as some degree of unused capacity gives a firm more flexibility in production planning. It also affords a cushion against excess or overdrawn positions in other customer accounts.

In view of the oligopolistic character of the banking industry in general, it is also pertinent to pose the question whether unused overdrafts reflect imperfect competition in the credit market and whether they are related to the existence of credit rationing. Insofar as variations in the price of credit (rate of interest) do not equalize the supply and demand for credit, there may arise simultaneously excess capacity (excess reserves) and idle capacity (unused credit) on the supply side and a “fringe of unsatisfied borrowers”32 indicative of excess demand. But the Keynesian fringe of unsatisfied borrowers could be said to exist only if banks were unable to lend more because of insufficient lending capacity, i.e., they would be willing to lend more in the event of an increase in their cash reserves. Credit rationing in this sense is rare in modern banking systems in view of their access to additional lending capacity in the form of reserves of government securities, central bank credit, etc. But credit rationing in the sense of a qualitative screening of credit demand based on evaluation of creditworthiness is of the essence of sound banking and is no indication of imperfection in the credit market. The unsatisfied borrowers are mostly those who fail to measure up to the qualitative standards of creditworthiness laid down by banks. Moreover, unused overdrafts do not inhibit more lending to otherwise eligible borrowers. It is also important to stress that there are good reasons why a bank does not always strive for a “fully loaned up” position, as it has to reconcile the requirements of safety, liquidity, and profitability; if the resulting policy leads to unused overdraft limits or unused lending power (excess reserves), it cannot be regarded as a deviation from long-term profit maximization.

IV. Unused Overdrafts and Monetary Policy

The analysis of the preceding sections suggests that the element of idle credit in unused overdraft limits is apt to be exaggerated. Can it then be maintained that unused overdrafts really hamper the effectiveness of monetary policy?33 An affirmative answer might be somewhat on the following lines. Unused overdrafts represent a ready source of liquidity that can aggravate inflationary pressures, since the initiative for drawing on the limits is largely with the customers. Once the limits are sanctioned the timing and magnitude of customer drawings become an unpredictable element, as it is difficult for a bank to take restrictive actions or even attitudes. This would appear to introduce an element of uncertainty into the formulation and execution of credit policy. But the type of complication that is often attributed to overdrafts is really the lag between the announcement of a tight credit policy and its actual effects owing to the existence of unused limits. To this extent the overdraft system does complicate the mechanics of monetary policy, but it could not really be said to hamper its effectiveness since it is open to the monetary authorities to regulate both limits and advances through periodic review and regulation of outstanding and fresh overdraft limits and the advances against them.

In this respect British experience exposes the limitations of a credit policy that relies mainly on regulation of limits. The Committee of London Clearing Bankers in its evidence to the Radcliffe Committee stated: “The ‘squeeze’ operates mainly through a reduction of ‘limits,’ beneath which the actual borrowing may fluctuate widely. It is therefore by no means certain that a reduction of £X million in sanctioned limits will bring about a reduction of equal size in advances. Indeed an aggregate reduction in limits might take place without any curtailment of actual borrowings at any particular date and vice versa: less availment is made of limits in a period of contracting trade.”34

The level of existing limits and the rate of approval of fresh limits as well as the level of actual advances are all equally important in the formulation of an effective credit policy. If these variables are kept under simultaneous control by the monetary authorities, many of the problems associated with the existence of unused overdrafts would be eliminated or at least reduced to manageable proportions. This calls for adequate statistics of the size, distribution, and time pattern of unused limits and for policy responses that take account of the possibilities of changes in the ratio of unused overdrafts to limits.

V. The Treatment of Unused Overdrafts in Banking and Monetary Statistics

The analysis in Sections III and IV, which broadly suggests that the economic significance of unused overdrafts may be far less than indicated by their size, would also, by the same token, justify their exclusion from the concept of money supply and from systems of financial accounts. Most existing definitions of money supply as used by central banks and by the Fund’s Monetary Surveys in IFS exclude “unused overdraft facilities.” Moreover, balance sheets of commercial banks do not show unused overdrafts even as a contingent liability, like guarantees or acceptances. In fact, commercial bankers35 regard unused overdrafts quite differently from contra-items, like acceptances. There is a definite legal commitment on the part of customers to provide funds on maturity of acceptances whereas the total of unused overdrafts does not commit the banker to necessarily provide these facilities on demand all at once. It is therefore not a significant determinant of the lending banker’s behavior.36 This has its counterpart in company accounts that give no indication of a firm’s liquidity in the form of access to unused bank overdrafts. Similarly, estimates of idle balances37 made in investigations of liquidity preferences do not allow for unused overdrafts. All this follows from the fact that it is difficult to attach any real economic significance to the bare aggregate of unused overdrafts. Although Keynes in the early 1930’s was among the first to raise the question of treating unused overdrafts as part of bank money,38 his general position could be interpreted as a case for understanding the role of unused overdrafts and the need to collect adequate statistics on unused limits rather than a dogmatic insistence on treating them as an integral part of money. This is consistent with the position of the Macmillan Committee, of which Keynes was a member, which was more inclined to accept the commercial banker’s interpretation of unused overdrafts. Perhaps this explains why the Radcliffe Committee did not refer explicitly to unused overdrafts as one of the elements of total liquidity or to their bearing on monetary policy. Among other expert inquiries, only New Zealand’s Royal Commission on Monetary, Banking, and Credit Systems in 1956 made an explicit case (paragraphs 156 and 160 of its Report) for inclusion of unused overdrafts in money supply on the grounds of their immediate availability and general acceptability. But the Reserve Bank of New Zealand justified the exclusion of unused overdrafts from money supply on the ground that “they are not freely transferable from one person to another; some of them are merely nominal and may never be used; and finally many bank customers could obtain unexercised overdraft authorities if they so desired, but this fact is not recorded in the statistics.”39The Irish Commission of Inquiry into Banking, Currency and Credit (see footnote 1), while emphasizing the importance of unused overdrafts, did not make any specific recommendations regarding their definitional or statistical treatment.

Even on statistical grounds the case for including unused overdrafts in the definition of money supply appears questionable. Mr. William White’s study40 of the data for Finland, New Zealand, Norway, the Philippines, and Sweden came to the conclusion that the inclusion of unused overdrafts apparently does nothing to improve the value of money supply data. One of the major statistical findings of this study was that the variability of velocity of circulation consequent upon the inclusion of unused overdrafts was so small relative to the conventional velocity of money (excluding unused overdrafts) as to be unimportant (Table 8).

Table 8.

Selected Countries: Differences Between Year-to-Year Percentage Changes in the Two Velocities of Circulation

(Average percentage change in velocity of money plus unused overdrafts minus percentage change in velocity of money)

article image
Source: William H. White, “Interpreting Monetary Statistics When Overdrafting Is Prevalent” (mimeographed paper, September 20, 1956), Table 1, p. 19.

But, irrespective of the logical and statistical case against inclusion of unused overdrafts in the concept of money supply, it would be useful to have complete data on their size and fluctuations as well as distribution by economic categories and size of accounts, which would facilitate more effective control by the monetary authorities of limits and advances. Moreover, “unused credit lines taken by themselves may still constitute a valuable series for forecasting money demand and money income in the short run, for many of the observed changes in unused credit lines are unconnected with the occurrence of unforeseeable events and are instead symptoms of foreseen future events—planned increases or decreases in bank-financed expenditures over a period starting a few weeks or months after the credit-line adjustment was made.”41 The analysis of the present study, in particular its revelation of the importance of seasonal factors as a causal influence in variations of unused overdrafts, also suggests that such data may have some predictive value.


Table 9.

New Zealand: Unused Overdrafts1

article image
Sources: Reserve Bank of New Zealand, Statistical Summary and Bulletin.

Columns 1, 2, 4, and 6 show the average of monthly figures.

Includes term loans and export finance until April 24, 1963.

Based on average of daily figures for each month. Represents excess of trading bank balances—including borrowing, if any, from the Reserve Bank—over statutory minimum balances.

Average of nine months.

Ten months.

Table 10.

New Zealands Monthly Variations in Unused Credit, July 1959-April 1967

article image
article image
Source: Reserve Bank of New Zealand, Bulletin.

From October 1963 a sum of £NZ 19 million, allocated as a result of redefinition of overdraft arrangements, is included.

Table 11.

New Zealand: Distribution of Unused Overdrafts by Classes of Borrowers, 1960–671

article image
Source: Derived from figures of limits and advances by categories of borrowers in Reserve Bank of New Zealand, Bulletin.

(a) represents limits in millions of New Zealand pounds; (b) represents advances in millions of New Zealand pounds; (c) represents unused overdrafts as percentage of limits.

Table 12.

Sweden: Unused Bank Credit, 1955–66

article image
Source: International Monetary Fund, International Financial Statistics.
Table 13.

Sweden: Unused Overdrafts, 1957–60

article image
Source: International Monetary Fund, International Financial Statistics.
Table 14.

Sweden: Unused Overdrafts, 1961–66

article image
Source: International Monetary Fund, International Financial Statistics.
Table 15.

Australia: Unused Overdrafts of Major Trading Banks, July 1960-January 1967

article image
Sources: Reserve Bank of Australia, Statistical Bulletin (Sydney); Australian Financial Review (Sydney); Commonwealth Treasury, Treasury Information Bulletin (Canberra).

Excludes temporary advances to wool buyers and term loans.