Vol. 6, No. 4
SUMMARY
There is an ongoing debate as to whether Australian banks are earning higher margins (or "spreads") than overseas banks between the interest rates paid to depositors and the rates charged on loans. This debate really concerns whether deregulation has produced as much competition as expected.
Differences between countries' banking systems make it difficult to be conclusive about international differences in banks' margins, although there are some indicators that point to higher margins here.
Moreover, while the underlying profitability of major Australian banks is now not only lower than before deregulation but apparently close to the average for other companies, this does not necessarily mean that the increase in competition has been adequate. There is some evidence that the cost of providing bank services is higher than necessary and that deficiencies in the competitive framework are inhibiting the competing away of such excess costs.
The apparent inadequacy of competition is importantly a product of government regulation and government involvement in the industry, including excessive "prudential" regulation, the existence of government-owned and guaranteed banks, and the protection of existing major banks against take-over.
There may be a case for a new "Campbell Committee" report to assess these (and other) factors inhibiting competition in the banking industry and what might be done to change them. A reduction in the costs of financial intermediation would help encourage investment and generally improve the functioning of the economy.
INTRODUCTION
For some time now, the House of Representatives Standing Committee on Banking, Finance and Public Administration has been investigating the "spread" (or margin) between banks' cost of funds and their lending rates, against the background of claims that Australian banks earn margins that are higher than overseas banks. The main object of this investigation has been to assess these claims, the extent to which financial deregulation has resulted in additional competition and benefits to consumers, and whether such additional competition and benefits are judged to be adequate. At the request of the Committee, I made a submission to it and this paper is an edited version of that submission.
At the outset, it is important to recognise that it is essential to a healthy banking system that banks earn returns on capital comparable with those in other industries. Failure to do so will result in inadequate capitalisation of the banking industry, with potentially adverse prudential and efficiency consequences affecting the economy as a whole. Banks thus need to earn an adequate "margin" of profit.
Further, competition is not limited to price. Indeed, it is apparent that consumers have benefited from an increase in the range of services provided by banks. However, the expectation would be that deregulation should also have led to some increase in price competition.
In its evidence to the Committee, the Australian Bankers' Association (ABA) appeared to suggest that the profitability of banks, rather than their interest margins, may be the best indicator of the state of competition in the banking industry, both domestically and in terms of international comparisons. The ABA's submission to the Committee of 18 May included data showing a declining trend in banks' return on shareholders' funds, both over time and relative to other industries, and also showing that in 1991 Australian banks' profitability was not especially high compared to overseas countries.
Table 1: Return on Shareholders' Funds (%)
| Banks | All company average | |
| 1980 | 15.3 | 12.2 |
| 1981 | 17.1 | 10.7 |
| 1982 | 15.6 | 9.0 |
| 1983 | 14.3 | 8.0 |
| 1984 | 15.8 | 9.5 |
| 1985 | 14.1 | 9.6 |
| 1986 | 11.4 | 10.8 |
| 1987 | 11.3 | 9.4 |
| 1988 | 12.6 | 10.2 |
| 1989 | 11.9 | 11.1 |
| 1990 | 9.8 | 9.0 |
| 1991 | 7.8 | 4.5 |
| 1992 | -4.9 | 1.3 |
Source: ABA Submission (from STATEX database)
There seems no doubt that the decline in banks' profitability in the early 1980s period of deregulation, and immediately after, did reflect increased competition and a move towards more "normal" or average returns on capital. However, returns remained above "normal" up to 1989 and the reductions between 1989 and 1992 are not necessarily an indicator of increased competition since then. Those reductions may simply reflect the effects of the recession and the resultant large increase in bad debts and non-performing loans.
Indeed, financial ratios derived from data in the ABA submission confirm that increased funding for bad and doubtful debts was the major factor in the recent decline in profitability (see Table 2). These ratios also suggest that the 1993 figure for the return on average total assets was back to pre-1989 levels. (1)
Table 2: Australian Banks -- Financial Ratios (%)
| 1987 | 1988 | 1989 | 1990 | 1991 | 1992 | 1993 | |
| Interest income as a % of ATA | 11.77 | 11.27 | 12.64 | 13.67 | 11.30 | 8.81 | 8.35 |
| Interest expense as a % of ATA | 9.66 | 8.82 | 10.10 | 11.32 | 9.07 | 647 | 5.41 |
| Net-interest income as a % of ATA | 2.15 | 2.50 | 2.56 | 2.46 | 2.35 | 2.41 | 2.93 |
| Non-interest income as a % of ATA | 2.67 | 2,03 | 2.05 | 2.19 | 2.23 | 2.37 | 1.56 |
| Gross income as a % of ATA | 4.82 | 4.95 | 523 | 5.17 | 4.59 | 4.83 | 4.49 |
| Operating expenses as a % of ATA | 3.59 | 3.46 | 3.82 | 3.88 | 3.40 | 3.14 | 2.92 |
| Funding for bad and doubtful provisions as a % of ATA | 0.25 | 0.32 | 0.84 | 1.11 | 0.94 | 0.81 | 0.46 |
| Abnormal item as a % of ATA | 0.00 | 0.03 | -0.21 | -0.77 | -0.38 | 0.05 | -0.06 |
| Income tax as a % ol ATA | 0.44 | 0.49 | 0.15 | 0.36 | 0.17 | 0.35 | 0.43 |
| Profit after tax as a % of ATA | 0.55 | 0.80 | 0.21 | -0.96 | -0.29 | 0.58 | 0.63 |
Source: Australian Bankers Association
Note: "ATA" = Average total assets
Data comparing net and "gross" interest spreads (gross spreads exclude the effects of interest on non-performing loans) also confirm that the loss of interest on non-performing loans was the main factor in the recent decline in profitability. Thus, gross spreads have been relatively stable since 1982 (see Chart 1).
This indicates the need to examine not simply the trend in "margins" (of which "profitability" is one such) but the factors behind such trends. Moreover, any international comparisons would need to examine developments over a period of time, as well as taking account of the extent of competition in overseas countries themselves.
There is also a question as to whether an assessment of profit margins can itself provide definitive conclusions in regard to the extent of price competition. Given the need for banks to earn an adequate profit, the same profit margin, for example, may be the product of, on the one hand, high costs and high charges that reflect an oligopolistic situation which results in an inefficient use of resources and, on the other hand, low costs and low charges that reflect a more competitive market. Clearly, the nation benefits if banking services of the same quality can be provided with resources equal to (say) only 5 per cent of GDP rather than (say) 6 per cent of GDP even though the banks' profit margin may be the same in each case.
In this regard it is noted that the managing director of Westpac Banking Corporation, Mr Robert Joss, was reported in the Australian Financial Review of 26 August 1993 as claiming that banks ought to be able to service their client base for $500 a head compared with the present cost of $700 a head, or $13 billion. This implies that there could be potential for banks to reduce costs by up to 40 per cent or by $3.5-$4 billion, which compares with net interest income earned by the 4 major banks in the year ended September 1993 of just over $12 billion, or 3 per cent on average total assets. If such cost reduction potential were to be realised, there would also be potential to reduce net interest income without reducing profit margins.
Accordingly, while the attachment examines the various definitions of margins and the trends therein, the main focus of this paper is on the factors affecting banks' costs and inhibiting the potential for reducing those costs.
INTERNATIONAL COMPARISONS OF MARGINS
A major influence in the debate on banks' margins has been evidence that Australian banks earn a bigger spread than some overseas banks between the interest they pay depositors and the interest they charge on loans. However, all international comparisons have to be approached with considerable caution because of differences in circumstances between countries, and between industries in different countries. This is certainly true in banking where there are significant differences between the structure of banking systems and the roles played by banks relative to other financial institutions and other parts of the capital market. In particular, there are significant differences between banking systems in most European countries and the Australian system.
In his letter to the Committee of 14 December 1993, the Governor of the Reserve Bank argued that reliable data permitting international comparisons of bank margins "do not exist". He noted that data compiled by Salomon Bros. and the OECD for 1990 showed Australia at the top in the Salomon Bros. figures but more in the middle of the field in the OECD figures. He pointed out that the Salomon Bros. figures excluded bank bills from banks' assets and that, once adjusted for that omission, they were broadly similar to OECD figures.
However, the Governor's comments are not necessarily the end of the matter. Indeed, the fact that the Reserve Bank is now assembling a detailed database on net interest spreads of banks in the OECD area, cross-classified by various attributes and dimensions of banking, indicates that it may be possible to develop international comparisons that will give a better indication of relative levels of competitiveness.
It is also evident that individual Australian banks are seeking to benchmark themselves against "international best practice", which implies that international comparisons are being made by those banks themselves. Indeed, the major banks publish comparisons between their overseas and Australian operations. While these overseas operations are relatively more involved in the wholesale market where spreads and margins are necessarily lower, the recent extent of these differences should be noted.
Table 3: Interest Spreads and Margins -- Australian and Overseas Operations of Major Banks
| Half Year to Mar 94 % | Half Year to Sep 93 % | Half Year to Mar 93 % | ||
| National Interest Spread Interest Margin | Australia Overseas Australia Overseas | 4.3 3.3 6.2 3.7 | 4.5 3.0 5.6 3.4 | 4.0 3.1 5.4 3.4 |
| Westpac Interest Spread Interest Margin | Australia Overseas Australia Overseas | 3.1 1.9 4.0 2.2 | 2.8 1.7 3.7 2.1 | 2.6 1.8 3.6 2.0 |
| ANZ Interest Spread Interest Margin | Australia Overseas Australia Overseas | 3.3 2.1 4.1 2.5 | 3.3 1.7 4.1 2.2 | 2.9 1.7 3.7 2.1 |
Note: Interest spreads are after allowing for interest forgone on non-accrual loans.
It also seems clear that many of those involved in attempting to assess internationally the relative strength and profitability of Australian banks have concluded that their capacity to generate higher spreads is considerable. In Moody's "Australian Banking System Study" of April 1994, for example, it is suggested that
The core earning capacity of major Australian banks is very good, primarily because of the broad net interest margin they can generate in their local market. Australian banks' spread between cost of funds and yield on loans is high, even compared with those of other very profitable banking systems in the US, Canada, and the UK.
and
The banks' interest rate spreads in residential mortgages are also quite good. Excluding a "teaser" rate with a term of one year, residential mortgages generate a spread that falls between 3 per cent and 4 per cent; by comparison, the UK spreads in this area run 2 per cent, and in Canada, they have been between 2 per cent and 3 per cent. The attractiveness of Australian residential mortgage spreads is magnified by the fact that losses associated with residential mortgages are minuscule and that residential mortgages only require half the regulatory capital needed to support a commercial loan.
Presently, the Australian banks' net interest margins are constrained because of the cost to fund a heavy load of bad loans. This cost should ebb with the expected decline in non-performing assets.
INFLUENCES ON BANKS' COSTS
As noted, in considering whether competition in the banking industry is adequate, it is necessary to go beyond an examination of margins per se. The following examines factors influencing banks' costs.
BANKS' BRANCH NETWORKS
Before deregulation, when there was little scope for competing on price, the major method of competing at the retail level was through a branch network. This almost certainly led to the establishment of a considerable number of branches which were uneconomic but whose costs were able to be recovered either through "excess" spreads or by having higher lending and deposit rates than were necessary. The existence of such an extensive network of branches at the time of deregulation undoubtedly made it more difficult for a new competitor to establish itself as, without undertaking a take-over, such a competitor faced the prospect of establishing a network of additional uneconomic branches and an extended and expensive price war to obtain market share. (2)
With the development of new technology, and with the associated lesser reliance on the use of cash, it has become increasingly practicable and efficient to conduct "banking business" outside normal banking hours and outside bank branches. Facilities such as ATMs, EFTPOS, Credit Cards and banking via telephone are examples.
However, even with these technological developments and the reduced physical contact required with banks, deregulation has been accompanied by little rationalisation of bank branches. Indeed, the total number of bank branches has actually increased from 6,239 in 1983 to 7,064 in 1993 and the notional number of the Australian population "served" per branch has increased only fractionally (see Table 5 attached on Bank Branches and Population). (3) While this increase of 825 in total bank branches reflects the establishment of new banks (mainly those that were previously building societies, which now account for around 800 bank branches), the very small reduction in bank branches of the four major banks plus the State Bank of Victoria -- from 5,641 in 1983 to 5,468 in 1993 (about 3 per cent) -- is one indication that banks have not been under great competitive pressure to reduce costs.
The position of the Commonwealth Bank is particularly noteworthy in this regard: since acquiring the State Bank of Victoria, the Commonwealth Bank appears to have reduced its total branches only from 1,791 to 1,756. This failure of the Commonwealth to effect any substantive rationalisation of its branch network is relevant to the point made below about the adverse effects of government intervention in the banking industry. Given the relative importance of the Commonwealth in the banking system, the fact that it has been inhibited in implementing efficiency measures has also had implications for the policies of other major banks.
OTHER COSTS
The foregoing is not to suggest that the major banks have not taken steps to improve productivity and reduce costs. There has, in particular, been a substantial reduction in employees by the four major banks: according to KPMG figures, between 1990 and 1993 their employees were cut by over 23,000 or almost 13 per cent (see attached Table 6 on Banking Employees, Productivity and Margins) and this figure has probably increased further since then. (4) However, this reduction may be more a reflection of the effects of recession and the increase in non-performing loans rather than of competition forcing a reduction in costs. The saving from a 23,000 reduction in employees is probably of the order of $700-$900 million p.a., or only about 5-7 per cent to total operating expenses.
In the ABA submission to the Committee (page 9) it is suggested that the ratio of banks' operating expenses to operating income was reduced from 78 per cent in 1990 to around 62 per cent in 1993. This contrasts, however, with KPMG figures indicating that the ratio in 1993 was not noticeably different from pre-recession levels (see attached Table 6 on Banking Employees, Productivity and Margins). Further, even with the large reduction in employees of the four major banks plus the SBV, each employee of those banks was notionally "serving" fewer of the Australian population in 1993 than in 1983: 921 compared to 1,154; that is, there is an implied reduction in productivity.
The ABA acknowledges "the need to make further cost and productivity improvements to achieve 'world best practice' " (5) and it is noted that there now appears to be a considerable difference between the best practice Australian bank (National) and other major banks (see attached Table 6 on Banking Employees, Productivity and Margins). If the other 3 banks were to have achieved the same 1993 operating expense/operating income ratio as the National, their operating expenses would have been about $1.5 billion lower in total. This lends weight to the comments by Mr Joss to which reference has already been made.
It is important in this context to recognise that the banks are operating within an industrial relations system which gives trade unions considerable capacity to resist or slow changes in work practices and which also gives unions greater capacity to obtain employee benefits from productivity improvements rather than to have such improvements reflected in lower costs for the benefit of bank customers/shareholders. Without access to details of industrial negotiations between individual banks and the respective unions, and to details of wage increases and other changes consequential on award restructuring and enterprise bargaining, it is difficult for an outsider to assess the extent of the constraints under which banks operate.
Anecdotal evidence suggests, however, that they may be of not inconsiderable importance in slowing the process of improving efficiency of use of resources by the banking sector, particularly in circumstances where competitive pressures are reduced by other factors discussed below. It is noted, in particular, that banks list the "willingness of employees to accept change" as an important factor in determining the pace and extent of reform and, while this is obviously a matter that banks have to take into account, it may be carrying more weight than is appropriate in the interests of improving efficiency. It is surprising that neither the ABA (6) nor the Reserve Bank have referred to possible industrial relations constraints in their submissions to the committee.
GOVERNMENT REGULATION OF AND INTERVENTION IN THE BANKING INDUSTRY
There is no scope to examine in this Backgrounder the full extent of government regulation of, and intervention in, the banking industry and the implications of that for the extent of competition. I suggest, however, that the effect of such government involvement is to restrict severely competition and to create a situation which provides reduced incentive for the major banks to effect structural changes.
By way of background, economic textbooks tell us that, to establish a perfectly contestable market, it is necessary to create conditions in which firms can enter and, if they choose, exit without losing the money they invested. (7) Such circumstances can ensure adequate competition because if higher than average returns are being earned on capital, new firms will enter and drive down prices to the point where "excess" profits are eliminated. Alternatively, if "normal" returns are being earned but cost inefficiencies exist, new firms will enter and provide the same product at lower cost and lower prices.
The Australian banking industry more closely resembles an oligopoly than one where perfect contestability exists. While new entry is possible, it is in practice very difficult for new entrants to become substantive participants by engaging in strong competition to obtain market share, without the risk of losing their capital or a substantial part of it. A major reason for this is that the existing four major banks operate in an environment which is regulated with a view to protecting the depositors of existing banks. Indeed, on one interpretation of statements by the present Governor of the Reserve Bank, it could be said that depositors have a de facto guarantee against loss. Thus, while the previous Governor stated in 1985 that "the Banking Act is silent in regard to the priority to be given to depositors. So the legislation is less than a guarantee to depositors of full repayment", (8) in the case of each press statement on the "runs" in 1990 on the Bank of Melbourne and Metway Bank, the present Governor indicated that the Reserve Bank would ensure that depositors are "fully protected". (9)
The effect is to create a considerable degree of inertia amongst the customers of the respective banks such that it requires a major deterioration in the performance of an existing bank before there is a substantial switching of custom. This situation inevitably operates to reduce the incentive for the banking industry to engage in competitive activity. That reduced incentive is reinforced by the fact that the existing four major banks are operating in a market in which they are effectively protected under foreign investment policy from take-over by a major foreign bank (this would be deemed against the national interest), as well as being effectively protected under domestic competition policy from take-over by each other. In any event, the prohibition under the Banks (Shareholding) Act against any one shareholder holding more than 14.99 per cent of the capital of an Australian bank makes it difficult for a take-over to be mounted other than by a bid mounted from a minority shareholding position and involving a very substantial outlay of capital. This also means that existing boards and managements of the major banks are protected against removal by dissident shareholders except in unusual circumstances.
The foregoing combination of circumstances produces the situation, not uncommon in an oligopoly, where tacit collusion tends to occur, so that a follow-the-leader type approach tends to develop in both pricing and the introduction of new products. This reflects the fact that for any one bank to try to obtain an increase in market share through a deliberate strategy of stepped-up and sustained competition would almost certainly require a major and extended programme of price competition that would be extremely costly and which might have only limited success. In those circumstances the alternative of continuing to earn normal profits and not upsetting the apple cart is a relatively attractive one.
GOVERNMENT-OWNED AND GUARANTEED BANKS
Another important way in which government involvement in the banking industry affects competitive conditions is through the ownership of banks and the provision of specific guarantees for depositors of those banks. In addition to the banks still owned and operated by the States of NSW, South Australia, Western Australia and Tasmania, the Commonwealth Government holds a majority interest in the Commonwealth Bank and it has included a specific guarantee in legislation passed (ironically) with the avowed intention of modernising the Bank so that it could compete in a deregulated market. As at 30 June 1993 these government-owned and -operated banks with specific government-backed guarantees accounted for 28.5 per cent of total $A assets of the banking system.
As with most other government-owned and -operated enterprises, these banks have a history of operating with higher cost structures and paying less attention to return on capital than the private sector banks. The Commonwealth Bank has, as already noted, a very large branch network and it is the bank with the largest $A assets.
It is evident that these banks have, at least up until very recently, operated with a significant political "brief", albeit rarely specifically stated publicly. Their political masters tended to encourage or condone activities which were uneconomic but whose cost was borne by the taxpayer in the form of lower returns to Consolidated Revenue on assets employed. Governments have also, at least until recently, been able to avoid proper capitalisation by relying on the government guarantee as a substitute for capital.
The existence of a situation where a substantial proportion of the assets of the banking industry has been operated under a government guarantee, but with less concern regarding returns, has clearly reduced the capacity for the other major banks to engage in active competition. The government banks could more afford, in a sense, to be "loss leaders" as the lower returns would not necessarily affect their standing. While the "loss leader" behaviour of some State Banks in the late 1980s eventually got them into serious trouble, their activities also had flow-through implications for the behaviour of other banks.
MONETARY POLICY
An important element of the banking industry's operations is the assessment of risk and the appropriate risk margins to be applied to the provision of finance to individuals and companies. This margin will obviously be affected both by past experience with the operation of monetary policy and by the banks' assessment of the capacity of the monetary authorities to operate policy in the future so as to reduce fluctuations in the economy and, in consequence, also to reduce bad debts and non-performing loans.
In this regard, there can be little doubt that the extremely poor performance of the authorities in the operation of monetary policy in the 1980s has added to risk margins and to the costs of financial intermediation generally. (It is also true that monetary policy was asked to bear too much of the burden of keeping domestic spending within reasonable bounds.) Of course, through an excessive easing in the criteria for assessing credit-worthiness the banks themselves were partly to blame for the problems which emerged. However, there remains a need for the authorities to make an unequivocal commitment to the maintenance of low inflation as the primary goal of monetary policy and to provide clearly and unequivocally for the operation of policy to be free of political interference.
CONCLUSION
There is no doubt that the removal of most controls over price and quantity in the banking industry (10) has led to increased competition which has resulted in some reduction in the profitability of the banking industry, and some reduction in (gross) interest spreads, by comparison with the pre-deregulation period. Abstracting from the recession-induced increase in bad debts and non-performing loans, the reduction in profitability does not appear, however, to have led to below-"normal" returns and the industry has continued to be able to raise substantial amounts of capital from the market. This capacity, of course, is important to the ongoing health of the industry and the financial system generally.
The question remains whether the increase in competition has reduced the cost of financial intermediation to the extent that might be expected in a so-called deregulated market. Such cost reductions need not, of course, have any effect one way or another on profit margins or even interest spreads. At a minimum, one might have expected that a deregulated market would have resulted in some narrowing of gross interest spreads in the post-1982 period and a much greater emphasis on charging for banking transactions. This has not happened.
My assessment, which must of necessity involve a considerable judgemental component, is that the reduction in costs has also been much less than it should have been, with the result that Australia remains "over-banked" and has more resources than it should in the industry. This situation exists primarily because of the very substantial degree of regulation of the industry, the large direct government involvement in the industry, and the provision of specific government guarantees to government-owned banks. Australia's industrial relations system and the preferential position given to trade unions have also probably inhibited cost-reducing reforms.
To the extent that government regulation involves prudential requirements dealing with capital adequacy and the like, the Australian industry is in a similar position to its counterparts overseas and there may be only limited scope to reduce such regulation so as to increase the potential for competitive forces to operate in the industry. Even so, there appears to be scope for basing government regulation more on underpinning the financial system and acknowledging that individual banks and even their depositors should have the potential to experience losses. The risk of "contagion" is overstated. Provided the central bank ensures an adequate supply of liquidity for the system as a whole, situations where individual financial institutions experience serious difficulties should be capable of being handled by allowing such institutions to be taken over without threatening the financial system as a whole. Moreover, the extensive regulation of entry and takeover warrants careful review. For example, the objective of the Banks (Shareholding) Act limit on ownership by an individual/company might better be met by relying on the prudential requirement limiting the proportion of a bank's lending to one individual or company. Equally, it may be that, rather than reducing competition, allowing a take-over of an existing major bank by another major bank or even a foreign bank might actually lead to increased competition for market share, as well as for improved management.
The privatisation of the government-owned banks, and the removal of their government guarantee, would also help create level playing field conditions that are more likely to be conducive to competition to reduce costs.
There may be a case for a new "Campbell Committee" report to assess these various factors inhibiting competition in the banking industry and what might be done to change them. Such a Committee could take the opportunity to review the increased regulation and supervision of financial institutions since the 1980s, which has been based on the accepted wisdom that the over-borrowing and over-lending that occurred in the 1980s could have been prevented by greater regulation, and a failure to recognise that the heart of the problem lay with the inadequacies of monetary policy.
ATTACHMENT -- BANK INTEREST MARGINS
Confusion often arises because, while there is a variety of possible definitions of bank interest "margins", commentators frequently do not make it clear which margin they are talking about. Moreover, some comparisons which are made are inappropriate. It is not, for example, meaningful to compare the interest rate on a selected deposit and a selected loan product as banks' funds come, in effect, from a general pool.
However, it may be possible to draw some conclusions from the margin between the average total interest rate paid by banks and the average rate on a selected loan product. For example, the fact that after the de-control of new housing lending in 1986 the margin of around 4 per cent for the major banks between average housing rates and the average total interest rate paid did not fall prior to 1992-93 suggests inadequate competition, given the relatively low risk on housing loans. (11) Since 1992-93 the availability of a lower fixed rate for the first year of a housing loan (which reportedly now constitute about half of new housing loans) has somewhat reduced the average margin on housing loans (see Chart 2 on Housing Margins). (12) This recent development is obviously a reflection of greater competition amongst financial institutions lending for housing, recently highlighted by the entry into the home mortgage market of National Mutual and the ACTU. However, this burst of competition emerged at a time when the demand for other loans from banks was very weak and it remains to be seen whether it is sustained. Moreover, the new fixed-rate bank loans would constitute only a small proportion of banks' total outstanding housing loans.
Comparisons between the average total interest rate paid and average rates on selected loan products may thus provide some indication of the extent of competition. It is understood that banks argue that this is not necessarily the case and that, in determining rates for particular products, regard is paid by them simply to conditions in the market for that product. However, while very competitive conditions in the market for a particular product may dominate in the short term, beyond the short term regard must obviously be paid to the average cost of funds and the margin thereon. If the margin is too low to be sustained given overall profitability objectives, banks will have to increase the margin or take other compensatory action.
In fact, one of the tests of whether there is effective competition is whether attempts are being made to reduce margins on particular products. In a competitive market the margin between banks' average cost of funds and their lending rates on various products should constantly be subjected to competitive assessments in the light of possible changes in risk and technology, and to attempts by individual institutions (both within and outside the banking sector) to acquire market share of one particular product or another.
It is also important to note that changes in the margin on a particular product may reflect the elimination of cross-subsidies which existed in a previously regulated market. The narrowing of margins on large business loans, and the widening of margins on small business loans, are good examples of this phenomenon.
Comment is also sometimes directed to the margin represented by the additional interest payable over the reference rate published by a bank for a variable loan product. This margin reflects the banks' assessments of the relative riskiness of individual borrowers for a particular loan product and is not therefore necessarily an indicator of the state of competition. It is noted in this regard, however, that the range of margins as between smaller and larger businesses in the ABA submission appears, on the surface, surprisingly small (see Table 4 below).
Table 4: Weighted Average Interest Rates, December 1993 (per cent)
| Smaller Business | Larger Business | All Business | |||||
| Under $10,000 | $100,000 to $500,000 | Average | $500,000 to $2m | Over $2m | Average | Average | |
| Variable rate loans | 11.1 | 10.7 | 10.9 | 10.3 | 9.8 | 10.0 | 10.5 |
| Fixed rate loans | 10.4 | 9.7 | 10.0 | 9.7 | 8.3 | 8.7 | 9.2 |
| Bill Facilities | 8.3 | 8.2 | 8.2 | 7.9 | 6.6 | 6.9 | 7.2 |
Source: ABA Submission
It could be, of course, that larger businesses are still cross-subsidising smaller ones to some extent, as was the case before financial deregulation. However, given the considerable potential for larger businesses now to go direct to the market for funds, it seems unlikely that any remaining cross-subsidisation is significant. A possible alternative explanation might be that the narrowness of the margin range is an indication of inadequate competition in lending to smaller businesses, so that the riskier propositions are not accommodated.
Another margin used by analysts is the difference between interest received and interest paid as a proportion of total assets. Table 2 shows the trend in this margin between 1987 and 1993. However, while it may be useful to examine changes in this margin over a short period, the potential for the composition of assets to shift as a result of changes in capital or off-balance sheet transactions (such as bills) may cause this margin to vary over time for reasons unconnected with profitability. It is, again, not necessarily a good indicator of competition.
Perhaps the most accurate indicator of the extent of competitive pressures is the difference between the average total interest received and the average total interest paid. This margin is usually known as the interest "spread".
However, as it does not include income from fees and charges, it is possible that any change in the interest spread, or any absence of change, could be offset by a relatively higher or lower proportion of income coming from fees and charges. A comprehensive assessment of changes in competitive pressures would thus need to include an assessment of changes in fees and charges as well as in net interest spread on the income side.
It is noted that the "gross" interest spread (that is, the net spread plus interest on non-performing loans) has been relatively stable since 1982, although there has been a relatively small reduction by comparison with the pre-1982 period (see Chart 1 on Domestic Interest Spreads). Prima facie, one might have expected, in the more deregulated environment since 1982, some narrowing of this spread and, to the extent that this was reducing profits below normal rates, a compensating adjustment elsewhere, such as through increased income from fees or reductions in net operating costs. There is also a question as to whether the much lower level of inflation might be expected to have reduced the nominal spread.
Of course, banks point out that there have been factors (such as the removal of controls on rates for "new" housing loans, the "loss" of a substantial volume of non-interest bearing deposits, and the "take-over" by them of riskier customers of finance companies) which have either added to costs or removed an artificial barrier on charges and have therefore tended to "push" margins higher.
However, it could equally be pointed out that deregulation should have reduced margins on, for example, large business loans. More generally, if margins have tended to widen in some areas as a result of deregulation they should have tended to narrow in other areas unless the industry is in a position to establish higher-than-average overall profitability. It is submitted that whatever may have happened in the case of the market for some particular products does not deny the general proposition that overall gross spreads should have narrowed.
* * *
This Backgrounder is a slightly edited version of our submission on bank interest margins to the House of Representatives Standing Committee on Banking, Finance and Public Administration.
ATTACHMENT
Table 5: Bank Branches and Population
| Branches | ||||||||||||
| Major Banks | 1983 | 1984 | 1985 | 1986 | 1987 | 1988 | 1989 | 1990 | 1991 | 1992 | 1993 | |
| ANZ | 965 | 972 | 990 | 1109 | 1105 | 1092 | n.a. | 1092 | 1094 | 1260 | 1209 | |
| CBA | 1025 | 1019 | 1015 | 999 | 937 | 971 | n.a. | 936 | 1813 | 1806 | 1756 | |
| (1287) | (1280) | (1278) | (1277) | (1273) | (1272) | (1261) | (1258) | |||||
| WBC | 1530 | 1389 | 1330 | 1315 | 1306 | 1281 | n.a. | 1301 | 1301 | 1299 | 1255 | |
| NAB | 1314 | 1321 | 1243 | 1331 | 1286 | 1280 | n.a. | 1286 | 1273 | 1261 | 124S | |
| SBV | 545 | 547 | 538 | 532 | 533 | 534 | n.a. | 533 | - | - | - | |
| Total | 5379 | 5248 | 5116 | 5286 | 5217 | 5158 | n.a. | 5148 | 5481 | 5626 | 5468 | |
| (5641) | (5509) | (5379) | (5564) | (5563) | (5459) | n.a. | (5470) | |||||
| Other Banks | 1143 | 1130 | 1320 | 1446 | 1639 | 1697 | n.a. | 1960 | 1436 | 1294 | 1596 | |
| Total | 5977 | 5831 | 5898 | 620O | 6323 | 6321 | n.a. | 6575 | 6917 | 6920 | 7064 | |
| (6239) | (6092) | (6161) | (6478) | (6609) | (6622) | n.a. | (6897) | |||||
| Population | ||||||||||||
| 1983 | 1984 | 1985 | 1986 | 1987 | 1988 | 1989 | 1990 | 1991 | 1992 | 1993 | ||
| Total (000s) | 15394 | 15579 | 15768 | 16018 | 16264 | 16532 | 16614 | 17065 | 17284 | 17483 | 17622 | |
| Branches per Million | (405) | (391) | (390) | (404) | (406) | (400) | (404) | 40O | 396 | 401 | ||
| Population Per Branch | Total | 2576 | 2672 | 2677 | 2584 | 2572 | 2615 | 2595 | 2499 | 2526 | 2495 | |
| (2467) | (2557) | (2562) | (2473) | (2461) | (2497) | (2474) | ||||||
| ANZ | 15952 | 16027 | 15947 | 14443 | 14718 | 14961 | 15627 | 15798 | 13875 | 14575 | ||
| CBA | 13613 | 15288 | 15554 | 16034 | 16478 | 17025 | 18231 | 9533 | 9680 | 10035 | ||
| (11961) | (12171) | (12354) | (12543) | (12776) | (12997) | (13565) | ||||||
| WBC | 10061 | 11215 | 11870 | 12180 | 12453 | 12905 | 13116 | 13285 | 13458 | 14041 | ||
| NAB | 11715 | 11793 | 12701 | 12034 | 12646 | 12915 | 13269 | 13577 | 13864 | 14120 | ||
| SBV | 28245 | 28480 | 29345 | 30109 | 30514 | 30958 | 32016 | |||||
| Others | 13468 | 13786 | 11960 | 11077 | 9231 | 9741 | 8706 | 12036 | 13510 | 11041 | ||
Source: ABS and Reserve Bank. Figures in brackets for CBA are supplied by CBA and include "sub branches or service centres" wnich did not have lending facilities.
Table 6: Banking Employees, Productivity and Margins
| Employees | |||||||||||
| 1983 | 1984 | 1985 | 1986 | 1987 | 1988 | 1989 | 1990 | 1991 | 1992 | 1993 | |
| ANZ | 25951 | 36789 | 38031 | 39018 | 41187 | 42445 | 44023 | 48182 | 46261 | 43977 | 41737 |
| CBC | 31713 | 35043 | 36853 | 37495 | 31388 | 35109 | 37114 | 37957 | 48267 | 45790 | 42329 |
| NAB | 22137 | 22474 | 20540 | 21060 | 24333 | 34645 | 34695 | 40997 | 39911 | 38377 | 43053 |
| WBC | 34285 | 35556 | 37635 | 37876 | 37282 | 46506 | 44925 | 45395 | 42431 | 39253 | 33724 |
| SBV | 7500 | 7766 | 7834 | 9270 | 10714 | 10372 | 11253 | 11496 | |||
| Total | 121626 | 137628 | 140890 | 144719 | 144904 | 169077 | 172010 | 184027 | 176870 | 167397 | 160843 |
| Productivity (Operating Expenses/Operating Income) | |||||||||||
| 1986 | 1987 | 1988 | 1989 | 1990 | 1991 | 1992 | 1993 | ||||
| ANZ | 74.2 | 67.1 | 70.8 | 73.3 | 72.2 | 67.6 | 73.0 | 74.6 | |||
| CBC | 75.4 | 80.1 | 71.3 | 67.7 | 61.7 | 50.9 | 67.7 | 71.9 | |||
| NAB | 69.6 | 68.3 | 66.6 | 61.5 | 58.2 | 58.4 | 60.8 | 58.6 | |||
| WBC | 74.2 | 72.2 | 66.1 | 72.0 | 59.2 | 59.9 | 82.0 | 66.1 | |||
| Total | 73.3 | 71.6 | 68.6 | 68.8 | 62.7 | 59.1 | 70.5 | 67.0 | |||
| All Banks | 72.2 | 69.7 | 61.7 | 64.9 | 61.4 | 70.8 | 66.5 | ||||
| (3.2) | (3.3) | (3.1) | (3.1} | (3.1) | (3.3) | (3.1) | |||||
| Productivity -- Operating Income Per Employee | |||||||||||
| 1989 | 1990 | 1991 | 1992 | 1993 | |||||||
| ANZ | 83 | 91 | 100 | 101 | 104 | ||||||
| CSC | 74 | 89 | 119 | 99 | 99 | ||||||
| NAB | 100 | 123 | 126 | 136 | 152 | ||||||
| WBC | 87 | 114 | 124 | 113 | 122 | ||||||
| Total | 86 | 104 | 117 | 112 | 119 | ||||||
| Net Interest Income/Average Total Assets | |||||||||||
| 1985 | 1986 | 1987 | 1988 | 1989 | 1990 | 1991 | 1992 | 1993 | |||
| ANZ | 3.14 | 2.81 | 2.65 | 2.92 | 2.96 | 2.65 | 2.63 | 2.44 | 2.53 | ||
| CBC | 3.62 | 3.83 | 3.56 | 4.00 | 3.89 | 3.74 | 3.79 | 3.68 | 3.37 | ||
| NAB | 3.03 | 2.87 | 3.09 | 4.21 | 3.84 | 3.79 | 3.78 | 3.66 | 3.72 | ||
| WBC | 3.74 | 3.52 | 3.52 | 3.44 | 2.97 | 2.93 | 2.00 | 2.38 | 2.44 | ||
| Total | 3.38 | 3.25 | 3.21 | 3.64 | 3.34 | 3.22 | 3.18 | 3.00 | 3.01 | ||
Source: KPMG Peat Marwick -- Various annual surveys of the performance of financial institutions. Figures in brackets are for operating expenses as a proportion of average total assets.
ENDNOTES
1. However, the 1994 Financial Institutions Performance Survey by KPMG shows that the 1993 return on net assets was still lower than pre-1989 levels.
2. Of course, a number of new banks have been established with retail branch networks. However, most of these have been conversions from building societies. The new foreign banks have fewer than 100 branches between them and hold only about 10 per cent of total $A assets of the banking system.
3. Using the bracketed figures for the pre-1990 period.
4. It is noted that the ABA submission of 17 May 1994 refers to a reduction of 31,500 since 1990 for all banks.
5. Evidence to Committee, 18 May 1964, page 62.
6. It is understood, however, that the ABA itself has no responsibility for industrial relations, which is handled by individual banks.
7. Unlike a perfectly competitive market, a perfectly contestable market does not require, therefore, a considerable number of sizeable participants.
8. "The Prudential Supervision of Banks" by R.A. Johnston, RBA Bulletin, March 1985.
9. See press statements by the Governor of 16 July 1990 and 3 October 1990.
10. A control still exists on interest rates on "old" (pre-1986) housing loans.
11. Some initial increase in the margin on housing loans after decontrol was to be expected, of course.
12. The average margin shown in Chart 2 for 1992-93 is that for variable rate housing loans only. The shaded section shows the average margin on fixed rate loans.
