Report on Certain Issues Arising out of the Report by the Competition Commission on the Supply of Banking Services by Clearing Banks to Small and Medium-Sized Enterprises
By Sir Bryan Carsberg
1. The Competition Commission has prepared a report on the supply of banking services by clearing banks to small and medium-sized enterprises (SMEs) in accordance with a reference made on 20 March 2000. As part of its investigation, the Commission considered whether the banks were making excessive profits, that is profits significantly in excess of their cost of capital. It found that the four major banks were earning excessive profits on their SME business. I have been asked to give my views on the methodology used to measure those profits and its application to the case under consideration. This report sets out my views.
Usefulness of Accounting Measures
2. The Commission has based its study of profitability on conventional accounting information supplied by the banks, calculating rates of return as the ratio of profit to equity capital employed and comparing these rates with the costs of equity capital calculated in accordance with the capital asset pricing model. Accordingly, I begin with some general comments on the use of accounting information for measurement of rate of return. The economic theory which underlies investment decision making, valuation and related matters rests on a model that sees the objective of economic activity as the generation of potential for consumption. It therefore focuses on the cash effects of the activity (or cash equivalents) - their amounts and timing - and leads to the conclusion that the most satisfactory basis for assessing rate of return is a calculation of the internal rate of return on a discounted cash flow basis. For a whole business, this is possible only after the business has come to the end of its life and been wound up and is therefore not suitable for the kind of investigation under consideration.
3. Accounting methods are intended to provide a basis for measuring profitability on a periodic basis. They involve recognising and measuring assets and liabilities. Equity capital employed is defined as assets minus liabilities and profit is the change in equity over a period, adjusted for capital introduced or withdrawn. In concept, assets are regarded as access to future benefits (as accepted in the Commission's report and used by it in its analysis): future cash flows or cash equivalents. This provides a theoretical link with discounted cash flow analysis. Assets are initially recognised at cost although they may subsequently be remeasured (revalued). The process of asset measurement in practice varies according to the purpose to which the measurement will be put. For example, when accounting is used to report on performance to a company's shareholders, a cautious view is taken in the measurement of assets: if great uncertainty exists as to whether an asset exists or has value, the asset will be omitted from the report or given a low measurement. This is because the undesirable effects that may follow overvaluation of assets are thought to outweigh the undesirable effects from undervaluation. A different view may be taken for uses such as the assessment of profitability for purposes of a monopoly investigation.
4. Whatever view is taken of the details of asset measurement, the accounting assessment of rates of return must remain imperfect. This can be seen by considering the case of a new business which involves significant initial investment and requires time to establish its reputation and build up its customer base. Even though the tangible and intangible assets acquired will be recognised as assets, and depreciation of the assets will be delayed to be matched against the resulting revenues, the business may well report an accounting loss in the early years. For example, it may have to employ a management team of a certain minimum size, at a cost that will exceed the revenues earned from the small customer base in the early years. This will be considered worthwhile if the business is expected to earn, in due course, profits large enough to outweigh the early losses. Study of a single year, early in the maturity of such a business, may suggest that the business is earning excessive profits. But a single year cannot validly be considered in isolation in a case like this because, given the need to incur losses in the early years, higher accounting profits in the later years are necessary to the achievement of a discounted cash flow return equal to the cost of capital. Some would argue that additional intangible assets - for example, "start-up costs" - should be recognised in the early years so as to eliminate the losses in those years and also eliminate the spurious impression of excessive profits in later years. However, accounting authorities have consistently refused to give support to recognition of such an asset for purposes of reporting to shareholders. Furthermore, whatever view is taken of some start-up costs, no method exists, or can practically exist, for identifying all such costs, identifying the part of the costs of a management team that should be treated as an asset in the early years and dealing with the other matters required to bring the accounting rate of return into line with the internal rate of return.
5. Generalisation of the above analysis has led some to argue that accounting numbers should not be used for measurement of profitability, at least in certain cases such as monopoly investigations. This appears to be the view of Professor Franklin Fisher (see Appendix 13.1). However, in my opinion, and, I believe, in the opinions of the great majority of people with financial expertise, this is too extreme a view. Use of a measure of periodic performance is essential to everyday business life and such use is made on innumerable occasions. Accounting measures are used because, although imperfect, they are the best available. And if they are used with an awareness of their limitations and with adjustments to make them as useful as possible for the purpose in hand, they can serve us well. Their limitations are most serious for businesses in a start-up phase, or recently emerged from a start-up phase, or businesses that are experiencing unusual and perhaps erratic growth or have been through major reconstructions. For mature and reasonably stable businesses - and the clearing banks appear to fall into this category - fewer difficulties are present. Accounting measures should still not be regarded as giving definitive computations of profitability - computations that can be shown to be absolutely correct. Their indications should perhaps be regarded in probabilistic terms. If adjustments are made for the imperfections of accounting measures as far as possible and, in a stable situation, profits appear excessive by a wide margin, a very high probability may reasonably be said to exist that profits are indeed excessive. I note that the main thrust of bank views, as given in chapters 7 and 10 to 13 of the Commission's report, is to call for modifications within the conventional accounting framework rather than to reject that framework.
6. One further comment on the relationship between the accounting rate of return and the internal rate of return may be of interest. If the cost of capital (the minimum satisfactory return) is constant over time and the accounting profit is always equal to the cost of capital times the accounting value of equity capital, the internal rate of return is equal to the cost of capital over the life of the business. This result holds good even if some assets are not recognised as they should be. Revenues will have a different pattern over time from what they would have had if the assets were recognised but this will not prevent shareholders receiving a satisfactory return. The result does not hold good strictly when the cost of capital varies over time because overs and unders associated with asset recognition do not then necessarily cancel out in discounted cash flow terms. Nevertheless, the result may be true as an approximation and may provide some assurance, particularly when one can be confident that profits have exceeded the cost of capital in the past.
Adjustments to Conventional Accounting Measures
7. I turn next to consider the adjustments that should be made to conventional accounting calculations in order that they should be as useful as possible for the investigation of the Commission. I take the base to be information prepared in accordance with standards for reporting to shareholders. The main areas to consider are whether all assets are recognised and whether they are measured appropriately, together with related effects on profit. As noted above, a deliberately conservative view is taken of asset measurement in reporting to shareholders. For assessment of the possible existence of excessive profits, in my opinion, it is appropriate to take a balanced view, avoiding any bias towards undervaluation.
Asset Valuation
8. I first consider the measurement of assets. In many situations, conventional accounting measures assets at their original cost or cost less depreciation. However, this is not the best basis for assessing performance where reliability of reporting by one group to another is not a key concern. I agree with the Commission when it says (paragraph 2.272) that deprival value is the appropriate basis for the case in hand. I also agree that in most practical circumstances, the replacement cost option under deprival value is the appropriate one. In essence, under deprival value analysis, if an asset would be worth replacing (were replacement to be necessary), replacement cost is the relevant measure. Realisable value and value in use, the two other possible valuations under deprival value concepts, would be relevant only where an asset should be sold or used up and not replaced. Replacement cost puts an upper bound to the valuation of an asset, assuming replacement to be possible, because one would not pay more for an asset or group of assets than the price at which it can be replaced. It is worth noting that, given very active markets for all relevant resources, including assets of various ages, replacement cost, realisable value and value in use should not differ very much.
9. I note that reference was made (see paragraph 2.278) to the fair value concept used by US authorities for valuation in certain circumstances. As the text of the Commission report notes, fair value is defined as the price that would be agreed for the transfer of an asset between a well informed and willing buyer and a well informed and willing seller. That definition, in effect, calls up some of the conditions for an active market and does not lead to a conclusion significantly different from the deprival value concept.
10. It may be worth noting that traditional accounting methods for calculating depreciation can distort accounting rate of return as an indicator of discounted cash flow rate of return. If usage of an asset and operating costs, including maintenance, are reasonably stable, and straight line depreciation is employed, the rate of return on the asset can appear to be increasing as the asset ages. This is because an approximately constant profit is compared with a decreasing level of capital employed. The accounting literature shows that, under certain assumptions, so-called "annuity depreciation", perhaps with an adjustment for changes in the price of the asset, can give greater consistency between the accounting rate of return and the internal rate of return; and original cost, depreciated in this way, can give a good estimate of the replacement cost of an asset of the relevant age. The analysis becomes complicated when patterns of usage and operating cost vary over the life of an asset and, in any case, the significance of the effect depends on the asset structure of the whole business. If the business has a well-balanced mix of assets of different ages and this mix does not change much from year to year, the distortion has little importance overall because under-assessment of the rate of return on young assets is offset by over-assessment of the rate of return on old assets.
11. In my opinion, the Commission has handled the measurement issues well. It has based its measurement of intangible assets on replacement cost. It has recognised the relevance of current values for land and buildings even though it concluded that, in most cases, adjustments to the base numbers were not needed because they could be assumed to approximate the relevant measure already. It did not discuss the depreciation issue explained above (paragraph 10) but it seems clear that any adjustment for this matter would have been immaterial in effect in the case on hand.
Intangible Assets
12. Turning to the question of whether all assets are properly recognised, I next discuss the treatment of intangible assets. These present particular difficulty. The assets are not easily identifiable in concept. They can not be observed directly and their original cost and hence their replacement cost can not be easily identified. Moreover, it is very difficult to assess their lives. One may feel, for example, that a promotional campaign qualifies, in principle, as an asset because it produces future benefits. However, it is not clear which costs should be included as contributing to the future benefits. (For example, should employment costs of permanent staff who worked on the project be counted?) And it is not clear how long the benefits last. The lives are not fundamentally unknowable - one could conceive of controlled experiments to assess the duration of various effects - but tests that would give really good evidence are impracticable and little information is systematically available even in areas where some could be gathered. These difficulties explain why accounting authorities are reluctant to sanction including intangible assets in routine company reporting. However, intangible assets are relevant for assessing profitability. I believe that the Commission has adopted the best approach available in that it has examined expenditures that could be expected to produce future benefits and applied its judgement in estimating the kind of life that could be expected for an asset of the type concerned.
13. The Commission's consideration of intangible assets focused on three main categories, staff acquisition, customer acquisition and information technology. In all cases, the Commission appears to have given full consideration to the claims of the various banks regarding what assets should be counted. It has exercised its judgement regarding what should be counted as an asset and also regarding the appropriate lives of the items concerned for the purposes of the investigation. In none of these judgements does the Commission seem to me to have behaved unreasonably.
Staff Acquisition
14. In the case of staff acquisition, among the major costs that the Commission decided should not be viewed as creating assets were costs of maternity leave and career breaks, the acquisition of temporary staff and relocation costs for lower paid employees. I take the same view of these items. The Commission has adopted a five year life for the depreciation of staff acquisition costs and I see this as being in a reasonable range, perhaps rather on the generous (high) side.
Customer Acquisition
15. Customer acquisition costs involve particularly difficult issues. The Commission begins its discussion of its conclusions in this area (paragraph 2.296) with an expression of discomfort about recognising such costs as creating an asset. Its reservations are understandable because of the impracticability of establishing the effects of individual expenditures with a high level of confidence. However, I believe that some asset should be recognised for the purposes of the Commission's investigation because promotional expenditures can be expected to bring future benefits and managers undertake them in expectation of a return.
16. The Commission adopted a seven year life for purposes of depreciation of customer acquisition costs. It said (paragraph 2.307) that this accorded with its understanding of the "average life" of an SME customer. As the Commission recognised, the average length of benefit from promotional expenditure may be less than the average customer life, and I regard the use of a seven year life as rather generous (high) although this has to be considered alongside other comments that I make below.
17. The Commission evidently considered carefully the submissions by the banks regarding the costs that should be treated as creating assets. The main issues that arise from this consideration are broad issues of principle. The Commission decided to reduce the amount attributed to the asset "customer acquisition" for three reasons: it saw some of the expenditure as neutralising the expenditure of other banks, some of the remainder as wasted and some as retaining existing customers rather than attracting new ones. It considered the proportions of expenditure that should be disallowed on the grounds that it fell into these categories and concluded inter alia that 80 percent of advertising and marketing expenditure should be disallowed (paragraph 2.313a). I am doubtful about the rationale for the disallowance. With regard to the expenditure to neutralise the expenditure of other banks, I can see that this may be regarded as having no value for the economy as a whole. However, it is, so to speak, a cost of having competitive markets (imperfect though the competition may be). An individual bank is (presumably) better off spending the money to neutralise the expenditure of other banks than not spending it and losing business. Spending this money is a valid business practice in pursuit of shareholders' interests and can be expected to yield a return for the individual bank concerned. The same applies to expenditure to retain customers.
18. The treatment of wasted expenditure is more difficult. I am reminded of the debate that has raged in accounting over the treatment of the costs incurred by oil companies in searching for oil. Some have advocated a "successful efforts" approach - that the costs of abortive exploration such as "dry holes" should be written off against profits straight away and not counted with the assets. Others have advocated a "full costs" approach - that the cost of the exploration programme should all be capitalised (treated as an asset), provided that the value of the oil found exceeds the total cost. The advocates of the successful efforts approach argue that only expenditure that is directly related to a source of future benefit should be counted as an asset. The advocates of the full cost approach argue that, because it is impossible to tell with certainty in advance which efforts will be successful, the cost of the entire search programme is necessary to obtain the assets - the productive oil wells - and all should therefore be regarded as creating an asset. For some time, the authoritative accounting literature worldwide allowed the two approaches as alternatives because of the strength of opinion on both sides. Recently, there has been a movement towards the successful efforts approach. The same arguments can be made about successful and unsuccessful promotional expenditures. One can not tell in advance which expenditures are going to be unsuccessful and so the total expenditure is necessary to achieve the overall results. Accordingly, I have some sympathy for the full cost approach in the context of the Commission's investigation and hesitate to agree with the disallowance of advertising expenditures assumed to be abortive.
19. Two other categories of expenditure are worthy of comment in relation to customer acquisition, sales incentives such as the cost of free banking, and the costs of credit assessments to borrowers. The Commission rejected both for inclusion in the asset measurement and I agree. I see the cost of free banking as a particular pricing mechanism chosen by the banks to optimise their overall net revenues - not as an investment. And the costs of credit assessment are related to a customer tariff that involves an initial fee as well as periodic charges. The costs of credit assessment could be regarded as creating an asset but they are effectively recovered in the initial charge to customers. (An analogy is in the position of telephone companies where the costs of connecting a new customer could be treated as an asset except to the extent that they are immediately recovered in a connection charge.)
20. Overall, on the question of capitalising customer acquisition costs, I believe that the Commission has adopted a reasonable approach. The issues are complex and different accountants will take different views about the details. I would not have disallowed such a high proportion of some of the expenditures but would have adopted a shorter life than the seven years adopted by the Commission (and the Commission gave the disallowance of expenditures as one reason for allowing a relatively long life). I do not think that the overall effect of my leanings would produce a result that was significantly different for the overall conclusion.
Information Technology
21. The third main category of expenditure on intangible assets is information technology expenditure. Although it raises some difficult issues, this area is less controversial than the previous two. A major difference between the positions of the Commission and the banks was on the lives to be assumed for depreciation of the asset. The banks proposed ten years - or even infinity, in one case - whereas the Commission preferred four or five years. I agree with the Commission's position which is more in line with general practice.
Other Specific Intangibles
22. The Commission's report notes that one of the banks proposed, late in the enquiry, the recognition of large additional intangible assets (paragraph 2.327). The assets fell under the headings of brand, reputation, licence, product and infrastructure. The claim under brand was the largest and was supported by an ingenious analysis by M&C Saatchi Ltd. This analysis measured the brand strength of two other organisations, assessed the cost of establishing those brands and applied the derived rate of cost to the brand strength of the bank concerned. I agree with the Commission when it concludes that the approach is unconvincing (paragraph 2.330). The procedure is not shown with sufficient force to relate to the costs of the bank concerned and one would suppose, as the Commission did, that, in general, such costs as were capable of being linked to the SME business would have been dealt with under the heading of customer acquisition costs. Similarly, I agree with the Commission's rejection of the other costs raised in paragraph 2.327.
Goodwill
23. The Commission's report notes that consideration was given to the inclusion of goodwill as an intangible asset (paragraph 2.342). Goodwill is recognised in accounting when a business is purchased (or brought in through merger). The amount attributed to it is the difference between the value of the whole business and the net amount attributed to recognised assets less liabilities. In other words, it is a balancing figure. The amount that is attributed to goodwill is lower, the higher the amounts that can be attributed to individual intangible assets - indeed much of the amount of goodwill is normally assumed to arise because of the non-recognition of intangible assets, assets which the Commission has endeavoured to recognise as discussed above. The existence of goodwill above the amount attributed to intangible assets might be taken as an indication of the need to search for other such assets. However, the value of a business, and hence goodwill, is determined by an assessment of the present value of its potential future cash flows and it may reflect the expectation of being able to earn excessive profits in future, cost economies and so on. Furthermore, one cannot know whether a business is correctly valued or whether the value agreed reflects an over-optimistic view of future prospects. For these reasons, I think that the Commission has been reasonable in insisting that the recognition of intangibles should be restricted to assets that are specifically identified and associated with costs incurred.
Depreciation of Intangibles
24. I will add at this point a comment on depreciation as it applies specifically to intangible assets. Normally, accounting involves the identification of the cost of assets, adds that cost to the asset account and deducts depreciation which then becomes a charge against profits. Sometimes, the expenditure on asset acquisition during a year can be assumed to equal depreciation and that expenditure is written off against profits in lieu of a depreciation charge. Suppose, for example, that a business has ten identical machines, one acquired each year over the past ten years, that the machines have a life of ten years, they have no residual value and there have been no changes in the prices of machines. The business intends to continue to buy one new asset each year and will thereby maintain its asset structure. The cost of the new machine would equal the required depreciation charge and the cost could be charged against profits in lieu of depreciation. This equivalence applies only where the asset can be assumed to be approximately constant from year to year. (It would not have applied when the business was building up from not having any assets to the steady-state position of having ten.) The same principles apply to intangible assets as to tangible ones. The Commission has assumed that the asset values for the intangibles it has recognised are approximately the same from year to year. Accordingly, it has estimated the asset amounts, taking account of depreciation and the need to measure the asset at replacement cost, and added the amount to capital employed. But it has assumed that current annual expenditure is equal to required depreciation so that no adjustment to profits is required. The Commission's approach seems entirely reasonable. It could be argued that replacement cost was presumably increasing from year to year and that the increase in asset value associated with this should be added to profit. The Commission has not done this. The issue is complex and controversial and would not make a significant difference in the case under consideration.
Stock Market Value/Book Value Ratio
25. It is convenient for me to deal, at this stage, with a challenge to the methodology used by the Commission, put forward in a paper by Charles River Associates (Appendix 13.2) (CRA). CRA note that the measurement of equity capital employed is critical to the assessment of super-normal profits and that several banks have stock-market values equal to two or three times their accounting book values; and they suggest that "the 'correct' capital base is 'somewhere' between book value and market value." They provide a table showing the ratio of market value to book value by sectors for most of the sectors represented in the FTSE 350. Several of the sectors have ratios in excess of four, some of them to a remarkable extent. CRA recognise that stock market value will exceed book value where monopoly profits are expected to be earned in the future: market value will include the discounted value of those profits. However, they argue that large differences also exist for companies that evidently do not have monopoly power. One can see that some of the cases are clearly explained (to a considerable extent) by non-recognition of intangible assets (software - non-recognition of development costs, pharmaceuticals - non-recognition of research and development). And one has to consider the suggestion that the very existence of a gap between stock market value and book value is evidence of the existence of unrecognised intangible assets, in cases where there is no monopoly power. The market values of the four major banks under consideration may include some capitalisation of excessive profits. But the difference between stock market value and book value for banks that do not enjoy a monopoly position may represent unrecognised intangible assets and the four major banks may have amounts of unrecognised assets to a similar extent (perhaps in proportion to other assets).
26. CRA suggest that their observations about the relationship between stock market values and book values indicate the appropriateness of applying to the book values of the major banks an uplift derived from the market value/book value ratios for a comparable set of companies without monopoly power. The comparable companies suggested by CRA are other UK and US banks which are said not to be suspected of abusing a monopoly position. CRA put forward a ratio of 2.75 times book value for consideration by the Commission as an appropriate uplift. However, CRA do not put forward detailed arguments to show that these companies are truly comparable as regards operating conditions and in other ways.
27. The Commission evidently found the suggestion of CRA unpersuasive and so do I. A major weakness of the CRA position is the lack of a theoretical framework that fits all the evidence and gives an understanding of the forces at work. Let us consider the possible explanations for the phenomenon of stock market values greatly in excess of book values in greater detail. One notices that current high stock market values are associated with unusually high price earnings ratios. Authoritative commentators in the Financial Times have been noting this phenomenon over the past year and suggesting that stock market values may be "too high". For example, The Lex Column on 2 January 2002 pointed out that the price-earnings ratio for the FTSE World Index on the basis of projections of earnings that may be optimistic is 20.6 "in the very upper range of its recent history". The same source points out that if growth for UK equities equals the long term real growth rate of 2.5%, with a dividend yield of about 2.5%, investors will earn 5% in real terms - i.e. about 8% in money terms. If investors in the major banks are to earn 11percent in money terms (the cost of capital calculated by the Commission for banks), with a dividend yield of 2% to 4%, growth averaging 8 percent must be achieved. One wonders whether this is realistic even though it is implied by a market consensus. In strictly competitive markets, one would expect the existence of stock market values in excess of required capital to attract entry by new firms which would push up asset prices and bring down prices of the goods and services produced until some closer approximation between stock market values and book values were reached. This is a long term tendency. With this background, one way of viewing the gap between stock market values and book values is that market values have been driven up in recent years, perhaps as investors have come to accept a lower average risk premium for holding equities; and the forces that bring about adjustments to restore equilibrium among profits, asset prices and stock market values have not yet had time to work. It is perhaps a feature of modern business activity that economies of scale and lack of divisibility of assets are becoming ever more important and the tendency for stock market values to be close to book values is accordingly becoming weaker - as conditions change, markets do not quickly adjust to whatever approximation to a new equilibrium is attainable.
28. It remains a possibility that unrecorded assets are part of the explanation of the gap between stock market values and book values. However, the Commission has conducted a careful and reasonable investigation to assess the existence and amount of such assets. A weakness of the argument that additional unrecorded assets exist is that the banks have apparently not recorded those assets as part of a system for managing, maintaining them and enhancing them. Nor is there any evidence of profits that have been unsatisfactorily low in the past as a result of deducting from profits costs that should have been capitalised. I do not think it would be reasonable to suppose that all of the companies contributing to the high market/book ratios in the FTSE 350 have large unrecorded intangible assets. Accordingly, I prefer an explanation of the position in terms of a disequilibrium and I believe that the Commission exercised its judgement reasonably in refusing any further adjustment to the book values of assets.
Equity Capital Employed
29. The next major item that the Commission had to deal with in its accounting analysis related to equity capital employed. In most industries, the capital employed would be built up from a consideration of the assets required to carry on the business and the liabilities which contribute to their financing. In the case of the clearing banks, however, different considerations apply because of regulatory requirements. These requirements specify a level of equity capital that must be provided having regard to the risks of the particular position of the bank concerned. Legal responsibility for bank regulation rests domestically with the Financial Services Authority (FSA), but a large measure of international co-operation now takes place on banking regulation because of the global implications of failure of a major bank. International co-operation on regulation takes place under the auspices of the Basel Committee on Banking Supervision. In attributing equity capital employed to the SME business of the banks, the Commission first calculated the amount required under current FSA/Basel requirements.
30. In January 2001, the Basel Committee released the details of proposals for a new approach to the determination of required levels of capital. This would take a more sophisticated approach to the assessment of risks and the relationship between risks and required capital. However, more recently, the Basel Committee announced that the new approach could not be expected to be finalised until sometime in 2002 and would not be implemented until 2005. Consideration was still being given to various modifications and, of particular relevance to the Commission's report, the Committee believed that further efforts were needed to ensure appropriate treatment of credit exposures related to SMEs. Three of the major banks covered by the Commission's enquiry put forward estimates of the capital that would be required for the SME business under the new Basel recommendations. These varied widely. The variations and the fact that the Committee have not yet agreed on the way forward show that considerable uncertainty remains about the levels of capital that will be required. The Commission could only apply its judgement. It did so and added 20percent to its first calculation of capital to allow for possible increases in requirements under the new recommendations. It clearly took great care to consider all relevant factors and study the views of the banks on the matter. Its final decision seems entirely reasonable.
back to top
31. The Commission also allowed for a 13 percent increase in capital for the banks' intangible assets which it believes should be recognised. The basic profitability of the banks' SME business was calculated with a component for bank lending, obtained from the volume of business multiplied by the margin earned on that business. Since some of the funds for loans are provided by equity capital, the Commission has added back an appropriate proportion of the cost of other finance for loans assumed in the margin calculation. Some of the equity capital might have been assumed to finance tangible assets and the intangible assets recognised for the purposes of the investigation. However, the financing of these items can be assumed to be shared among several sources of finance and, on that basis, the share of equity is immaterial. All of this seems to me to have been appropriately dealt with by the Commission.
Cost Variability
32. Some accounting issues arose because of suggestions that the years that were the focus of the analysis by the Commission were unrepresentative of the longer term. Three are perhaps worth specific comment, pension costs, bad debts and restructuring costs.
33. First, I comment on the treatment of pension costs. The banks operate pension funds into which they pay amounts to settle their obligations on the basis of actuarial advice. At the time of the Commission's investigation, the funds' investment performance had been unusually good and so the banks were not required to make the normal contributions into the funds; and, consequently, employment expenses were shown at lower than normal amounts in the profit computations. The Commission substituted higher expenses incorporating normal pension costs. This seems to me to be the appropriate treatment and not controversial.
34. Secondly, I comment on the incidence of bad debts. The pattern of bad debts experienced in banking appears to vary significantly in relation to macroeconomic conditions for the period concerned. In the three years that were the focus of the Commission's investigation, 1998, 1999 and 2000, bad debts were relatively low, a good deal below the average for the period 1989 to 2000, considered as a reference period. The banks argued that profits for the years under consideration should be calculated by including a charge for bad debts that is representative of the whole economic cycle. The Commission considered this claim with evident care. It reviewed bad debt experience for each year from 1989 to 2000 for the major banks involved and noted the differences between "recession years" and other years. It accepted the argument of the banks in principle but thought that it should not simply take the long period average for the bad debt rate because the banks would have learned from the experience of the early 1990s when bad debts were very high and improved their lending and credit control techniques. On this basis, the Commission allowed a bad debt rate of 1.2percent, compared with an actual rate of about 0.5percent for the largest banks in the most recent full year. The decision about what bad debt rate to allow is a matter of judgement of what is likely to happen in the future, given the basis of past experience. There is no accepted accounting technique that can be referred to for guidance nor is experience in other industries likely to be helpful. It seems to me that the Commission's approach is entirely reasonable and has been undertaken with evident care.
35. Thirdly, one of the banks had restructuring costs for one of the years under consideration. These were shown as an exceptional item in the bank's profit computation. The Commission ignored them for the purposes of its analysis. These costs could be deemed relevant to the overall economic assessment of performance - they would feature in a discounted cash flow analysis to determine the rate of return - even though, for purposes of an accounting analysis, they should perhaps be spread over a number of years. A different treatment would not have any significance for the overall conclusions.
Line of Business Allocation
36. Accounting issues of a different, but nevertheless potentially important, nature arise in attempts to identify the profitability of a particular segment of business activity in a company that has other activities; in the case under consideration the need is to identify the results of services to SMEs separately from other banking activities. To identify the results of SME services, it is necessary to allocate revenues and expenses to the different activities. In many cases, this is relatively easy but difficulties arise where there are cases of jointness. For example, some of the costs of maintaining a network of branches may arise at a fixed level whether a bank has any one of personal customers, SME customers and large corporate customers, or two of these, or all three. And strictly there is no demonstrably appropriate basis for allocating such costs among the activities that benefit. Different rule-of-thumb methods can be identified but none can be justified strictly. A similar difficulty might be found in relation to marketing costs where a personal customer was also in control of an SME customer and the two accounts are likely to be held at the same bank for reasons of convenience.
37. There is an extensive accounting literature on allocation problems of this kind. Some are of minor importance, relating more to convenience of allocation in a routine system than to dealing with costs that are literally joint. However, it appears from the text of the Commission's report that allocation problems have not been particularly controversial in the investigation under consideration. The Commission appears largely to have accepted the approach put forward by the banks. It may be the case that strictly joint costs are not very large in the case of banks and, even if they are large, it is probably reasonable that different lines of business should be expected to contribute to joint costs in proportion to some measure chosen to represent the use of the resources concerned. I do not think that the Commission's analysis is vulnerable to challenge on such grounds. A related point (because it was based on the same analysis of costs) concerns possible disallowance of costs by the Commission on grounds that they represented a level of inefficiency of operation. This was much more controversial. However, the Commission reasonably considered all the arguments put forward and eventually made such an adjustment only for one bank.
Conclusion
38. My analysis in this report has recognised that there are many difficulties in the application of accounting analysis to the determination of rate of profit. Independent experts would be likely to reach differing conclusions at least in points of detail. However, I believe that the framework adopted by the Commission is sound and would be accepted as appropriate by most independent experts. I would have judged some of the details of application of the framework differently from the Commission but that is in the nature of such analysis. The Commission has been guided in its judgement by findings in the course of meetings with representatives of the banks and sifting through a great amount of written evidence and representation. Had I undertaken such a process, my own conclusions might have differed from those given in this paper. Be that as it may, I think it is very unlikely that I would have reached conclusions that differed significantly from those of the Commission and I am clear that the conclusions of the Commission do not fall outside a range of reasonable conclusions.