AustLII Home | Databases | WorldLII | Search | Feedback

Journal of Law and Financial Management

USyd
You are here:  AustLII >> Databases >> Journal of Law and Financial Management >> 2002 >> [2002] JlLawFinMgmt 2

Database Search | Name Search | Recent Articles | Noteup | LawCite | Help

Goddard, Garry K; Walker, Greg --- "Competition Analysis of Bank Mergers in Australia" [2002] JlLawFinMgmt 2; (2002) 1(1) Journal of Law and Financial Management 6

Competition Analysis of Bank Mergers in Australia

Garry K Goddard [+] and Greg Walker[#]


Abstract

This paper reviews the competition analysis of bank mergers in Australia, a regulatory task that is undertaken by the Australian Competition and Consumer Commission (ACCC) under the Trade Practices Act 1974 (Cth), s 50. As a preliminary step, some background is provided on ACCC merger assessment procedure. This is followed by a discussion of the three most significant bank merger investigations undertaken by the ACCC, including the ACCC’s most recent and comprehensive investigation into the Commonwealth Bank of Australia’s acquisition of Colonial Limited. Five major themes running through these investigations are then identified and discussed. The five themes include: the use of market share thresholds and analysis of coordinated conduct by the ACCC; market definitions, especially the impact of electronic banking and convergence on definitions; contestability and entry barriers in retail banking; the importance of regional banks in promoting competition; and the nature of competition remedies accepted by the ACCC. Some brief observations on the implications of all of this for future bank mergers are then presented by way of conclusion.

Introduction

In Australia, few, if any, corporate mergers or acquisitions generate greater public interest and concern than those between major financial institutions, and especially those involving large retail banks. The extent of public sensitivity has been such that, since May 1990, mergers between the four largest banks have been explicitly prevented by public policy stances adopted by successive Federal Governments, first by the “six pillars” policy of the former Labor Government and, since April 1997, by the so-called “four pillars” policy of the Coalition. It is interesting, given the high profile of bank mergers as an area of public policy debate in this country, that there is not an extensive academic literature on the law and economics of assessing the competition (antitrust) effects of bank mergers in Australia.[1] In the United States, by comparison, there is a vast literature on the antitrust implications of bank mergers.[2] While the US literature can provide some useful analytical insights, the very different nature of the two banking systems and of the regulatory procedures adopted limits the practical value of much of this material in the Australian context.

The importance of this topic is highlighted by the debate on bank mergers policy, which, if anything, has intensified since 1997 when the Financial System Inquiry (FSI), or, as it became known, the Wallis Inquiry into the Australian financial system, reported to government on this issue. Wallis recommended that the six pillars policy should be abolished and bank merger proposals should be left to be assessed by the appropriate regulatory authorities.[3] For bank mergers the appropriate authorities are the Australian Prudential Regulation Authority (APRA) for prudential effects and the Australian Competition and Consumer Commission’s (ACCC) for competition effects. The FSI recommendations on mergers policy were amongst the very few not to be adopted by government and implemented in full over the past five years; the policy ban on mergers between the four major banks was retained by the Treasurer and remains in force to the present day.

In commenting on this policy impasse, Harper has suggested that: “The ACCC and APRA are more than adequate to the task of assessing these likely effects and imposing conditions, as necessary, on merger proposals.”[4] Other recent papers have expressed negative views on the adequacy and accuracy of ACCC determinations in previous bank merger cases. Rose has criticised the analysis and remedies adopted by the ACCC in the 1997 Westpac takeover of the Bank of Melbourne.[5] Valentine and Ford have taken a contrary view to the ACCC on the competitiveness of the Australian banking industry. They argue that the level of competition is greater than that accepted by the ACCC and that ultimately, removal of the four pillars policy will be a political rather than an economic decision.[6]

The ACCC (and its predecessor, the Trade Practices Commission) have assessed a number of bank mergers over the past decade. However, three of these merger cases stand out as being most significant in providing an insight into the major issues and problems in assessing bank mergers and pursuing competition remedies. These three cases are: Westpac Banking Corporation’s (WBC) 1995 acquisition of Challenge Bank in Western Australia, WBC’s 1997 acquisition of the Bank of Melbourne Limited (BML), and the Commonwealth Bank of Australia’s (CBA) 2000 acquisition of Colonial Limited (Colonial). In addition to these three cases, the ACCC also examined the 1995 Advance Bank takeover of the Bank of South Australia and the 1996 St George Bank takeover of the Advance Bank. However, neither of these two cases proved particularly problematic from a competition perspective, and so their illustrative value is more limited.

This paper discusses the ACCC approach to assessing the competitive impact of bank mergers and acquisitions in recent years, and the competition remedies accepted by the ACCC in the more problematic cases. The paper is focussed rather narrowly on the above topic and does not aim to address wider issues, such as the recent review of credit card schemes by the Reserve Bank of Australia or the current (Dawson Committee) review of Australia’s competition laws and their administration. The aims of the paper are essentially twofold. First, to provide a comprehensive summary of recent bank merger investigations in Australia in order to illustrate how the approach of the ACCC has evolved over time. Secondly, to comment on the appropriateness of the ACCC approach in these specific cases and the views of its critics. Both authors were involved with the ACCC assessment of the CBA merger with Colonial and hence the arguments presented in ACCC publications on that merger reflect to a considerable degree our private views on that merger. However, as noted in the disclaimer, we remain solely responsible for the views in this paper.[7]

The paper proceeds by providing an outline of the merger review process employed by the ACCC (section 2), an exposition and analysis of the WBC / Challenge and WBC / BML cases (section 3), and the recent CBA / Colonial case (section 4). Major themes and critical issues running through all these bank merger investigations, including: the use of market share thresholds and analysis of coordinated conduct; market definitions, especially the impact of electronic banking and convergence on definitions; contestability and entry barriers in retail banking; the importance of regional banks in promoting competition; and the nature of competition remedies accepted by the ACCC are then discussed (section 5) before conclusions are presented (section 6).

2. Merger Review Process

In Australia, an established protocol exists for coordinating those regulatory agencies that play a central role in assessing bank mergers. Specifically, the ACCC analyses the competition effects of the proposed merger. APRA examines the prudential aspects of the merger. Both agencies undertake concurrent assessments and then report their findings to the Federal Treasurer, who then makes a final “national interest” decision, encompassing both the competition and prudential issues but also addressing wider public policy concerns (these could include, for example, employment, regional development and banking sector efficiency issues).[8]

The competition assessment of bank mergers is therefore only one part of the overall process of regulatory approval for such transactions in Australia. Nonetheless, it is a crucial element given the concentrated nature of the banking industry and community concerns about the price, quality and availability of banking services, especially in rural and regional areas. The key piece of legislation that pertains to the competition assessment of mergers by the ACCC is the Trade Practices Act 1974 (Cth), s 50 (TPA). Section 50 prohibits mergers or acquisitions within Australia which would have the effect, or are likely to have the effect, of substantially lessening competition (SLC) in a substantial market for goods or services.[9]

At a broad level of conceptualisation, what the merger provisions of the TPA seek to do is to provide a legal framework within which an assessment of the social desirability of particular merger transactions can be made using economic concepts and tools. The economic analysis sets out to ascertain the degree of market power that a merged firm would or could enjoy post-merger and to assess whether the exercise of that market power is likely to lead to such economic detriment to consumers and other firms in the economy that the proposed transaction should not be allowed to proceed. In administering s 50 of the TPA the ACCC has adopted a five-stage process to assess market power, consisting of the following elements of analysis:

In considering market definition, the ACCC aims to define all the product markets likely to be impacted by a proposed transaction. Put simply, the aim of the exercise is to identify those products or groups of closely related products over which the merged entity may be able to exercise market power post-merger. For each product market, all sources and potential sources of close substitutes for the merged firm’s products are identified and included, and then an appropriate geographic aspect for that product market is ascertained. In theory, product substitution can be estimated by calculating (cross-price) demand and supply elasticities between various products that are potential substitutes. In practice, however, the data available to the ACCC may preclude such precise determination, and it may prove necessary to use a wider range of information (both qualitative and quantitative) about consumer and producer behaviour in order to form a view about the nature of the product and geographic markets that apply in a particular industry.[11] In bank merger cases this can be a complex task since banks and other financial institutions usually produce a wide range of products that are offered in multiple locations - implying a number of product markets with varying geographic dimensions. As befits the complexity involved in this crucial first step in the assessment process, it provokes conflicting views and the ACCC’s views on bank market definition are not universally accepted. This is discussed further below.

Once meaningful definitions of markets have been arrived at, the ACCC calculates the post-merger market share of the merged entity in each of those markets and compares these to its concentration thresholds. The thresholds are used as a guide to help identify those markets in which a substantial lessening of competition is considered most likely to occur and where further analysis is required. Specifically, where the merged entity has a market share of 40 percent or more (post-merger), the ACCC considers that the potential for unilateral (monopoly) market power exists. Where the merged entity has a market share of 15 percent or more and the four largest firms have a combined market share (CR-4) of 75 percent or more (post-merger), then the ACCC considers the potential for coordinated or interdependent (oligopoly) market power exists. Criticisms of this step in the ACCC analysis focus on issues such as whether the concentration ratio (CR) measure of market share is the best measure (e.g. the choice of CR-4 as opposed to say a CR-3 ratio for the threshold or the use of Herfindahl Hirshman Index (HHI) measures instead) and the theoretical and empirical basis for identifying potential for coordinated market power where the CR-4 thresholds are exceeded.

Generally, if the combined market shares of the two parties are below those threshold levels then a substantial lessening of competition is considered unlikely and the ACCC would not examine that market further. If the merger thresholds are exceeded, the ACCC considers two supply-side responses that could mitigate the anti-competitive effects of any market power arising from the accretion of market share post-merger. The first of these is the potential for imports to successfully penetrate a market and thus provide an effective antidote to the exercise of market power.[12] The second is the potential for entry of new players into a market, and for them to successfully compete against the incumbent players, thereby imposing pricing discipline. If either or both imports and low entry barriers imply an effective constraint on the use of market power post-merger, then a substantial lessening of competition would generally be viewed as unlikely. For bank mergers, the issue of new entrants / contestable markets has generated considerable debate. The debate is closely linked to the dynamics of the technology and innovation in the banking industry and consideration of this issue is clearly evident in the ACCC’s analysis of successive bank merger cases. We discuss the ACCC view on entry / contestability in banking and those of its critics below.

If, on the other hand, neither imports nor new entry appear likely to restrain the market power of the merged entity post-merger, then the ACCC will consider a range of other factors which may have a bearing on the anti-competitive impact of a merger in each market. Chief amongst these other factors are: the degree of countervailing power in the market; the dynamic characteristics of the market, including growth, innovation and product differentiation; the likelihood that the acquisition would result in the removal from the market of a vigorous and effective competitor; and the nature and extent of vertical integration in the market. This is a non-exhaustive list and does not preclude the ACCC from considering any other factors, including the impact of “efficiencies”.

Given the importance of efficiency arguments as a rationale for bank mergers, it should be noted at this point that the ACCC’s analysis under s 50 is confined to a narrow interpretation of efficiencies: that is, to those efficiencies that cannot be achieved through other means and which contribute to increased competition in the markets impacted by the merger, say through increased output, product innovation or reduced prices.[13] A broader interpretation of efficiency claims, based on offsetting public benefits[14] can apply under the “authorisation” provisions of the TPA. However, the ACCC has so far never considered a bank merger via the authorisation route. It is possible that such an approach may be attempted in the future if, following the abolition of the four pillars policy, mergers involving two or more of the four major banks were to eventuate. Supporters of bank mergers typically emphasise the efficiency benefits such mergers should generate.

3. ACCC Assessment of Westpac’s Acquisitions of Challenge and Bank of Melbourne

This section looks at two bank merger assessments during the 1990s, each of which were significant landmarks in terms of the evolution of competition analysis for bank mergers in Australia.

The first is a September 1995 investigation by the former Trade Practices Commission (TPC) - a forerunner of the ACCC - of the acquisition of the Western Australian regional bank, Challenge Bank, by Westpac (WBC). It is important for several reasons.[15] First is the TPC’s use of a “cluster approach” to product market definition, where the relevant market consists of a “cluster” of related products that tend to be consumed and / or sold as a bundle. Second, the adoption of a state-based geographic definition. Third, the TPC’s identification of significant barriers to entry in retail banking. Finally, there is the TPC’s recognition of the important role of regional banks in providing competition in Australian banking markets. It should be noted that the focus of this merger assessment was on retail banking due to the market alignment of Challenge Bank.

The “cluster approach” to market definition identified retail banking services as a single product market comprising a bundle of retail products that encompassed deposits, loans, and payments services. Whilst acknowledging “partial and at times vigorous competition for particular banking products from non-bank suppliers”, the TPC took the view that “there continues to be a competitive advantage in supplying a full range of banking services”.[16] The retail banking market was considered to have a state-wide geographic dimension. In taking this view, the TPC rejected arguments that the relevant geographic market was national in scope, and that the retail banking cluster was part of a broader financial services market that included life insurance, superannuation and other non-bank products. [17]

Significant barriers to entry for retail banking were identified by the TPC.[18] Entry barriers arose from the need for extensive retail branch and agency networks, electronic banking networks, customer inertia (a reluctance amongst customers to switch suppliers), scale and scope economies, the importance of brand awareness and the advertising expenditures required to maintain brand, small market size (of the geographically isolated Western Australian market) and significant sunk costs.

Although the TPC considered the Western Australian market for retail banking services highly concentrated, with the threshold screens breached for the deposit component of the market, it decided not to oppose the WBC / Challenge merger. The presence of another strong regional bank in Western Australia (BankWest) contributed significantly to the Commission’s decision not to oppose the WBC / Challenge merger.[19]

The WBC / Challenge case serves to illustrate the Commission’s assessment process. The foundation stone of this process was the definition of the relevant product and geographic markets. Subsequent decisions concerning the competition impact of a proposed merger and consideration of any competition remedies flow from market definition. The identification of possible competition problems in this case were triggered by high levels of concentration in the retail banking market in WA, and for deposits in particular. Imports from out-of-state banks were considered negligible and significant barriers to entry for retail banking were identified, eliminating these factors as potential sources of post-merger competition. However, in the final stage of its assessment process, the Commission identified the presence of another regional bank in Western Australia as a significant “other factor” mitigating against a substantial lessening of competition (SLC) in an otherwise problematic bank market.

The TPC concluded that regional banks had an important role to play in stimulating competition in all Australian banking markets. They were considered to be efficient, innovative, geographically focused, close to their customers and sufficiently differentiated from the “look-a-like” majors to provide an incentive for the major banks to remain competitive.[20] In the wake of the WBC / Challenge decision, the comments on the competitive significance of regional banks became amplified by some commentators in to what became known as the “Fels Policy” of retaining at least one regional bank in each state market. The clearest annunciation of this view is can probably be found in the following passage taken from a speech by the ACCC Chairman (Professor Allan Fels) shortly after the WBC / Challenge decision:

Any MTB [Major Trading Bank] acquisitions of regional banks would be scrutinised very carefully. In any State with only one major regional bank the Commission would be especially concerned. The concern is that such an acquisition at this time would be likely to substantially lessen competition. The Commission would need persuasive evidence to be convinced otherwise. Even in States with more than one substantial regional bank, the Commission would still examine any MTB acquisition of a regional very closely. [21]

This position on the importance of regional banks was reiterated by the ACCC in its 1996 decision not to oppose the St George / Advance merger when it stated: “This merger does not affect the ACCC’s long held view on the importance of having a strong regional bank in each State”. [22]

The second landmark bank merger decision during the 1990s was the ACCC’s July 1997 decision regarding the WBC acquisition of the Bank of Melbourne (BMC). This assessment marked an important shift in the ACCC’s approach to assessing banking markets.[23] In analysing the competitive impact of the WBC / BML merger the ACCC’s approach to market definition responded to significant changes in the way financial products were being purchased and sold. In particular, the rapid growth and success of non-bank mortgage originators had seen many bank customers unbundling their home mortgages from the remainder of their banking products. Thus, while noting that “product bundling is still occurring to a significant degree”, the ACCC recognised that “there is also a considerable number of customers who are willing to unbundle their banking requirements to obtain the best priced products where unbundling costs are significantly less than the potential pricing benefit.” [24]

In the WBC / BML case, the ACCC identified six product markets for “retail banking”.[25] Market share threshold testing lead the ACCC to conclude that three of these markets were unlikely to substantially lessen competition. Those three markets (with their associated geographic dimensions shown in brackets) were: home loans (becoming national); a cluster of small business banking products (considered most likely a local market, but state based figures were used for the analysis) and personal loans (state). For the other three markets the market share threshold tests for coordinated conduct were crossed. These three markets were: transaction accounts (state); deposit / term savings accounts (state) and credit cards (state).

However, following detailed competition analysis for each of these markets in accordance with the assessment procedure noted earlier, the ACCC concluded that a substantial lessening of competition was likely to occur only in the market for transaction accounts in the State of Victoria.[26] Concerns about a lack of competition in this product line were based on ACCC findings that: market and institutional conditions appeared conducive to coordinated conduct, especially between the four major banks; there were low penetration of imports (due to the importance of branches in delivering this product line and the slow take-up of remote delivery systems like on-line banking); high barriers to entry (linked in large part to branch and electronic payments networks); and considerable customer inertia (resulting from high search and switching costs associated with transaction account products). In reaching its conclusions on the (state-based) geographic scope and negative competition effects arising in the market for transaction accounts, the ACCC relied on a considerable amount of survey evidence that demonstrated the importance of bank branches to both retail customers and banking institutions. Much of this survey evidence had been provided by the recently completed Financial System Inquiry (FSI) into the Australian financial system.[27]

As noted in the introduction to this paper, mergers policy and regulation had been one of the central issues that the FSI had been asked to consider, and it made a number of significant recommendations for a more liberal approach to mergers policy and ownership restrictions.[28] While consideration of these policy issues is beyond the scope of this paper, it should be noted here that the FSI also presented more detailed “findings” on the ACCC’s approach to the competition assessment of bank mergers, in particular the market definitions adopted in the WBC / Challenge case. These were not formal recommendations, but the ACCC was obviously mindful of these comments / suggestions during the WBC / BML investigation and the approach adopted needs to be seen in this context.[29]

In response to the ACCC’s competition concerns regarding the transaction account market in Victoria, the parties to the WBC / BML merger offered court-enforceable undertakings under the TPA, s 87B.[30] The undertakings were designed to ease entry into this market and to maintain some of the non-price competition that had been exhibited by BML pre-merger. A crucial element of the WBC / BML undertakings were provisions requiring the merged entity to provide “access” to its electronic (ATM and EFTPOS) networks on reasonable commercial terms to small and new competitors. The ACCC considered these undertakings were sufficient to remedy the substantial lessening of competition identified and consequently did not oppose the merger.[31]

For the purposes of this paper, there are several key points that should be noted about the WBC / BML assessment. First, there is the shift in market definition away from the broader cluster approach towards the acceptance of individual product lines with varying geographic dimensions. In comparative terms, this puts the ACCC’s approach to bank mergers much closer to that of the Canadian Competition Bureau[32] and of the US Department of Justice in regard to defining markets, rather than the more stylised cluster approach used by the US Federal Reserve.[33] Second, there is the identification of a substantial lessening of competition in only one of the six product markets (ie the Victorian market for transaction accounts) which could not be overcome by “other factors” as in the WBC / Challenge case. Third, there is the ACCC’s acceptance of a combination of “quasi-structural” (access) undertakings and “behavioural” (price, product or service) undertakings in this particular case, as a competition remedy to the identified substantial lessening of competition. As regards the ACCC’s long held view on regional banks, it is perhaps pertinent to note that WBC had painted its acquisition as a “reverse takeover” whereby the merged entity would retain its BML brand name and a degree of autonomy within the Westpac Banking Group.

Despite shifts in the ACCC approach evident in the WBC / BML case, the ACCC decision has been strongly criticised by Rose.[34] He makes a number of important claims, especially in regard to coordinated conduct, including: transaction account products are not homogenous; prices are not transparent, especially where relationship discounts are offered; non-price competition and innovation are strong; posted prices were hard to observe in practice; and the cost structures of the four majors are not similar. We question several of these claims as well as a number of other issues raised by Rose, but this is left to section 5 below following a review of the CBA / Colonial merger.

4. ACCC Assessment of the Commonwealth/Colonial Merger

On 10 March 2000 the Commonwealth Bank of Australia (CBA) announced that it would buy Colonial. At the time of the announcement, the CBA was Australia’s second largest bank with consolidated assets of $161 billion. It was the largest in terms of housing loans and retail deposits and had Australia’s largest branch network, with 1,118 branches and 3,900 agencies. Colonial was a diversified financial services group with substantial insurance, retirement savings and funds management operations, as well as a considerable banking business. Colonial had consolidated on-balance assets of $54.4 billion. Colonial’s banking operations were the sixth largest in Australia and focused mainly in NSW, through its 1994 acquisition of the former State Bank of NSW, and in Tasmania, through its 1999 acquisition of the former Trust Bank.

The assessment of the proposed CBA / Colonial merger entailed the largest and most complex bank merger investigation ever undertaken by the ACCC. The assessment was completed in late May 2000, with the ACCC ultimately deciding not oppose the merger, but only after accepting significant competition related undertakings from the CBA.[35] The ACCC faced two major challenges in assessing the CBA / Colonial merger proposal. The first arose from claims of convergence in certain banking and investment products, which raised the possibility of broader market definitions extending beyond retail banking. The second related to the identification of appropriate competition remedies for those markets where the ACCC had identified competition problems.

Market Definitions

The CBA, in its submission to the ACCC, proposed a set of market definitions that challenged those established in the WBC / BML decision. A summary of the banking market definitions proposed by the CBA and those accepted by the ACCC is provided in Table 1. As can be seen, the major areas of difference related to retail / personal banking products and small business banking services.[36]

The ACCC rejected the CBA claim for a single national market in retail deposits and investments that included cash management trusts, managed funds and direct holdings of equities and other securities. It retained its previous WBC / BML definitions which excluded these additional investment products.[37] The ACCC also rejected the CBA claim for a single national consumer finance market, deciding instead to extend its previous definitions to identify four distinct consumer finance markets.

The ACCC retained its previous geographic definition for small-to-medium enterprise (SME) business banking services as well as identifying a separate product market for agribusiness lending. The historical strength of both banks in rural lending and their extensive contemporary rural and regional branch networks in the two states impacted most heavily by this merger, namely NSW and Tasmania, called for a special focus on agribusiness banking. The SME product definition retained the product cluster approach adopted in the WBC / BML case, with special consideration being given to credit and debit card merchant servicing for SMEs.[38]

For a group of (non-bank) financial services products that were not of critical importance to the CBA / Colonial case, the ACCC did not commit to firm product and geographic market definitions. Where these are shown in the tables they refer only to broad categories of products, each of which may contain a number of individual product markets with varying geographic dimensions.

A critical issue in the CBA / Colonial case was the extent to which market dynamics and in particular the impact of electronic banking justified a redrafting of market boundaries: both in terms of their product and geographic dimensions. Further discussion of this important issue is contained in section 5 below, however it is here noted that the whilst the ACCC rejected most of the claims for convergence in the product dimension of the markets examined (that is previously single product markets merging into a broader cluster market), it did accept that new technology and new non-branch service delivery mechanisms were contributing to geographic convergence in some of these markets (that is previously state-bounded markets merging to become national in scope).

Table 1: Product Market Definitions

CBA Definitions
ACCC Definitions
WBC/BML Case
CBA/Colonial Case
Personal Banking Products


Deposits and Investments (National)
Transaction accounts (State)
Deposit/Term Savings Products (State)
Transaction accounts (State)
Deposit/Term Savings Products (State)
Retail Investment Products (Not defined)
Consumer Finance (National)
Personal Loans (State)
Credit Cards (State)
Personal Loans (State - National)
Hybrid Personal Lending (National)
Credit Card Issuing (State - National)
Credit Card Merchant Servicing for smalll businesses (Local – State)
House Financing (National)
Home Lending (State- National)
Home Lending (National)
Business Banking


Small Business Lending (National)
Small Business Cluster (Local – State)
No Change


Agribusiness Lending (Local – State)
Corporate Finance (National)
Large Corporate Banking (National)
No Change
Non-Banking Products


Retirement Savings (National)

Retirement Savings (Not defined)
Wholesale Funds Management (National)

Wholesale Funds Management (Not defined)
General Insurance (National)
Life Insurance (National)

General Insurance (Not defined)
Life Insurance (Not defined)

Problematic Markets

The product markets that raised the most significant competition concerns for the ACCC in the CBA / Colonial case, were those for transaction accounts, deposit products and SME banking. The merger thresholds were exceeded for all three of these products in Tasmania and for transaction accounts and SME banking in NSW - hence a substantial lessening of competition was considered possible in a total of five markets.[39] Two other markets - credit card issuing and agribusiness banking in NSW - raised concerns but were ultimately not considered likely to result in a substantial lessening of competition as a consequence of this merger case. Despite both markets having levels of concentration that were likely to exceed the threshold screens for coordinated conduct (with probable CR-4 > 75%), the evidence revealed that Colonial had a relatively small share of the national credit card issuing market and was a declining force in the NSW agribusiness market, which was experiencing a growth in competition from new entrants.[40]

In NSW, Colonial had been a vigorous retail competitor. It has been an innovative leader in home equity lending (via its Viridian product). Colonial appeared to have successfully completed its takeover of the Trust Bank of Tasmania and seemingly was achieving efficiency gains for this bank. It was also mooted in the financial press as a possible buyer for the Bank of Scotland’s share of Bank West. These acquisitions signalled a corporate strategy involving Colonial’s possible transformation from regional to national banking status. Altogether therefore, at a broader level of consideration, Colonial could be seen as a vigorous and effective competitor in these markets and a potential future national bank.

New competitors may be able to enter the geographic markets affected adversely by Colonial’s removal but in rural areas this will not be easy. Entry could come about via electronic delivery based operations, assuming the provision of modern telecommunications facilities to rural and regional areas and assuming widespread customer acceptance of such services. New entrants have appeared in the agribusiness market.[41] Substitutes for retail banking services have been provided by credit unions and / or building societies in some rural areas where banks have closed branches. Bendigo Bank has been pursuing a community bank model in regional areas, including NSW, and, at the time of the merger investigation, had recently announced plans to enter the Tasmanian market through an alliance with established non-bank players in Tasmania. Bendigo has subsequently pursued this strategy and has established a retail banking presence in the Tasmanian market. The extent to which the expansion of these newer players into regional areas represents clear evidence of lower entry barriers and an effective pricing constraint on larger regional and national banks is, however, yet to be confirmed.[42]

After considering the above examples of entry, the role of imports and other merger factors, the ACCC remained concerned that the proposed CBA / Colonial merger would lead to a substantial lessening of competition for transaction accounts, deposit products and SME banking in Tasmania, and for transaction accounts and SME banking in regional (non-metropolitan) areas of NSW.[43] In this regard, it is worth noting that a number of the key ACCC findings in the CBA / Colonial investigation (eg local to regional markets, limited penetration of imports and high entry barriers linked to electronic networks and customer inertia) had also been identified in the earlier WBC / BML investigation, and have been echoed by competition regulators and public inquiries in other jurisdictions.[44]

Remedies

In examining the CBA / Colonial merger proposal, the ACCC considered whether the likely anti-competitive effect it had identified in retail banking markets in Tasmania and in regional NSW could be remedied by accepting appropriate s 87B undertakings offered by the CBA.[45] Such undertakings can provide a flexible and proportionate instrument for the parties to a transaction to modify a merger proposal in circumstances where the ACCC identifies competition problems, so those competition problems can be removed or at least minimised, rather than only facing the option of having the transaction blocked.

Traditional merger remedies are “structural” in nature, and usually involve the divestiture of parts of the acquired business to smaller competitors or new entrants in the market(s) affected - this is the ACCC’s “first best” approach to merger undertakings.[46] In this case, however, major complications existed with the pursuit of structural remedies with regard to Trust Bank, these included: difficulties in identifying a package of assets for sale due to the advanced stage of integration of Trust Bank into Colonial’s banking operations; concerns about the disruption to Trust Bank customers due to the uncertainty of another sale process; and limitations on the potential field of bidders due to competition and other regulatory constraints.[47] Given these constraints, and the identification of problems in NSW markets also, the ACCC considered whether the competition problems it had identified could be overcome by an alternative combination of “quasi-structural” and “behavioural” undertakings.

Quasi-structural undertakings were intended to improve the contestability of retail banking markets in Tasmania and regional NSW. More precisely, these included obligations for the merging parties to: provide access to their EFTPOS and ATM networks, and to their credit and debit card processing, and cheque, item and loan processing services, on reasonable commercial terms and conditions; and to facilitate the provision of information technology and telecommunications services from their suppliers and joint venture partners; for new or smaller financial institutions (ie other than the major banks) proposing to enter or expand their provision of financial services in regional NSW and /or in Tasmania.[48] Behavioural undertakings were designed to limit the scope for the exercise of any market power that may accrue to the merged CBA / Colonial entity post-merger in those same geographic areas. These required the merged entity to: set rates, fees and charges on all products, and to maintain the quality of service, to customers in regional NSW and Tasmania, at an equivalent or more favourable level than in metropolitan NSW; and to provide to customers in regional NSW and Tasmania the same product range as is available to customers in metropolitan NSW, subject to certain conditions relating to customer demand and delivery costs.[49]

This mix of remedies was ultimately considered by the ACCC to be adequate.[50] The logical link from market definitions to the identification of competition problems to the pursuit of competition remedies can again be seen in the CBA / Colonial case. As with WBC / BML, a mix of access and behavioural undertakings was accepted as a remedy to an identified substantial lessening of competition in the markets considered problematic by the ACCC. Section 5 below picks up on the broader theme of competition remedies in Australian bank mergers, and sounds a cautionary note on their precedential value.

5. General Themes and Critical Issues

Following on from the overview of the ACCC’s most significant bank merger investigations provided above, this section examines several key themes and critical issues derived from these past cases and likely to continue to apply to any future bank mergers.

Thresholds and Coordinated Conduct

The first theme relates to the efficacy of the ACCC approach to assessing merger proposals in oligopolistic industries and whether this approach is adequate to deal with bank mergers. Section 2 outlined the ACCC merger guidelines. An important issue is whether these guidelines are flexible enough to apply across all industries and in particular to the unique characteristics of the banking industry. Although the initial threshold screening step in the guidelines may appear to be prescriptive and inflexible, there is a degree of flexibility built into the subsequent steps that help avoid prescriptive outcomes. As well, in the bank merger cases considered in this paper, the procedure has been applied reasonably consistently. Further, evidence of shifts in some aspects of the ACCC approach suggests the procedures are capable of dealing with the complexities of a dynamic industry such as banking.

Legitimate criticism can be directed at the choice of threshold levels, although in many ways paying too much attention to this aspect of the ACCC review process overlooks their role as a screening device rather than as an analytical tool. For example, Rose argues that a three firm concentration ratio (CR-3) would be superior to the CR-4 ratio used by the ACCC and he cites empirical evidence from the industrial organisation literature to support this position.[51] There is no universally agreed theoretical basis for selecting a particular firm number for the market share threshold (CR) used to identify potential oligopoly market power, and with respect to the technical aspects of structural measures: “... the most commonly used measure of concentration in recent times is the Herfindahl-Hirschman Index.”[52] This index possesses superior technical qualities not found in the CR measure, including greater sensitivity to the change in concentration resulting from a proposed merger. It is “... because the Herfindahl index combines information about the number and the size distribution of firms, [that] it is usually regarded as the preferred measure of market concentration.”53 The HHI is employed widely by US competition authorities, which as a general rule tend to rely more heavily on concentration threshold numbers in forming a final view on mergers than does the ACCC.

Switching to an HHI would still not overcome the need to pick a threshold screen for a substantial lessening of competition test but this is really not the substantive issue in this matter. The critical issue is the inability of the industrial organisation literature to provide a clear and unambiguous indicator of the division between a competitive and a non-competitive industry structure for oligopolistic industries. The ACCC mergers procedure, as with any other assessment approach which relies on concentration measures, is thus at risk of inappropriately including or excluding merger proposals in a substantial lessening of competition test that could lead to an incorrect final determination. The ACCC provides little in the way of theoretical or empirical discussion on the threshold levels adopted in its merger guidelines, and in this respect is justifiably criticised by Rose. Rose does ground his criticism in both empirical and theoretical literature.

On the issue of coordinated conduct, we would argue for a more cautious interpretation of the literature than that adopted by Rose.[54] Rose draws heavily on the Stigler model of oligopolistic behaviour to critique the ACCC conclusions on coordinated conduct in the WBC / BML decision. This model predicts inherent instability in oligopolistic interaction. The forces that drive oligopolists to coordinated activity also provide a strong impetus for cheating on each other. There are many useful insights to be gleaned from Stigler’s approach, but the outcomes in practice are very dependent on a range of factors including, amongst others, the length and stability of the repeated interaction between large firms in the industry, similarities in cost structures, the relative value and frequency of sales, and the ease of detecting and punishing non-cooperative behaviour (cheating). While we would not argue that collusion of any type is necessarily present in the Australian banking industry, Rose has perhaps been too quick to dismiss the concerns that a competition regulator may have with a merger in a relatively stable retail industry dominated by a few large players of similar structure, where information about prices and sales volumes is readily available.[55] In this regard we note that concerns about high levels of concentration and the potential for coordinated market power were also identified by the RBA report on credit card interchange fees.[56]

Given the multiplicity of factors that could potentially impact on oligopolistic interaction, there are a number of theoretical models that predict different outcomes. Of interest to the Australian banking industry is the proposition that tacit collusion is more likely in an oligopolistic industry with firms of equal sizes and that: “... as the number of equal sized firms in an industry falls, tacit collusion becomes easier to sustain. When firms are of equal size, there are no asymmetries in the incentive to punish other firms for cheating.”[57] A critical factor in the outcome of merger activity is the reaction of firms outside the merger and the subsequent changes in industry structure. If firms outside the merger are large enough to react in an aggressive manner they may steal market share from the merged firm and prevent total industry output from being reduced. Indeed, if following the merger market share flows to a dominant firm outside the merger, and economies of scale are present, then average industry costs may be reduced and the merger could be welfare enhancing. However, as Gans concludes from his recent review of the behavioural literature on merger analysis: “there is no simple relationship between concentration, competition and welfare.”[58]

The Australian retail banking industry structure continues to be dominated by the four majors. They continue to be relatively equally sized. Past merger activity has impacted heavily on the regional banks. Many have disappeared and ironically this has been driven, at least in part, by the four pillars policy causing the majors to seek second tier merger targets. It has had the effect of minimising the potential for additional competitors to grow and disrupt this oligopolistic structure. Still, the evidence outlined in the previous two sections suggests it would be naive to simply apply any theoretical oligopoly model at the industry level without careful regard to the commercial realities of the industry.

Market Definitions and Convergence Issues

The cornerstone of merger assessment remains that of market definition. In the WBC / Challenge case the relevant market had been defined as a single market for retail banking, albeit with separate product categories identified for deposit accounts, loans, and payments. In the WBC / BML case six separate product markets for retail banking were defined. In the CBA / Colonial investigation further retail banking and business banking markets were identified and grafted onto the WBC / BML market definitions.

Rose was particularly critical of the ACCC’s market definitions for the WBC / BML case. A number of the criticisms he raises deserve consideration.

In particular, Rose argued against a single transaction accounts product market based on references to statements by the ACCC that the industry contains “... many hundreds of differentiated banking products and services”.[59] Rose takes this as sufficient evidence to dispute the ACCC definition of a separate transaction accounts market arguing that transaction account products are not homogenous and prices are not transparent, especially where relationship discounts are offered. However, this line of argument fails to appreciate the significant shift in the ACCC approach to unbundling retail product markets since WBC / Challenge, with closely substitutable products being grouped into the same product market category, and the evidence it drew upon to reach this conclusion in WBC / BML. It is also at odds, as revealed in section 4, with the broader retail market definitions argued by the CBA in its merger submission and those suggested more recently by Valentine and Ford.[60]

With regard to his point on posted prices and the lack of transparency in pricing, we would argue that prices for transaction products are transparent (even if somewhat complex) - being clearly indicated to retail customers in a variety of mediums including posters, mail outs and the internet. The issue he raises about relationship discounts is significant, as this has been an important feature of Australian banking in recent years. Whilst it could potentially generate price discrimination, such discounts are heavily focused on high net worth customers. A case could be made for a separate market definition for such customers, however, data limitations prevented exploration of this notion in both the WBC / BML case and the subsequent CBA / Colonial case.

There remains however a large and significant component of the retail transaction accounts market that relates to publicly posted prices. This market is also characterised by considerable customer inertia. Rose did not consider this point but it is a significant factor in drawing the market definition for this retail product market and in assessing its competitive structure. Another feature of transaction accounts pricing is the rise of transaction fees on such accounts in recent years. It is well documented that banks have suffered a decrease in interest margins and in response have increased fees on a wide range of services. Survey data on bank fees are now published by the RBA.[61] It is evident from these surveys that fees on household transaction accounts are growing faster than fees on other bank products.[62] This experience is consistent with the ACCC claim that retail transaction accounts are amongst the least competitive of bank product markets and hence have the potential for coordinated activity. If data were available to identify that segment of the retail market in receipt of relationship discounts, it is probable that a refined retail transactions market could be identified for non-discount customers that is even less competitive.

The growth in total fee income has not been sufficient to compensate retail banks for the loss of income from shrinking interest margins. Rather the major sources of income compensation have come from expanding assets and lowering costs. Valentine and Ford conclude from this evidence that banking must be a competitive industry because it has not been able to recoup all of its lost income, arising from squeezed interest margins, via increased fees.[63] However, this evidence is also consistent with the thesis that banks are operating in a number of differentiated markets - some of which are more competitive than others. As interest margins have been squeezed in the more competitive markets, banks shifted to fee increases but, subject to constraints imposed by market demand curves, the largest fee increases would be imposed in the least competitive of their product markets. According to the ACCC analysis in WBC / BML and CBA / Colonial, transaction accounts markets were identified as raising competition concerns and this would be consistent with the thesis advanced above.[64]

A critical issue that arose in the CBA / Colonial case was the degree to which market convergence (in both its product and geographic dimensions) had taken place since the WBC / BML assessment. Product convergence is associated with customer’s switching between product types that previously were not acceptable as close substitutes. In the extreme, this would see a single product market for all financial service products. Subject to data availability, product convergence should be reflected in changing cross-elasticities of demand and changes in consumer holdings of different product types. Geographical convergence involves the merging of previously separate geographical market areas into one larger market. Alternatively, it can be seen as the extension of one market area to include surrounding areas. For Australian banking, geographical market extension is seen as extending through local to state and finally national geographical boundaries.[65]

The driving forces behind both forms of convergence include; product and delivery innovation (predominantly electronic based access advances), diversification of product offerings by financial institutions (increasingly outside of their traditional areas of operation), product awareness (influenced by the spread of information and marketing advances), and diminishing customer inertia (willingness to use new products and delivery systems).

While there is no question that retail banks and other financial institutions are offering an expanded range of financial services products, genuine product market convergence requires a significant proportion of customers to be prepared to switch between traditional and new products. As noted earlier, the CBA in its submission to the ACCC had argued for product convergence between traditional bank savings and transaction products, and what have traditionally been regarded as non-bank investment products, including equities. Valentine and Ford adopt a similar broad product definition to that adopted by the CBA for retail products and, in taking this view, they focus on the same factors of potential ease of switching across product types, particularly due to expanded electronic access systems, common involvement by institutions across products, production complementarities and common national deposit and loan rates charged by banks.[66] Whilst the latter is a particularly compelling point we maintain the critical factor is the preparedness of customers to substitute products. The evidence suggests there remains a considerable degree of customer inertia in some markets and hence separate product markets can still exist within the broad category of retail banking and investments.

Attempts to estimate cross-price demand elasticities for traditional deposit products and equity investments were undertaken as part of our analysis for the ACCC and whilst the results were not conclusive, we still failed to find significant cross-price effects between deposit products and equity investments. However there was suggestive evidence that equity holdings were driven mainly by equity returns (own price effects) and household borrowing, and deposit holdings were driven by household income (wealth / income effects).[67] Collaborative evidence of a simple nature can be seen in Table 2 below. In the 12 year period from June quarter 1989 to June quarter 2001, the ratio of household sector holdings of bank deposits to household disposable income per quarter changed very little. In stark contrast the household sector more than doubled its holdings of equities (including holdings in public trusts) relative to household income. A similar dramatic increase occurred in bank loans as a proportion of household disposable income.

These results support the claim that bank deposits and equity investments belong in separate product markets and have not been readily substitutable from the perspective of consumers. They raise questions about the extent to which convergence of traditional bank savings / transaction products has occurred with newer non-bank investment products.

We also note that the role of electronic service delivery channels in determining market definitions is somewhat confused and liable to misinterpretation. On the one hand, electronic delivery may have no impact on the product dimension of a market. It may merely change the manner in which the financial service is delivered (eg via EFTPOS, telephone or the internet rather than over the branch counter). It will not contribute to product convergence unless it stimulates product substitution. The impact of electronic service delivery on geographic convergence is a separate issue. If it enables a particular service to be delivered outside the previous geographic area serviced say by a local bank branch, then it will impact on the geographic dimension of that product market and hence contribute to a geographical extension of this market. It is expected that on-line services will be a significant factor in this process. However, at the time of the CBA / Colonial analysis, the evidence available supported the view that internet banking, despite its rapid growth, still represented a very small proportion of total banking transactions.[68] None-the-less, for certain retail banking products the ACCC accepted that the geographic dimension had widened, for example home lending was considered a national market in the CBA / Colonial analysis, whereas it was identified as approaching national in the earlier WBC / BML assessment.

Contestability and Barriers to Entry

The ACCC has identified significant barriers to entry into retail banking in all three of the bank merger investigations discussed here, although their extent varies from product market to product market. An additional issue in the CBA / Colonial case was the lower contestability of rural and regional banking markets relative to metropolitan markets, suggesting that barriers vary depending on the geographic location as well as the product market in question.

Barriers to entry and the issue of contestability in banking is a highly significant and contentious issue. It also has a dynamic dimension in respect of the extent to which entry barriers (if they exist) are currently being lowered by new technology and service delivery mechanisms employed by financial institutions. Traditionally, high entry barriers in retail banking had been represented by such attributes as: extensive branch and agency networks; customer inertia, as reflected in the considerable time and effort (opportunity cost) required to compare and change suppliers for products such as transaction accounts and credit cards (relative to the perceived benefits); and the importance of brand name and a perception of security, especially for savings products. More recently, new technology, remote service delivery strategies and the appearance of niche suppliers for certain banking products has arguably provided evidence of a lowering of entry barriers. Is this sufficient to support claims by some that the banking industry is contestable?[69]

Economies of scale have long been of considerable interest for the banking industry, however there is considerable dispute within the literature as to whether economies of scale constitute an entry barrier. In the Stigler sense, economies of scale do not constitute a barrier to entry if entrants and incumbents can access the same cost curve.[70] Entrants are not precluded from attaining the same cost levels as incumbents, they simply have yet to achieve the required production levels. However, an alternative interpretation of costs and entry barriers is provided by Demsetz who argues that focus on narrow costs of production by existing incumbents is unproductive if it ignores “ ... the costs that must be incurred to create and to maintain a good reputation, to bear the risks of innovation, and to build a scale of operations appropriate to the economical servicing of consumer demands...” and these include “... entitlements to the fruits of past investment.”[71]

In Australian banking there has been a need to build a large customer base in order to operate successfully in the retail markets. A number of delivery models are available but all bring with them significant risk for new entrants. This is particularly so given that the incumbents have all existing delivery modes covered and have demonstrated a capacity to steadily deliver cost efficiencies. The majors have a long history of investment in reputation and customer bases as well as physical delivery networks. The risks are high for potential new entrants if they are unable to enter and exit without incurring substantial capital losses. Ease of entry and exit is the essential feature of a contestable market.[72] Recent history also suggests that although many new entrants have appeared in the Australian banking industry, they have found it difficult to gain significant market share. In September 2001, APRA listed 53 licensed banks in Australia, however 45 of these banks had a market share of less than 2%.[73] Very few of these banks have successfully entered the retail sector of the industry. As well, the 5th and 6th ranking retail banks, St George (6%) and Suncorp-Metway (3%), entered banking with an established retail base from their previous existence as retail non-bank financial intermediaries.

Table 2: Household Sector Holdings of Financial Assets


June 1989
June 2001
Increase in Ratio
Bank Deposits / Household Disposable Income
1.60
1.74
8.8%
Equities / Household Disposable Income
0.92
2.36
256.5%
Bank Loans / Household Disposable Income
1.59
3.14
197.5%
Equiteies / Bank Loans
0.58
0.75
29.3%

What the CBA / Colonial and WBC / BML investigations, and the pattern of new entrant activity in recent years, suggest, is that many aspects of retail banking may still have considerable entry barriers associated with them. While niche players may cherry-pick certain high-margin product markets, such as home lending, other products such as transaction accounts may prove less attractive for niche players to enter. Electronic delivery through ATM’s, call centres and internet banking may appear to lower these barriers, but these new delivery systems can require significant investments, especially ATM networks. Further, evidence from the United States suggests that internet only banks (“clicks”) have been far less successful in generating profitable customer growth than traditional bricks and mortar banks (“bricks”) that are expanding into on-line delivery, coupled with retention of their branch network (“bricks and clicks”).[74] In Australia the progress of ING Direct should be closely monitored to see if its “clicks” model is able to successfully compete with the majors extensive “bricks and clicks” models. It is starting from a low market share of around only 0.8%.[75]

For the incumbents, the risk is that a new entrant such as ING, operating with a different delivery mode, may prove itself capable of capturing a significant market share. Then the dual delivery networks (with considerable sunk investments) could turn into a significant cost disadvantage. If so, then the majors have another incentive to engage in merger activity. Rationalisation of existing branch networks will not only bring cost savings, they will also minimise the potential cost disadvantage associated with a “bricks” network, should the banking consumer change their preferences and fully embrace a “clicks” delivery mode.

We suggest the theoretical arguments supporting Australian banking industry contestability should be tempered by commercial and empirical reality, and reference to the recent history of, at best small scale, de novo entry in to retail banking. This point is emphasised by Martin as a feature that is common to many industries.

Real-world data suggest that entry is a hazardous enterprise ... Most entrants have little if any influence on market performance because most entrants exit, and fairly quickly. Market performance is likely to depend much more on the flow of entrants who survive than on gross or net entry flows, and the theoretical and empirical analysis of entrant survival remains a field that is largely unexplored.[76]

The issue of new entrant survival and effective impact links to the important role of the regional banks in bank merger policy. As should be clear from the previous discussion of recent bank merger investigations, it is the regional banks, particularly those who have been new entrants, who have been removed from the market.

Valentine and Ford provide a completely different focus on the issue of contestability. They predict that a case for a breakdown in the four pillars policy will be proposed based on the need for mergers amongst the four majors to enable them to compete effectively against larger overseas banks in the international arena.[77] In terms of the TPA, such a case might be successfully proposed under the public interest test that applies under the authorisation provisions. We see little chance, however, of its acceptance by the ACCC under the present competition-only test that applies under s 50, given the focus on domestic retail markets in past ACCC decisions. Valentine and Ford argue that ultimately the future of mergers between the big four banks will be decided in the political arena; it is here globalisation arguments may carry much more weight.

Regional Banks

A consistent theme to emerge from the recent history of bank mergers in Australia is the role of the smaller regional banks in providing a competitive spur to the big four national banks. The importance of regional banks to the ACCC’s thinking on competition in retail banking was most clearly articulated in the WBC / Challenge case, where the decision not to oppose the merger was linked to the continued competitive vigour of the remaining Western Australian regional, Bank West. Indeed, as noted above, some observers came to see this as an ACCC “policy” of maintaining at least one regional bank in each state market. Yet, both the WBC / BML decision and the CBA / Colonial decision resulted in the removal of significant remaining state-based regional banks - BML from the Victorian market and Trust Bank from the Tasmanian market respectively. As noted below in regard to the remedies accepted in these two decisions, they can both be seen as somewhat special cases, and consequently the importance of retaining the competitive vigour of regional banks in state-based retail banking markets should not be dismissed.

To emphasise this point, it is worth quoting observations made about the NSW market in the press release accompanying the CBA / Colonial decision: “The ACCC’s concerns in NSW were partially alleviated by the presence of another well-established regional bank. In the absence of such competition in NSW, it is likely that the acquisition would have generated substantial competition concerns”.[78] Further, the health and vitality of Australia’s regional banks as low cost and profitable players in state markets has recently been confirmed by their robust share prices and favourable analyst reports.[79]

Precedential Value of Remedies

Observers and practitioners might be tempted to place some precedential value on the combination of access undertakings and behavioural undertakings that the ACCC has accepted as resolving the competition concerns identified in the WBC / BML and CBA / Colonial investigations. Others may see it as evidence of the ACCC shifting from structural undertakings to embracing alternative quantity (or price cap) undertakings, which some commentators have argued can help ensure that horizontal mergers are socially desirable while minimising the information burden on the regulator.[80] On both counts a word of warning and explanation is needed.

The first point that needs to be made is that the banking industry has a number attributes which make the pursuit of traditional antitrust remedies for mergers more difficult than is the case in almost any other industry.[81] This is especially the case in jurisdictions like Australia and Canada, which have banking systems that are characterised by a small number of large, nationally branched institutions. In such circumstances, other regulatory and policy considerations as well as the limited number of potential buyers may limit the antitrust remedies available to competition agencies.[82]

The second point to emphasise is that each merger assessment is unique to the set of facts associated with that case. The remedies seen by the ACCC as acceptable in any given case are also unique to the competition problems identified in that case.

In the WBC / BML case, only one of the six retail banking markets defined by the ACCC was considered likely to substantially lessen competition post-merger, namely the Victorian market for transaction accounts. The quasi-structural remedies accepted from WBC in that matter related specifically to that one market. Access undertakings relating to ATM and EFTPOS networks were intended to ease entry for new and smaller players wanting to offer transaction services in Victoria. The behavioural undertakings relating to management autonomy, preservation of rights for certain classes of BML customers and opening hours were considered by the ACCC only because of some unique features of the WBC takeover. Specifically, WBC had publicly painted the acquisition as a “reverse takeover” whereby the merged entity would enjoy a degree of autonomy within the Westpac Banking Group and use BML’s brand name and customer service culture to bolster its position in the Victorian market.

In the CBA / Colonial case a larger number of retail banking markets were identified as likely to substantially lessen competition post-merger than in the WBC / BML matter. In total there were five, encompassing three markets in Tasmania (transaction accounts, deposit products and small business banking) and two in regional NSW (transaction accounts and small business banking). Concerns were also expressed by the ACCC about competition in other markets (deposit products in regional NSW and credit cards nationally). Given the more pronounced competition problems identified in CBA / Colonial, especially in Tasmania, the ACCC did seriously explore structural undertakings as a potential remedy. However, as noted above, significant constraints existed in pursuing such a course. Another complicating factor in this particular case was the identification of NSW markets where substantial lessening of competition had also been seen as likely. These markets did not lend themselves so easily to large-scale divestiture remedies. While individual branch divestitures in specific locations in rural NSW may have provided an alternative structural solution, the ACCC could not be certain of a large field of interested and acceptable buyers for each of these branches. Declining full-branch representation by banks in rural and regional areas, and associated fear of declines in the product choice, pricing competition and service levels, has been a major concern to rural and regional consumers and policy makers alike.[83]

The behavioural undertakings provided by the CBA to ensure the benefits of pricing, service levels and product range in more competitive metropolitan` markets are available to consumers in regional NSW and in Tasmania should be seen in relation to the concerns about banking services in regional areas. The quasi-structural undertakings relating to access must also be seen in light of the challenges facing traditional (branch banking) models of service delivery in regional Australia. The access remedies were intended to assist the entry and expansion of financial institutions that are exploring new models for servicing customers in these areas of Australia, for example through partnerships with local communities (e.g. franchising or community banks) or through alliances with existing non-bank players (e.g. pastoral houses or trustee companies), as well as to assist those smaller institutions pursuing more traditional branch banking approaches.

6. Conclusions

This paper has attempted to add to the Australian literature regarding the competition (antitrust) analysis of mergers involving medium to large banking institutions. The key focus has been an analysis of the approach taken by the ACCC in three important bank merger investigations - WBC / Challenge in 1995, WBC / BML in 1997, and CBA / Colonial in 2000.

One key conclusion to emerge from the recent CBA / Colonial assessment is that individual product markets persist despite some product and geographic convergence in the financial services sector. Generally speaking, the geographic dimension was wider (tending to national) for bank assets (lending products) and was narrower (tending to local or state-based) for bank liabilities (deposit and transactions products) and for small business and agribusiness banking.

As regards the ACCC’s use of market share thresholds, we conclude that these are generally adequate and, in any case, their use as a screening device should mean that, in and of themselves, they are not crucial to any final conclusions reached in bank merger assessments. On the issue of coordinated conduct, we are not convinced by the arguments of some critics that the industrial organisation literature and / or consideration of commercial realities can provide a definitive rebuff to ACCC concerns of possible coordinated market power in the banking industry. Against this, we have sympathy with the argument that the ACCC has not always clearly articulated the basis for its concerns on coordinated conduct, either in terms of the pointers provided by the industrial organisation literature or the empirical evidence regarding interaction between firms in the markets in question.

Lowering of entry barriers and increased contestability has been accepted by the ACCC in some markets, such as in home lending, yet in other product markets entry barriers appear to remain high. The potential exists for new technology and service delivery mediums to impact on entry barriers in other product markets, although this process is yet to play itself out. Effective entry by new retail banking players into rural and regional areas remains problematic. In each case, what is relevant for competition analysis is the extent to which new sources of supply can effectively constrain the pricing (and non-price) behaviour of the merged entity in the post-merger environment.

The loss of regional banks (the medium size competitors in Australian banking) remains a concern, particularly as some, including Colonial, had the potential to grow to major national bank status. While niche players have emerged in some product markets, and some smaller retail banks have expanded their geographic coverage, for example Bendigo Bank, the four major national banks continue to dominate many categories of retail banking. In this context, the role of the regional banks in many state markets remains important.

Another key conclusion to emerge relates to limitations on the remedies available to competition authorities to address identified competition problems in concentrated banking systems like Australia’s, where other regulatory or policy objectives may also constrain the options available. Such constraints were encountered by the ACCC in the CBA / Colonial case. What the perceptive reader may also note is that in the presence of constraints on structural remedies, the scope for identifying adequate non-structural remedies to major bank mergers under the TPA, s 50 may be extremely limited in an increasingly concentrated banking industry. Or to put this another way, the scope for flexible regulatory responses to bank mergers may in fact diminish past a certain point.[84]

Each merger scenario needs to be considered in the context of the facts that apply at the time a particular merger is proposed, and so predictions about how the ACCC would treat future mergers in the banking industry must be treated cautiously. The response to the WBC / Challenge, WBC / BML, and CBA / Colonial cases may not be a good guide to the ACCC’s approach to future bank mergers that appear likely to lead to substantial anti-competitive detriment in a number of product and geographic markets. The latter circumstance is most likely to arise with mergers involving the big four retail banks; especially where there is consolidation amongst these players. This suggests that even if the four pillars policy were to be abandoned, mergers of this scale may encounter regulatory hurdles under the TPA, s 50 and if they were to proceed, may need to be dealt with via the broader public interest test that applies under the authorisation provisions of the TPA.


[1] The March 2000 issue of the Australian Economic Review has recently contributed to the literature on mergers policy and trade practices regulation in Australia through its policy forum entitled “Merger Policy in Australia”. See Aust Eco Rev 33 (2000) 65.

[2] See, for example, a symposium in the Summer 1999 edition of the Antitrust Bulletin entitled “Bank Mergers - Constructing Markets and Calculating Shares” and a 1999 “Special Issue on The Consolidation of the Financial Services Industry” in the Jnl of Banking and Finance. Of particular note are the papers by Gerald A Hanweck and Bernard Shull, ‘The Bank Merger Movement: Efficiency, Stability and Competitive Policy Concerns’ (1999) 44 Antitrust Bulletin 251, Anthony W Cyrnak and Timothy H Hannan, ‘Is the Cluster Still Valid in Defining Banking Markets? Evidence from a New Data Source’ (1999) 44 Antitrust Bulletin 313, and Allen N Berger, Rebecca S Demsetz and Philip E Strahan, ‘The Consolidation of the Financial Services Industry: Causes, Consequences, and Implications for the Future’ (1999) 23 Jnl of Banking and Finance 135.

[3] The FSI, which commenced in May 1996 and reported in March 1997, was headed by the then President of the Business Council of Australia, Stan Wallis. See the Financial System Inquiry, Final Report (AGPS, Canberra, 1997), pp 417-476 regarding bank mergers. A brief discussion of the FSI, and especially the impact on ACCC assessments of bank mergers at the time, is provided below.

[4] Ian R Harper, ‘Mergers in Financial Services: Why the Rush?’ (2000) 33 Aust Eco Rev 67, at 71-72.

[5] Jim Rose, ‘The ACCC and the Westpac / Bank of Melbourne Merger’ (1998) 26 ABLR 358.

[6] Tom Valentine and Guy Ford, ‘Bank Mergers in the Australian Financial System - Should the Pillars be Pulled Down?’ (2001) 20 Aust Eco Papers 4.

[7] Aspects of this paper are based on a more substantial monograph by the authors - Garry K Goddard and Greg Walker, Bank Mergers in Australia: Competition Assessment of the Commonwealth Bank of Australia’s Acquisition of Colonial Limited (Working Paper Series No. 2/01, Faculty of Commerce, Charles Sturt University, Bathurst, 2001, April).

[8] ACCC analysis of significant bank mergers is conducted independently but is reported to the Treasurer before a decision is announced. The process acknowledges the Federal Treasurer’s reserve powers in relation to bank merger approvals under the Banking Act 1959 (Cth) and the Financial Sector (Shareholdings) Act 1998 (Cth). Under s 14 (1) of the latter piece of legislation, the Treasurer may allow an application for a shareholding in a financial services company to exceed 15 percent where it is in the “national interest”. If a merger is considered not to be in the national interest it can be blocked by the Treasurer.

[9] Section 50 of the TPA prohibits mergers or acquisitions within Australia, which fail the SLC test. Mergers or acquisitions can also be assessed by the ACCC through the “authorisation” procedures contained in Pt VII of the TPA. Authorisation allows parties to a merger to point to a wide range of offsetting public benefits, which may enable a merger to proceed that would otherwise be opposed under s 50 of the TPA. See n 14 below.

[10] A more comprehensive discussion of the ACCC’s merger assessment procedures is provided in Australian Competition and Consumer Commission, Merger Guidelines (ACCC Procedural Guide Series, Canberra, 1999, June).

[11] In addition to the end uses and physical characteristics of the products in question, product definition can be informed by information such as pricing and marketing strategies of suppliers, including those of the merger proponents, interviews with market participants and regulators, consumer surveys and market research, and by data on price movements and relativities between products. Geographic definition can be informed by information on buyer commuting patterns, the delivery mechanisms used to supply the product, and consumer surveys and market research. See ibid, pp 35-37.

[12] It should be noted that where non-national geographic markets (eg state, regional or local) are identified, imports need not come only from off-shore suppliers but could also include Australian-based suppliers without a physical presence in that market.

[13] As with the antitrust of bank mergers, the vast bulk of the studies addressing the efficiency gains from bank mergers, including economies of scale, are from the United States. See in particular Hanweck and Shull, op cit, at 259-263, Stephen A Rhoades, ‘The Efficiency Effects of Bank Mergers: An Overview of Case Studies of Nine Mergers’ (1998) 22 Jnl of Banking and Finance 273, and Allen N Berger and David B Humphrey, ‘Efficiency of Financial Institutions: International Survey and Directions for Future Research’ (1997) 98 European Jnl of Operations Research 175. Analysis of the nature of efficiencies in Australian banking has been provided by Greg Walker, ‘Economies of Scale in Australian Banks 1978-1990’ (1998) 37 Aust Eco Papers 71, Greg Walker, ‘Cost Structures and Efficiency in Banking’ in Tom Valentine and Guy Ford (Eds), Readings in Australian Financial Management (Allen & Unwin, Sydney, 1999), pp 115-129, and Milind Sathye, ‘X-efficiency in Australian Banking: An Empirical Investigation’ (2001) 25 Jnl of Banking and Finance 613.

[14] Offsetting public benefits in the context of a merger authorisation refer to the potential benefits that would flow to the wider community from a proposed merger, such as expanded exports, increased employment or improved efficiencies. These may be sufficient to “offset” the likely anti-competitive detriment (ie SLC in the various markets) associated with a merger and allow a merger to proceed. For the application of the authorisation provisions of the TPA to merger cases see ACCC, op cit n 10, pp 62-74.

[15] A detailed discussion of the analysis underpinning this decision was provided at the time - see Trade Practices Commission, Westpac / Challenge: Background to TPC Decision (TPC, Canberra, 1995, September). A more wide-ranging discussion of the WBC / Challenge decision is provided in a speech delivered by the ACCC’s Chairman to the Australasian Banking Conference on 29 September 1995. The speech was later published in the first issue of the ACCC’s journal - see Allan Fels, ‘Competition Policy and Banking’ (1996) 1 ACCC Journal 15.

[16] TPC, op cit, p 2.

[17] See Fels, op cit, at 17-18.

[18] TPC, op cit, pp 7-8.

[19] Ibid, pp 9-10.

[20] Ibid, pp 8-9.

[21] See Fels, op cit, at 22.

[22] Australian Competition and Consumer Commission, ACCC Not to Oppose St George / Advance Bank Merger (ACCC News Release, Canberra, 1996, October). See also the unattributed summary in the ACCC’s journal - ‘St George Bank Ltd and Advance Bank Ltd’ (1996) 6 ACCC Journal 9.

[23] Australian Competition and Consumer Commission, Westpac Banking Corporation / Bank of Melbourne Limited: Background to Decision on Merger Proposal (ACCC, Canberra, 1997, July).

[24] Ibid, p 3.

[25] The ACCC also identified “corporate banking” as a further (national) market but, due to the Bank of Melbourne’s lack of presence in this market, it did not discuss this market in any detail. See ibid, pp 3-4.

[26] Ibid, p 35.

[27] For the evidence on bank branches see ibid, pp 12-13, and FSI, op cit. Evidence from a 1996 Leading Edge & AGB McNair survey suggested that 93 percent of consumers needed access to a bank branch at some point to do their day-to-day banking (FSI, p 451). Further evidence from an October 1996 RDA survey for Citibank indicated that over three quarters of consumers preferred to make deposits at a branch and over half preferred to make credit card payments at a branch (FSI, p 451). Evidence provided by the Reserve Bank of Australia suggested that virtually none of the regional banks (for which meaningful data was available) were able to raise deposits in states where they did not have any branch presence at all (FSI, p 448).

[28] Ibid, pp 417-476 covers mergers and acquisitions. The key FSI recommendation regarding merger policy was that the (six pillars) policy restricting mergers in the financial sector should be discarded and that merger approvals be subject only to competition assessment under s 50 of the TPA and prudential assessments under relevant insurance and banking laws. In the event, a policy ban on mergers between the big four banks was retained by the Treasurer and remains in force to the present day, although restrictions on mergers involving the two largest insurance companies were relaxed.

[29] The most salient of the FSI findings on ACCC merger review were, that: the cluster approach to product market definition should be “closely questioned”; retail transaction accounts and small business banking were likely to be critical in future bank merger assessments; the geographic definition for some retail banking products was moving toward national markets, although the extent of this varied depending on the product; and regional banks had provided important competitive pressure on the four majors but their position was not immutable going forward. See ibid, pp 417-416 & 428-473.

[30] Section 87B of the TPA makes provision for the ACCC to accept court enforceable undertakings as a remedy to a wide range of matters under the TPA - see Australian Competition and Consumer Commission, Section 87B of the Trade Practices Act (ACCC Procedural Guide Series, Canberra, 1999, August). In merger cases, parties to a transaction can offer s 87B undertakings to address competition concerns identified by the ACCC, and, if sufficient and workable, these undertakings may enable merger transactions which would otherwise be opposed by the ACCC, to proceed. See ibid, pp 9-11, and ACCC, op cit n 10, pp 75-82.

[31] Further detail on the WBC / BML merger assessment and the undertakings accepted by the Commission can be found in ACCC, op cit n 23, and Rose, op cit. More condensed summaries can be found in Donald G. McFetridge, Competition Policy Issues: Research Paper prepared for the Task Force on the Future of the Canadian Financial Services Sector (Task Force, Ottawa, 1998, September), pp 65-71, Dave Poddar, ‘Australian Bank Merger Policy - A Restraint on Efficiency?’ (1999) 20 European Com L Rev 40, at 41-43, and Garry K Goddard, ‘Bank Mergers Policy and Competition Law Enforcement: A Comparison of Recent Experience in Australia and Canada’ (2000) 15 Banking and Finance L Rev 181, at 211-221.

[32] On the Canadian approach see ibid, at 221-241.

[33] On the latter see Cyrnak and Hannan, op cit.

[34] See Rose, op cit.

[35] See Australian Competition and Consumer Commission, ACCC Not to Oppose Commonwealth Bank / Colonial Merger (ACCC Media Release and Background, Canberra, 2000, May). A more detailed analysis of the CBA / Colonial assessment, with particular emphasis on the definition of markets and the nature of the competition remedies accepted by the ACCC, can be found in Goddard and Walker, op cit, pp 14-34.

[36] It is not proposed to provide detailed coverage of the analysis leading to the rejection of the CBA definitions in this paper, although a brief description of the key product and geographic definitions adopted by the ACCC is provided below. Interested readers are encouraged to consult ibid, pp 17-24 for more detail on the theoretical and empirical underpinnings of the product and geographic market definitions employed in the CBA / Colonial investigation.

[37] The ACCC’s rejection of this claim was based in large part on a detailed consideration of various inherent characteristics (eg risk, return and liquidity) and consumer behaviour relating to these product types, including an estimation of the cross-price demand elasticities (substitutability) for bank deposits and equity investments. See ibid, pp 19-23.

[38] The cluster of banking services was seen to encompass credit products, transaction and cash facilities, advisory services, and credit card and debit card merchant services, identified also as a separate component of the credit card market. On the SME banking cluster, see ACCC, op cit n 35, pp 6-7, and, for credit card merchant servicing, pp 4-5.

[39] Ibid, pp 1 & 5-9. For transaction accounts, the merged entity was estimated to control close to 50 percent of the market in Tasmania and over 30 percent in NSW, where the CR-4 ratio would be close to 80 percent. For deposits, the merged entity was estimated to have more than 50 percent of the market in Tasmania. For NSW however, the threshold tests were not likely to be exceeded, with estimates of the merged entity at just under 25 percent and a CR-4 ratio of less than 60 percent. For SME banking the data were less reliable, however the merged entity was expected to control in excess of 50 percent of the relevant market in Tasmania and 25 to 33 percent in NSW, where the CR-4 was estimated at close to 80 percent.

[40] For credit card issuing see ibid, p 4, and, for agribusiness banking, p 9.

[41] Rabobank, a specialist foreign agribusiness bank, has expanded its presence in rural Australia via ownership of the Primary Industry Bank of Australia (PIBA). Elders had recently announced plans to provide banking services through its extensive rural branch network via a relationship with the Bendigo Bank.

[42] The expansion of newer players such as community banks and franchised credit unions into regional areas could be taken to suggest that entry barriers are low or reducing. Against this, it should be noted that these institutions are pursuing different business models to traditional retail banks and it is not clear the extent to which they offer a full range of products / services that are close substitutes to those offered by the larger regional and nationally-branched banks. It remains an open question as to whether these newer players can constrain the pricing behaviour of the merged CBA / Colonial entity and the other majors in regional areas. This issue will require a more detailed consideration by the ACCC in the future, especially if the four pillars policy were to be abandoned and mergers between the major banks are allowed to proceed.

[43] More detail on the analysis of imports, barriers to entry and other merger factors in the CBA / Colonial investigation is provided by Goddard and Walker, op cit, pp 24-28, see especially table 4.

[44] The importance of branch banking, customer inertia and local geographic markets for retail commercial banking has long been a cornerstone of US antitrust, see for example Hanweck and Shull, op cit, at 256-270, Myron L Kwast, ‘Bank Mergers: What Should Policy Makers Do?’ (1999) 23 Jnl of Banking and Finance 629, at 631-632, and Stephen A Rhoades, ‘Retail Commercial Banking: An Update on a Period of Extraordinary Change’ (2000) 16 Rev of Industrial Organisation 357, at 362-635. Concerns about the degree of competition in small business banking and support for an SME banking cluster have emerged in recent reviews of the UK banking system, see Don Cruikshank, Competition in UK Banking: A Report to the Chancellor of the Exchequer (Her Majesty’s Stationary Office, 2000, March), chs 4 & 5 and annex D2, and the Competition Commission (UK), Supply of Banking Services by Clearing Banks to Small and Medium-sized Enterprises (Report to Parliament by the Secretary of State for Trade and Industry and the Chancellor of the Exchequer, 2002, March). A cluster of small business banking services has also been identified in the US, see Kwast, op cit, at 630-631. Concerns about the role of interchange fees in electronic networks both as a barrier to providing transactions services and a source of higher prices for consumers has been highlighted recently in the UK, see Cruikshank, op cit, at ch 3 and annex D3 & D4, and in the US see David A Balto, ‘The Problem of Interchange Fees: Costs Without Benefits?’ (2000) 21 European Com L Rev 215. For a discussion of the similarities between the structure of the banking industry and the findings of bank merger assessments by competition regulators in Canada and Australia, see Goddard, op cit.

[45] See ACCC, op cit n 35, pp 7-9 on the CBA / Colonial undertakings.

[46] See ACCC, op cit n 10, pp 75-78, and ACCC, op cit n 30, pp 9-10.

[47] See ACCC, op cit n 35, pp 7-8.

[48] Ibid, p 8. Dispute resolution procedures (using independent experts) were included for each of these measures so as to expedite commercial negotiations and minimise the resource implications for the ACCC. The merged entity was also obliged to offer for sale or lease certain vacated banking sites in regional NSW and Tasmania, to new or smaller financial institutions, subject to certain qualifications.

[49] Ibid, pp 8-9. Metropolitan NSW was taken as the benchmark for setting the behavioural undertakings as it was generally regarded as the most competitive geographic area for the supply of financial services in Australia. To minimise the resource implications for the ACCC’s monitoring of these behavioural obligations, the undertakings contained provision for an independent monitor to be appointed for Tasmania and for regional NSW, and to report to the ACCC on compliance with these obligations.

[50] The undertakings were provided for a five-year period, with provision for the merged entity to be released from the undertakings after three years if the ACCC and CBA agreed that there have been no material breaches in that time. Readers who want to examine the precise conditions applying in the undertakings are advised to consult the full undertakings, which are in the public domain and are available through the ACCC from its public register (reference D00/14462) or from the ACCC website (www.accc.gov.au).

[51] See Rose, op cit, at 359-360.

[52] Joshua S Gans, ‘Competitive Balance Argument for Mergers’ (2000) 33 Aust Eco Rev 83, at 85.

[53] Stephen Martin, Advanced Industrial Economics (Blackwell, Cambridge Mass. 1993). Note the index is variously referred to as the Herfindahl-Hirschman, HHI or the Herfindahl index. It is measured as: H = S si2 where si is the market share of the ith firm in the industry. When applied in fractional as opposed to percentage share form, H=1 for a monopoly and H approaches 0 for a highly competitive industry with a large number of firms.

[54] Rose, op cit, at 360-363 draws on the industrial organisation literature to argue that conditions conducive to coordinated conduct (particularly the maintenance of tacitly collusive agreements through the easy detection and punishment of cheating) were not present in the Australian banking industry, and hence the ACCC should not have opposed the WBC / BML merger.

[55] In terms of a successfully coordinated strategy, it is possible that a small number of major banks could choose to ignore the products of fringe players, and focus only on each others key “indicator” items as benchmarks for increases in fees and charges (or reduced service levels / access options). Customer inertia would help facilitate such a strategy. Rose also fails to adequately address other aspects of the industry that may give competition regulators cause for concern regarding coordination, especially the presence of industry associations and fora, and shared operations or infrastructure. See ACCC, op cit n 10, pp 58-59.

[56] Reserve Bank of Australia, Reform of Credit Card Schemes in Australia (RBA, Sydney, 2001, December).

[57] See Gans, op cit, at 90.

[58] Ibid, at 84.

[59] See Rose, op cit, at 361, quoting from ACCC, op cit n 23, p 4.

[60] See Valentine and Ford, op cit, at 38.

[61] Reserve Bank of Australia, ‘Bank Fees in Australia’ (2001) July RBA Bulletin 1.

[62] Examples include: from 1995 to 2001, average account servicing fees rose from $2 to $5 and average counter withdrawal fees rose from $1 to $2.75. See ibid, p 3.

[63] See Valentine and Ford, op cit.

[64] It should also be noted that a more subtle constraint on increases in fees and charges for bank transaction products might also be felt by the majors through adverse public perceptions and political pressures placed on banks not to raise fees for what some regard as basic banking products.

[65] Geographic convergence would be evidenced by increasing numbers of consumers accessing particular products from suppliers without a physical presence within the previous market boundaries (e.g. significant penetration of a state-based market by an out-of-state supplier). However, it might also be evidenced by a traditional supplier retaining its customer base despite either the supplier or its customers relocating to a location that was once outside the geographical market dimension. In the extreme, geographic convergence / extension would see all geographic markets being not simply national but international in scope.

[66] See Valentine and Ford, op cit.

[67] See Goddard and Walker, op cit, p 23.

[68] See Australian Bureau of Statistics, Use of the Internet by Householders (Cat No. 8147.0). The May 1998 survey confirmed that less than 1 per cent of adults used internet banking. In its May 2000 survey the figure had risen to 5 per cent.

[69] See Harper, op cit, Rose, op cit, and Valentine and Ford, op cit. For example, Rose argues in regard to the WBC / BML analysis that some of the entry barriers identified by the ACCC (e.g. bank branches and brand awareness) are simply legitimate costs to doing business in the industry (on the scale and efficiency of incumbents) rather than entry barriers per se.

[70] See George J Stigler, The Organisation of Industry (Irwin, Homewood, 1968).

[71] Harold Demsetz, ‘Barriers to Entry’ (1982) 72 American Economic Review, p 56.

[72] See William J Baumol, ‘Contestable Markets: An Uprising in the Theory of Industry Structure’ (1982) 72 American Economic Review 1.

[73] Australian Prudential Regulation Authority, Australian Banking Statistics (APRA, 2001, September).

[74] See, for example, Australian Financial Review, ‘On Balance, Online Banking has Failed’, 12 January 2001, p 36.

[75] See APRA, op cit.

[76] See Martin, op cit.

[77] See Valentine and Ford, op cit.

[78] ACCC, op cit n 35, p 1.

[79] Regarding the recent performance of some of the regionals, see Australian Financial Review, ‘Regional Banks Outperform the Big Boys’, 30 January 2001, p MW1 & p MW5.

[80] On the former, see Nicole Tyson and Lynden Griggs, ‘Mergers and Metaphors’ (2001) 9 TPLJ 95. On the latter, for a theoretical exposition of this position, see Teresa Fels, Joshua S Gans and Stephen P King, ‘The Role of Undertakings in Regulatory Decision-Making’ (2000) 33 Aust Eco Rev 3.

[81] For a discussion of the multiple regulatory and policy attributes which may need to be considered in assessing bank mergers see Michael Andrews, Addressing the Prudential and Antitrust Aspects of Financial Sector Mergers and Acquisitions (Monetary and Exchange Affairs Department, Operational Paper No.MAE100/06, IMF, Washington, 2000, October).

[82] More detailed discussion of these constraints and their impact in the CBA / Colonial case is provided in Goddard and Walker, op cit, pp 28-34.

[83] Readers are referred to the House of Representatives (Hawker) Report on Regional Banking Services - Parliament of the Commonwealth of Australia, Regional Banking Services: Money Too Far Away (Report of the House of Representatives Standing Committee on Economics, Finance and Public Administration, AGPS, Canberra, 1999, March). Public policy concerns surrounding theses issues appear also to have contributed to commitments made by CBA, at the time of the announcement of the acquisition of Colonial, to maintain a minimum branch representation in certain locations in regional NSW and Tasmania and to maintain or expand the net employment effects of the transaction in each of these two regions taken as a whole. See Tyson and Griggs, op cit, at 98 for more detail on the additional commitments made by the CBA to the Tasmanian government. While these considerations are outside of the ACCC’s TPA, s 50 analysis of competition, they are clearly important to the Federal Treasurer’s “national interest” considerations.

[84] The cautionary note sounded in this paper is at odds with interpretations which simply draw a line from the WBC / BML case to the CBA / Colonial remedies and conjecture that subsequent bank mergers can and will be dealt with by the ACCC in the same manner. For such an interpretation see Tyson and Griggs, op cit, at 95 & 104-105. As noted above, such views fail to fully appreciate the extent of the problems facing alternative remedies in the CBA / Colonial case, and the case-by-case nature of the ACCC approach to merger assessments.

* The authors would like to thank Commissioner Ross Jones of the Australian Competition and Consumer Commission and those Commission staff involved in the Commonwealth / Colonial merger investigation, namely Alistair Davey, Melissa Randall, Anita George, Nicole Ryan and Mark Bailey. We are also grateful to Tony Cyrnak, Mike Andrews and Paul McCarthy, and to three anonymous referees, all of whom have provided helpful comments on earlier versions of this paper, however all omissions and errors remain those of the authors. The views expressed are those of the authors and do not necessarily represent the views of Charles River Associates or the Australian Competition and Consumer Commission.

[+] Principal, Charles River Associates (Asia Pacific) Pty Ltd, Melbourne. Formerly Director, Mergers and Asset Sales Branch, Australian Competition and Consumer Commission.

[#] Associate Professor of Banking and Finance, School of Management and Marketing, Charles Sturt University, Bathurst.§


AustLII: Copyright Policy | Disclaimers | Privacy Policy | Feedback
URL: http://www.austlii.edu.au/au/journals/JlLawFinMgmt/2002/2.html