RBA Annual Conference – 2007 Discussion

1. Ian Harper

The discussion of Claudio Borio's and Franklin Allen's papers left me thinking that history has not been kind to the Wallis Committee of Inquiry into the Australian Financial System. It would appear that banks complement financial markets, rather than substitute for them, to a far greater extent than assumed by the Committee. This even led to speculation that the Wallis architecture of creating an integrated prudential supervisor separate from the central bank might have been – at least in hindsight – a mistake.

Reading Kevin Davis' excellent overview of the literature on banking concentration, competition and stability left me feeling much better about Wallis. At the end of his paper, Kevin calls for a ‘considered and substantial review of bank merger policy arrangements’ and I could not agree more. Of the 115 recommendations presented to the Government in the Wallis Report, only one was rejected: the recommendation that there should be no outright ban on mergers among financial institutions but that all merger proposals should be considered on their merits through the usual channels of the Australian Competition and Consumer Commission (ACCC) and, in this case, the Australian Prudential Regulation Authority (APRA) and RBA investigations. Far from accepting our recommendation, the Treasurer announced that ‘for the time being’ he would not authorise mergers among any of Australia's four major banks. Thus the so-called ‘four pillars’ policy was born and it remains in place some 10 years later.

When queried, as he has been over the years, as to what might suffice to change the Treasurer's mind on four pillars, his usual answer is to refer to the perceived lack of competition in the banking industry and the need for this to change significantly before he would consider changing the policy. Kevin's paper tackles the supposed link between bank concentration and competition head on. His careful review of the literature reveals no compelling evidence of rising bank concentration in any region, including Australia; if anything, concentration appears to be falling. Furthermore, mergers in domestic markets have tended to increase the preponderance of mid-size banks even as three- and four-firm concentration ratios have remained static. The growth of cross-border banking has increased the presence of banks that operate in more than one national market, enhancing competition in markets nevertheless highly concentrated according to the usual measures.

In short, Kevin's literature review leads him to conclude that there is no clear link between concentration and competition in banking markets. It would appear that banking is becoming more competitive without becoming less concentrated. One might even conclude that it would remain competitive, or become no less so, if the industry became even more concentrated, although Kevin declines to go this far.

He warns that a thorough analysis of competitive conditions in banking would need to consider the various markets in which banks operate. The standard tests of competition in the literature, including the Panzar-Rosse H-statistic, assume a single-product firm whereas even moderate-sized banks operate across multiple markets. One would need to look carefully at competitive conditions in retail as opposed to wholesale markets and, within retail, at segments like transactions services and lending to small- and medium-sized enterprises, which have historically resisted normal competitive inroads.

In this context, Kevin identifies only one significant barrier to entry in retail banking that might usefully be removed as part of further enhancing competitive conditions. The current restriction barring foreign bank branches from taking deposits in Australia worth less than A$250,000 was designed for an earlier time when confidence in foreign bank supervisors was lower than it is today. The growing presence of foreign bank branches in Australia, their potential to enhance competition in concentrated domestic retail banking markets, and the internationalisation of bank supervisory standards through the Basel Committee all tend to undermine the rationale for this restriction.

Even given his misgivings about competitive conditions in retail markets, Kevin recognises the potential of continuing improvements in the power and reach of technology, the incidence of cross-border banking and the growth of mid-size domestic banks operating in retail as well as wholesale markets to further erode any link between bank concentration and competition. In such circumstances, it seems puzzling indeed that the Treasurer feels the need to take the option of mergers among the major Australian banks off the table. At the very least, allowing the ACCC to consider one or more merger applications would test the strength of the arguments and permit closer scrutiny of conditions in actual markets in an Australian context.

One of the fears that Kevin raises at this point is that removing the ban on mergers among the big four banks would set off a scramble to be the first to merge. He offers a brief game-theoretic analysis in support of his conclusion. An unseemly race to the ACCC, together with the political pressure that would inevitably accompany such a push, would militate against the careful analysis of the proposal which good public policy in this area would demand.

While I agree that fears of only one merger being allowed would spark just such a race, I do not agree that this need be the outcome. If the link between concentration and competition really is very weak or non-existent, there seems no reason to rule out ab initio two mergers among the majors, bringing four down to two. Of course, neither merger may be approved; but the decision to grant one and refuse the other would need to consider the competitive imbalance, including the potential for price leadership, which allowing one dominant bank to emerge might elicit. Then again, maybe even that would not matter if concentration really is not linked to competition.

Kevin's second theme is the absence of a clear link in the literature between bank concentration and the stability of the banking system. Here his paper echoes the conclusions of Franklin Allen's theoretical work, especially with Douglas Gale, in which he argues that banking instability has much more to do with the absence of complete markets and complete contracts in financial markets than bank concentration per se. Even though one might assume this conclusion should put to rest any concerns of the RBA or APRA that mergers among Australia's major banks might compromise systemic stability, Kevin makes the point that dealing with the failure of one or more very large banks post-merger is problematic. In light of this, the prudential authorities would need to be satisfied that allowing mergers among the majors would not lead their depositors, shareholders and directors to conclude that such a merged bank would simply be ‘too big to fail’ and therefore underwritten by the government de facto if not de jure. If this idea gained currency, a merger of majors could well exacerbate the risk of systemic failure by encouraging the merged bank to take on riskier assets than it otherwise would or should. The creation of one or more mega-banks might also play havoc with the proposed Financial Claims Compensation Scheme, which would be faced with concentrated risk among its insured depositors.

While these problems are real and, to some extent, mitigate the conclusion that concentration and stability are completely independent, the potential for heightened systemic risk following one or more mergers among the majors is not beyond the powers of the RBA and APRA to analyse, manage or oppose. Even if the ACCC could find no evidence of anti-competitive effects of bank mergers, the Treasurer, who would retain a right of veto under the Trade Practices Act 1974, would presumably block mergers determined to be contrary to the public interest on prudential grounds.

One of the ironies of the four pillars policy is that it actually increases the chances of one or more of Australia's major banks being the subject of a foreign takeover. Even though protected from domestic takeover, they are not immune to the dynamics of the global banking industry. Of course, the Treasurer again figures as the authority who must approve any foreign takeover of an Australian-owned entity. But, as Kevin again correctly points out, the circumstances of a serious tilt at one of our major banks by a foreign multi-national are hardly conducive to cool-headed analysis of the four pillars policy. It is likely that the Treasurer, even if only to shore up a decision to block a foreign takeover bid, might abandon four pillars in a rush in order to allow one or more of the domestic majors to mount a credible counter-bid. Neither Kevin nor I would wish to see the policy abandoned and mergers materialise without careful analysis. All the more reason to support Kevin's call for a measured review of the social benefits and costs of allowing mergers among the majors before circumstances force anyone's hand.

Notwithstanding the arguments Kevin advances, he and I both know that the major banks themselves take a different view. All four CEOs have spoken at one time or another against the four pillars policy. My question to Kevin is what he makes of this. Are the CEOs just wrong or self-serving or both? They tend to make one or more of the following claims:

  • the Australian majors are getting to be too small to participate in major capital market deals, thus losing valuable fee revenue (The lack of any Australian major among the banks leading the merger between BHP-Billiton and Western Mining Corporation, including the ANZ – BHP's near century-long domestic banking partner – is often mentioned in this regard.);
  • the Australian majors need a strong domestic base if they are to compete globally (The example of the Dutch banks, ABN/AMRO, ING and Rabobank, having been allowed to merge domestically and having since expanded internationally with great success is often cited as evidence of the so-called ‘national champions’ effect.);
  • the major banks need to spend large sums on IT, risk management systems and global brands if they are to compete with global banks, either here in Australia or overseas, and these expenditures require step-increases in investment which cannot be afforded without an increase in their scale; and
  • the only politically acceptable way to rationalise Australia's extensive network of bank branches is to allow mergers among the majors, which would keep bank branches in most locations but reduce wasteful duplication, leading to cost savings.

On the face of it, these do not seem to be ridiculous arguments. I would like to hear what Kevin says to the banks when they ask him the same questions they ask me!

I have argued elsewhere that another inquiry into the Australian financial system would be timely. I am pleased to see that Kevin agrees with me – not, I might add, an altogether common event! Ten years is a long time in financial markets and it would be wise to review the performance of Australia's new regulatory system in the light of a decade's experience. Not only would such a review consider in more detail the evolving role of banks versus markets in our financial system, but it could canvass in detail the issues surrounding mergers among Australia's major banks. It is high time that the four pillars policy was reconsidered.

2. General Discussion

There was a lengthy discussion of the nature of financial crises and how policies should address moral hazard concerns. In the context of problems in the US sub-prime mortgage market, one participant wondered whether it was possible to provide sufficient liquidity without providing respite for those who should face up to their earlier mistakes. Another participant asked whether there were ‘good crises’ that policy-makers should leave to run their course, or whether all crises were ‘bad’ due to incomplete markets. Franklin Allen thought that the difficult distinction between insolvency and illiquidity was at the core of these problems, and that preventing the latter would assist in avoiding the former. On completeness, he noted that those in the field of finance often argued that markets were complete because of the potential for dynamic trading, but ultimately he thought that the fact that crises were not that infrequent demonstrated that markets are not truly complete.

There were also a variety of opinions expressed about the optimal relationship between the monetary policy function of central banks and prudential regulation. For example, one participant argued that separating these functions made sense because monetary policy expertise was not the same as regulatory expertise and that central banks with responsibility for both functions may place insufficient weight on their regulatory responsibilities. On the other hand, Franklin Allen pointed out that the lender of last resort function of central banks is very important, citing the recent decision of the Fed to allow banks to use mortgages as collateral for their borrowings as an appropriate example. He thought that central banks had a key responsibility to provide liquidity so as to reduce asset-price volatility, which triggers bankruptcy and distress in a way that further exacerbates the original disturbance. The provision of liquidity by central banks was also important since markets often do not anticipate nor understand all possible states of the world. On the question of cooperation, he suggested that central banks might find it more difficult to carry out their responsibilities as the lender of last resort in countries in which supervisory responsibilities were dealt with outside of the central bank, and that this would certainly require a carefully coordinated response.

Much of the discussion was focused on the perceived shortcomings of particular markets and institutions. One participant was critical of the fact that in the securitisation market, originators of loans are not required to keep an equity tranche on their books. One participant argued that policy-makers somehow needed to focus more on underlying behaviours, particularly the factors that encouraged agents to all manage risks in the same way, whether it was because they adopted the same risk management ‘best practices’ – for example, the same value-at-risk models – or were over-reliant on the same prognosis from ratings agencies. In a similar vein, one participant questioned the usefulness of ‘stress testing’, arguing that in these exercises banks do not take sufficient account of contagion between institutions. While agreeing that stress testing has its limitations, Nigel Jenkinson argued that it was still useful in understanding the exposure of banks' balance sheets to shocks. Another participant agreed, saying that it was better to conduct imperfect stress tests than none at all.

On the issue of concentration, although there was a broad consensus that greater concentration in the banking sector does not necessarily inhibit competition, particularly when foreign banks are present, there was a lively debate on what this meant for the ‘four pillars’ policy. One view was that allowing the ACCC to consider mergers would make the policy more accountable, even if merger applications were ultimately knocked back. However, other participants doubted that further mergers would lead to efficiency gains and thought that the four major Australian banks had sufficiently large domestic bases to expand offshore.

This led to a discussion about the possible impact of further mergers on financial stability with some participants wondering whether they would make some institutions ‘too big to fail’, though it was pointed out that the Australian majors may have already achieved this status. More generally, Kevin Davis argued that deposit insurance schemes need not be threatened by large banks if governments stood ready to provide additional funds and there was a good recapitalisation plan in place. There was also a brief exchange about the effect of the Basel II Capital Accord on competition in the banking sector with one participant wondering whether it would undermine the competitiveness of smaller banks. However, Kevin Davis suggested that the Accord would merely formalise the existing advantage of big banks.