RBA Annual Conference – 1996 Discussion

1. Murray Sherwin

Richard Dale has provided us with an interesting survey of current regulatory and supervisory challenges. His focus is on three developments that are influencing the shape of financial markets – namely globalisation, functional integration, and innovation, especially in the form of derivative products.

Underlying the discussion of these three developments is the theme that, in essence, the world is becoming a more dangerous place for financial intermediaries, and therefore, for the supervisors and regulators of those intermediaries. As a consequence some regulatory response is required. Those developments deserve examination.


Globalisation is assumed to increase risks for banks because ‘it offers fresh channels for the transmission of financial shocks’. Of course, the other side of the argument, as Dale allows, is that globalisation also facilitates the wider diversification of economic risk. Why is globalisation a concern? A few possibilities occur:

  • financial intermediaries gain opportunities to expose themselves to risk in jurisdictions which they do not understand. That is no more than an argument that any new activity is likely to be risky as the new entrant works its way up the learning curve. However, it certainly isn't obvious, at least to me, that there is inherent or undue difficulty in managing that risk if the management has a mind to try;
  • national settlement systems are not integrated, leading to increased Herstatt risk as the volume of cross-border transactions grows. This is not a proposition that I would take issue with at all – it simply highlights the need to make progress in dealing with Herstatt risk generally; and
  • if the soundness of financial institutions is indeed dependent on supervision, then there is a greater chance that some risks will fall through the cracks when business spreads across different jurisdictions. That is also true, but is also increasingly being dealt with by national supervisory agencies.

Functional Integration

As with globalisation, those who worry about functional integration and its implications for financial risk would seem to be facing an increasingly ulcer-prone future. The blurring of boundaries between institutional groups within the financial sector is now well underway and is not about to reverse. It is not just that banks now sell insurance, or that insurance companies sell loans, or that they all deal in securities; more fundamentally, the financial products themselves are blurring at the edges. Banks are writing multi-option facilities (where the client can switch between loan and underwriting facilities) and hedging products are looking more like insurance. In essence, anything that involves cashflows can be present-valued, and risks that can be priced or hedged can be re-engineered and repackaged to meet the particular and diverse needs of some client or other. Given financial products which involve an amalgam of lending, underwriting, and insurance all wrapped into one, the traditional ‘functional’ basis of regulation and supervision becomes increasingly difficult to sustain.

If functional integration is inevitable, is it inevitably dangerous or risky? The argument that functional integration is dangerous seems to rest largely on the proposition that integration allows banks to exploit their cheaper funding, while taking on riskier activities such as securities dealing and derivatives trading. This, in turn, seems to hinge largely on the existence of deposit insurance or other similar safety nets for banks or their clients (this being the route which allows banks to lower their cost of funds relative to non-bank financial intermediaries).

Having banks involved in derivatives or securities trading is not, of itself, anything special. Banks have always invested in risky assets – indeed understanding, accepting, and managing risk is a key component of what banks do to add value to the community. A bank with exposure to market risk as well as credit risk is not inherently riskier than one with just exposure to credit risk. The real problem – as Dale notes – is one of the interaction of balance sheet risk and creditor protection. The problems of moral hazard and associated risk/return dislocations which emerge when deposit insurance or other protective devices are available to banks is well understood. The fact that such problems arise does not constitute a case for strenuously resisting functional integration. However, it may be a case for having a fundamental look at the nature of protection afforded to bank creditors.

Financial Innovation

The sources of increased risk associated with derivatives identified in Richard Dale's paper relate to two principal points:

  • balance sheets have become more opaque because derivatives are so complex and because derivatives allow banks to alter their risk positions rapidly, thereby rendering reported risk positions at best obsolete and irrelevant or, at worst, misleading; and
  • operational risk is increased because dealers are able to enter into very large unauthorised transactions.

On the first point, derivatives, in some senses, can be less opaque than traditional credit exposures. It is difficult to value many loans at any point, and to the extent that credit exposures involve embodied options or agency problems, predicting how their value will evolve over time is no simple exercise.

By contrast, it is possible to reduce the main risk dimensions of derivative instruments to their physical equivalent, and to aggregate the risk of a portfolio to a single value-at-risk (VAR) measure. While there is certainly some way to go before we are all talking the same language with respect to VAR measures, there is certainly the potential for derivative exposures to be more transparent than credit exposures.

On the question of how quickly risk can be shifted via derivatives, it does seem that disclosure can still be an effective means of constraining imprudent behaviours. A requirement to disclose peak intra-period exposures, as well as the more traditional end-of-period exposures, is a very powerful sanction on imprudent behaviour. Unless a bank believes that there is a good probability that it will not be around for the next period's report (in which case it does have an incentive to go for broke) the bank will find peak exposure reporting requirements a very real constraint on risk-taking.

There can be little dispute that the growth of derivatives trading has materially increased operational risk. A single employee can encounter both the incentives, and the means, to put the bank (or, indeed, any other entity) at risk. What Richard Dale is silent on is the relative strength and comparative advantage of private incentives in managing this particular risk, as opposed to public supervision approaches.

The New Zealand perspective is that the management and boards of financial institutions are best placed to combat operational risk. It certainly isn't obvious that extending international co-operation in supervision can do a lot to help. In any event, institutional failure arising from operational risk arguably poses less of a systemic threat than, say, default through credit risk. Failure from operational risk is essentially random, and there is little reason to suspect that the failure of one financial institution through operational risk says much about the solvency of others. By contrast, credit risk is much more likely to be correlated across financial institutions, and hence, give rise to more serious contagion risks.

Richard Dale usefully reviews the Barings failure and concludes that it illustrates important regulatory failings. The alternative view is that, from a systemic perspective, it was a damn fine little failure – the costs fell about where they should have, on management and shareholders; depositors lost nothing; the UK Government contributed nothing; markets continued relatively unscathed; a few smaller banks with large security operations had to pull their horns in a little bit, but there were no contagion effects. On the positive side, boards and management everywhere were given a graphic lesson in the importance of understanding operational risk, and for good measure, a demonstrably poor management team exited the market.

Richard Dale's conclusions, at several points, come back to suggestions that regulatory responses are called for to deal with the evolving risks identified. A common theme is the importance of enhanced co-operation between supervisors – in different countries, across different institutional groups, across different functional activities. The tone of my reactions to these various evolutionary forces in financial markets is somewhat different. To my mind the necessary agenda for regulators and supervisors includes the following:

  • to recognise that the developments to which Richard Dale refers – globalisation, functional integration and innovation – are not going to go away, and certainly will not be regulated out of existence;
  • to recognise that there are real welfare costs to regulatory responses that prevent the public from having access to innovations which are themselves being driven by fundamental changes in technology, consumer choice, or competitive advantage;
  • to be very modest about what supervision can achieve;
  • to begin working very hard on encouraging the public (and even more, the politicians) to be correspondingly modest in their expectations of financial sector regulation and supervision. The persistent political reaction to bank failures which sees more and more supervisory effort being thrown at banks is surely counterproductive;
  • to reorient the work of regulators and supervisors. The task, surely, is not to commit to finding ways to prevent bank failures. Rather, the key task is to put ourselves into a position where it is easier to allow banks, even big ones, to fail;
  • to clarify our thinking on what is really important in regulation and supervision and why. Only then will it be possible to concentrate regulatory and supervisory resources on the smallest possible core of the financial system. It seems clear that the key objective is systemic stability. Current developments world-wide in the implementation of RTGS payment systems are a major step forward in dealing with systemic risk. The next substantive challenge is to deal with Herstatt risk. As a goal for supervision, depositor protection must surely have a limited life expectancy – except perhaps in the ‘narrow bank’ context. Depositor protection complicates too many boundary issues in a world of growing functional integration, creates competitive neutrality problems between institutional groups, and is the source of much of the moral hazard problem that we grapple with. Moreover, the public seem increasingly prepared to go without depositor protection, as illustrated by the rush into mutual funds and similar savings vehicles in many countries; and
  • finally, when thinking about supervision and regulation, we need to focus very closely on how best to align private incentives with public interests. Public disclosure of financial risk material is a very powerful force in that regard and deserves a prominent position in the supervisors' armoury.

2. General Discussion

The discussion revolved around three main issues:

  • the goals of financial regulation;
  • firewalls; and
  • ‘are banks special?’.

Participants discussed a number of possible goals of financial regulatory policy. It was commented that the goals were sometimes unclear, and that existing regulatory policies were being asked to do too much. What was needed was to identify the core rationale for regulating particular institutions or activities.

Two main views were expressed. One view was that the main rationale for financial regulation arose from systemic risk and from the potential for liquidity crises. Banks and other deposit-takers were inherently vulnerable to liquidity problems because their main assets were non-marketable loans which were worth considerably less in liquidation than on a going-concern basis. This meant that liquidity problems could lead to cases of insolvency even in sound banks. The usual regulatory safeguards included capital standards, supervision and central bank liquidity support.

There was some debate as to whether there was an analogous liquidity risk associated with securities markets: that is, whether financial markets could become illiquid to an extent that would generate solvency problems in institutions. Some participants thought that this could not occur, since markets were always liquid if sellers were willing to accept a low enough price. Others thought there had been examples of market illiquidity which in certain circumstances could give rise to systemic concerns. The role of policy was to ensure that markets had adequate liquidity and to ensure that institutions were adequately capitalised to have a buffer against risk.

The second main view of regulatory policy was that the essential rationale came from the political imperative to protect depositors. Policy could not take a disinterested view of any loss of depositors' money, and there was a strong public demand to have some core of safe assets that could be held by risk-averse savers.

It was pointed out that these two views had overlapped to a considerable extent in traditional financial systems, where banks were clearly distinct from other financial institutions. Policies aimed at institutional solvency of banks had served the dual purpose of protecting depositors and promoting systemic stability. The situation became more complicated once banks were significantly engaging in activities outside of traditional banking business. If the scope of regulation and official support were extended to the whole of banks' expanded operations, it would undermine regulatory equality between banks and other institutions in the banks' new areas of activity.

The situation was further complicated by the unbundling of banks' core functions, and the increasing scope of banks and other financial institutions to specialise rather than offering a comprehensive range of services. These trends meant that policymakers would have to decide what combination of the traditional core activities was the real focus of regulatory concern. More generally, they would have to decide whether banks were to be regulated because they were called banks, or because of the particular activities they were engaged in. If the latter was the case, then those activities should be regulated equally for all institutions engaged in them.

A second major issue was the question of ‘firewalls’. Regulatory systems generally allowed banks to engage in a wide range of financial services through subsidiaries, but banks were not permitted to underwrite the solvency or performance of their subsidiaries with the bank's capital. In this way banks were intended to be protected from the effects of the failure of a subsidiary. Participants debated whether this kind of firewall could really be effective. Some participants thought that markets recognised the effectiveness of firewalls in some parts of the financial sector, by giving different credit ratings to different members of a financial group. Others argued that firewalls were generally ineffective. Banks could not allow their subsidiaries to fail without damage to their own reputations, and consumers generally did not recognise the distinction between banks and their subsidiaries. Indeed, the marketing advantage conferred by a bank's brand name relied on some public perception that banks stood behind their subsidiaries' products. This in turn was argued to give bank subsidiaries an unfair advantage in the markets in which they operate, because of the parent banks' perceived access to public support. This pointed to a need to strengthen the firewalls, for example though improved disclosure of the fact that bank capital could not be used to support a subsidiary.

The more general question of what makes banks special was also discussed. Participants discussed the traditional view that banks' special status came from their unique combination of activities: their role in the payments system, illiquid loan portfolios and deposit liabilities. A number of participants commented that this special position was being eroded. In particular the process of securitisation meant that banks could package and sell off their assets to an increasing extent, so their role as holders of non-marketable loans was being reduced. Similarly, the process of unbundling was changing the basic character of many financial enterprises. On the other hand it was pointed out that these processes take time and that it is not yet known how far they will go. Another view was that banks had originally been seen as special largely because of their role as retail deposit-takers. A corollary of this view was that the regulatory focus should not be on banks per se but on retail deposit-taking activities generally.