One of the consequences of the financial crisis of recent years has been a re-examination of the role of central banks in financial stability policy, and the associated notion of ‘macroprudential supervision’. I've been asked to talk today about the interaction between micro- and macroprudential supervision, and how we reflect that in the design of our regulatory institutions.

The first thing to say is that the distinction is in my view somewhat artificial. In concept, the micro/macro distinction corresponds roughly to the difference between idiosyncratic risk and market risk in finance theory. One refers to risk specific to a firm and the other to risks affecting all firms. Microprudential supervision might be thought of as the management of the first type of risk, and macro the second.

But while that distinction is conceptually clear, I think it would be overly simplistic to try to build a supervisory structure around it in practice. Certainly in Australia we don't try to do so. External observers sometimes use the shorthand that the RBA is the macroprudential agency and APRA the micro, but we wouldn't ourselves see it as being so clear-cut. APRA, as the bank supervisory agency, has a mandate to use its powers to promote financial stability for the system, not just to focus on the idiosyncratic risks faced by individual banks. And conversely the RBA as the central bank can't avoid taking some interest in individual institutions even though its financial stability mandate relates to the system as a whole.

The more meaningful distinction, I think, is not between the objectives of the key regulatory agencies (micro versus macro objectives) but between the powers available to them. APRA is the regulator of institutions – it supervises the banks, sets prudential standards and holds a wide range of directive and resolution powers. The RBA is the liquidity provider to the financial system, it has regulatory powers of its own in respect of clearing and settlement facilities and the payments system, and it incorporates financial stability assessments in its monetary policy decision process. Both agencies have a mandate for macro financial stability, though obviously APRA has the more specific mandate for risk management at the level of the individual institution.

With that background, I want to ask two questions about the design and conduct of macroprudential policy arrangements in general.

1. Is there a single right way for these powers to be allocated between the central bank and other regulatory agencies? And
2. Is there a single best practice framework, in terms of instruments and objectives, for these powers to be used?

I want to suggest that the answer to both of these questions is No. I'll then suggest a couple of important positive principles we need to stay focused on, regardless of the specifics of our national frameworks.

To start with the institutional question: is there a ‘right’ way to allocate prudential responsibilities between the central bank and other agencies?

Certainly a number of jurisdictions have come to the conclusion that they needed to bring more of the supervisory powers under the central bank. This has been perhaps clearest in the UK, where the supervisory functions of the FSA are to be brought fully within the Bank of England. A similar but less comprehensive shift in the same direction has been an important part of the regulatory response in the US and the euro area.

In some ways these developments seem like a natural response to the crisis. Cooperation between the holders of traditional central banking powers and those holding institutional regulatory powers is very important. Where that cooperation is considered to have been inadequate, it is natural that institutional reforms would be directed towards fixing the shortcomings. Bringing supervisory powers into the central bank is one way of doing that.

But it's not the only way these things can be arranged, and I can see both advantages and disadvantages to the more centralised approach. Clearly, the centralised approach maximises the scope for formalising the coordination between the two sets of functions. But a large institution with both central banking and supervisory powers may be hard to manage, and it may still be susceptible to a ‘two culture’ problem. I think this was one reason for the earlier trend towards separation of these functions a decade or more ago. My general observation is that we don't yet know how these more formalised arrangements will perform. The centralised model has to be made to work and to build up its own culture and track record of success.

That will take time, and it's far too early to say that any of the post-crisis arrangements represents a new norm of best practice. That is especially so when we consider the length of the typical financial cycle, where episodes of serious stress might be separated by intervals of a decade or more. It will be a long time before these new arrangements are tested by the next cycle.

I would also stress that the centralised model is not the only way to get effective coordination. In Australia we retain a model of separate policy agencies, with a coordinating structure to ensure they work together effectively. Our four main regulatory agencies (the RBA, APRA, ASIC and the Treasury) together comprise the Council of Financial Regulators, under the chairmanship of the RBA Governor. The Council has no statutory powers of its own, but works to coordinate the activities of the agencies in exercising their own powers. Where existing powers to act are inadequate, the Council will typically make recommendations to the government for legislative action to remedy that. In Australia we've found that these cooperative arrangements worked well during the crisis period.

People sometimes ask us how we've achieved that without a formalisation of cross-agency relations, or giving one agency primacy over the others. How has the informal coordination been possible where other countries have found it difficult to make that kind of arrangement work? Are we just nicer people? I'd like to say that the answer is Yes, but I think the less frivolous answer is that institutional history matters. We're fortunate to have built up a culture across the regulatory agencies where we regard cooperation with the other agencies as an important part of our job, and there is a strong expectation from the public and the government that we will continue to do so. That being the case, we're not inclined to disrupt an arrangement that's working. But clearly every country needs to tailor its institutional arrangements to its own circumstances and history.

My second question was about the framework of instruments and objectives for macroprudential policies. Here again I think we need to be suitably modest as to what we can say constitutes best practice.

I've spent much of my career working on monetary policy, and it strikes me that there are certain parallels between the current desire for a more rigorous framework for prudential policy and the search for rigor in monetary policy that occurred after the breakdown of old certainties in the 1980s and 1990s. The result of that search on the monetary policy front was a broadly agreed framework with a single instrument (the short-term interest rate) and a well-defined objective (inflation control), linked by a forecasting framework and an understanding of the feedback between instrument and objective. This broad schema has been thoroughly researched, and lies behind much of the thinking in actual monetary policy decision processes. There is also a well understood case for independence of the monetary policy process, and for transparency and clarity of objectives which has been widely built into institutional structures around the world.

It strikes me that there is nothing commensurate with this schema in the field of financial stability policy. Instead of a single instrument, we have a multiplicity of instruments (e.g. capital rules, some of which may be time varying, risk weights, LVRs, liquidity rules, infrastructure requirements, supervisory intensity) whose effects are difficult to quantify. And the objective of financial stability has no simple metric. It can't be summarised in a continuously observed number like the inflation rate, though obviously we know financial instability when we see it. Further, as I have argued, there is no single best practice model for the design and allocation of decision-making powers.

The obvious conclusion is that the pursuit of financial stability is a much more subtle and complex undertaking than the traditional monetary policy function of central banks. It necessarily involves not just quantitative frameworks, but detailed surveillance of risks guided by the perspective of historical experience. I expect these are things that central banks will be devoting increasing attention to in the years ahead.

I want to finish by emphasising two elements that must be at the core of effective macroprudential policies, regardless of the institutional arrangements in a given country.

The first is effective cooperation between those involved in traditional central bank activities and bank supervisors. Combining these functions in an expanded central bank is not in itself a guarantee of effective coordination, though it may obviously help. Key aspects include an effective flow of information across staff in the market operations and macroeconomic departments of a central bank and those working in the areas of financial stability and bank supervision. Regular meetings among these groups to focus on risks and vulnerabilities and to highlight warning signs can be very valuable. A culture of coordination among these areas is very important in a crisis because, in many instances, a stress situation is first evident in liquidity strains visible to the central bank, and the first responses may be calls on central bank liquidity.

The second element is effective bank supervision. By this I mean not just the standards and regulations but the capacity of supervisors to understand the risks that are being taken, to query them, and to ensure that they are being appropriately managed. This is a point that I know my APRA colleagues have been emphasising. A great deal of effort in recent years has gone into upgrading the prudential requirements on banks through revisions to the Basel standards and other measures. As important as these things are, their effects around the world will be only as good as the quality of implementation and the quality of supervision that builds on them.