Speech Developments in Financial Regulation
Glenn Stevens
Governor
Address to the Australian Business Economists Annual Dinner
Sydney –
- Audio 32.2MB
- PDF version 112KB
The financial crisis that engulfed global capital markets and brought a number of important international banks close to the brink last year has been followed by a good deal of soul searching among the regulatory community. In several countries, though not in Australia, regulatory structures and/or practices have been seen as inadequate. Work is proceeding to try to establish better arrangements so as to prevent the next crisis, or, more realistically, at least make it less costly – all the while seeking to avoid doing things that make it harder to recover from this one.
I propose to offer today some information and some observations about these developments and the associated issues. I won't go into the causes of the crisis per se; these have been covered at length before. The material offered is set in a global context, rather than an Australia-specific one. Just to be clear, if in the following paragraphs I am saying something about Australia, I will make that explicit.
Lessons from the crisis
What conclusions have governments and regulators around the world drawn from this episode?
There are many. But the most important ones can be organised under five relevant headings.
First, capital: there was not enough. In the case of global banks' trading books, a lot of risk accumulated and was not well measured. Capital held against complex structured products in particular was seriously insufficient. A good deal of risk was also supposed to be ‘off balance sheet’, but returned very quickly to major institutions once liquidity dried up. Moreover, some instruments were considered to be ‘capital’ but could not really absorb losses, at least while a bank remained operating.
Second, liquidity: not enough attention had been paid to the risk that, in the event of some market shock, funding liquidity could become much more difficult. Comfortable assumptions that markets for some instruments would remain liquid proved to be unfounded.
Third, the so-called ‘shadow banking system’: there were systemically important activities going on outside the ‘regulatory perimeter’. This included the activities of investment banks, hedge funds, finance companies, money market mutual funds and institutions that often had close ties to banks, such as special purpose vehicles, structured investment vehicles (SIVs) and conduits. These entities were typically less closely supervised or unregulated, but in some cases their risk-taking behaviour and subsequent travails had systemically significant impacts on the core financial system.
Fourth, cross-border arrangements: globally active banks and other entities operated apparently seamlessly across national borders and legal jurisdictions. But the structures to allow that were actually quite complex, and legal, supervisory and crisis-management arrangements remained nationally based. So when it came time to manage the process of deleveraging and winding up of some institutions, the degree of complexity was increased by the cross-border nature of the issues.
Finally, pro-cyclicality: the episode demonstrated – again – that the financial system imparts its own dynamic that reinforces the natural cyclical tendency in an economy. In good times, lenders and investors tend to be confident and act with less caution. Standards decline and banks come under pressure either to use ‘surplus’ capital or return it to shareholders. Backward-looking risk metrics present risk as low just when it is reaching dangerous levels and very high when everyone has already become much more risk averse. This all serves to fuel the boom and bust.
Many commentators have argued that accounting standards contributed to pro-cyclicality. Fair-value accounting and buoyant markets make for strong valuations that can boost recorded profit, but problems emerge when banks have to mark-to-market securities whose markets have effectively ceased to function. The incurred-loss basis for provisions – where an event has to occur before a provision can be made – promotes transparency in one sense. But it arguably inhibits the build-up of buffers in the good times to cushion against future losses, and prompts more provisioning during the turmoil, further harming profitability and confidence.
Other incentives have also been seen as adding to cyclical behaviour. Remuneration packages for some financial institutions' executives and employees appear to have been structured in a way that may have encouraged traders and managers to take excessive risks in activities that appeared profitable in the short term but led to large losses later on.
People have also pointed to the role of earlier regulatory changes, credit ratings, the complexity of instruments and weaknesses in market infrastructure, not to mention the long period of low global interest rates, the ‘global imbalances’ and so on as all playing a role. I don't want to underplay those factors, but the five above are the big themes on which I want to focus today.
What is being done?
Much has already been done, or is under way, to respond to these weaknesses by the various standard-setters, the Financial Stability Board and within the G-20 process. Many of the responses can be organised under the same five headings.
First among them is capital regulation. Now it is worth pointing out, before going on, that it is something of a stretch to suggest, as some commentators have, that the so-called Basel II framework was to blame for the crisis. In fact the build-up to the crisis occurred under the old Basel I capital rules: most global banks did not even implement Basel II until after the crisis had begun. US banks are still not using it. Basel II is not perfect but it addresses some of the shortcomings of Basel I that were identified by the financial crisis. It allows, for example, greater differentiation between different types of risk, introduces capital charges for off-balance sheet exposures to SIVs and conduits, and creates more neutral incentives between holding assets on balance sheet and securitising them. Had Basel II been in place, it might not have prevented the crisis but it would probably have helped matters.
Nonetheless, Basel II can be improved in the light of experience during the crisis. The Basel Committee on Bank Supervision is in the process of implementing numerous changes, which essentially require more capital and higher-quality capital. It has already finalised changes to risk weights on certain exposures related to securitisation, and issued new supervisory guidance on compensation, governance, risk management and concentrations of risk.[1] It has changed rules around disclosure and valuation practices, which will come into effect at the end of next year. The Committee is currently rethinking the amount of high-quality capital banks should hold, specifying what instruments should be included in that definition, and developing a non-risk-weighted simple leverage ratio as a supplement to the risk-weighted capital adequacy measures. The Committee intends to finalise this latter set of new capital rules by the end of this year and calibrate them in 2010 with a detailed ‘quantitative impact assessment’ to gauge their effects.
Second, efforts to bolster liquidity management are under way, with the Basel Committee planning to introduce a new global standard for funding liquidity soon. This is likely to require financial institutions to focus on adequate funding liquidity over longer time horizons, as well as resilience to more demanding stress scenarios. In line with this, APRA has recently released for consultation proposals to enhance liquidity risk management by authorised deposit-taking institutions in Australia.
Third, the regulatory perimeter is being extended. Some of the relatively less supervised institutions that were problematic prior to the crisis no longer exist – for example, US investment banks failed, or were converted into, or assumed by, regulated banks. But other institutions whose actions could on occasion be of systemic importance, such as hedge funds, are being subjected to more oversight.
Fourth, attempts are being made to help the cross-border issues with the creation of supervisory ‘colleges’ for large institutions. These are designed to promote better sharing of information across countries. Agencies on the Financial Stability Board are also working on protocols for cross-border crisis management.
Regarding pro-cyclicality, proposals are being developed for the use of capital regulations that would require banks to increase capital in the good times that can then be run down during a crisis. The proposals involve introducing target counter-cyclical capital buffers, above the re-designed minimum capital requirements. The Basel Committee is also working to promote the use of more forward-looking provisioning policies based on expected losses, rather than current arrangements that base provisions on losses already incurred.
Accounting standard-setters are continuing their work on international convergence and clarity in regulation. They have issued guidance emphasising the need for judgment in valuing mark-to-market assets when markets for those assets are inactive, are working to simplify the valuation rules for financial instruments, and are seeking to close loopholes that generated incentives for off-balance sheet activities.
Some observations
All of this is worthwhile work. It forms part of a comprehensive set of responses to the conclusions drawn from the crisis. Implementing it presents a very demanding schedule for regulators. I want to offer a few observations about what we might realistically expect over time. These are not criticisms, but rather nuances that are, in my view, worth noting.
The first is that the right balance needs to be struck between more regulation and more effective enforcement of existing regulation. There is no doubt that regulation can be improved after an event like this. Yet some jurisdictions ended up with very serious problems, while others did not, even though they were all, more or less, operating on the same internationally agreed framework for bank supervision. Why that was so remains a question of interest.
Secondly, on the assumption that most of these regulatory changes go ahead, one effect will presumably be to make the process of financial intermediation more costly. The intention, after all, is that lenders will operate with more capital against the risks they are taking. But capital is not free; shareholders have to be induced to supply it, and it will have to be paid for. High-quality liquid assets typically carry lower yields too, so mandating higher liquidity will have some (modest) cost as well. Admittedly it can be argued that shareholders of financial institutions will have a less risky investment and so should be prepared to accept lower returns. But customers of financial institutions – depositors and borrowers – will also pay via higher spreads between what lenders pay for funds and what they charge for loans. That is, they will pay more ex ante to use a safer financial system, as opposed to taxpayers having to pay large costs ex post to re-capitalise a riskier system that runs into trouble.
Now of course protecting the interests of taxpayers is very important, and there is no doubt that certain types of behaviour need to be backed with much more capital, if not severely curtailed or even stopped altogether. It is appropriate that pricing play a role in achieving that.
We should try to ensure, however, that the cost is no more than necessary. The most egregious behaviour was mainly that of 30 to 40 large, globally active banks. They have imposed very large costs on their own banking systems, economies and taxpayers, and on the global economy. But there are thousands of other banks in the world whose risk appetite did not get out of control, that have remained solvent, and that have not needed public capital injections. So it will be sensible to ensure, as far as we can, that the proposed measures act effectively to constrain the worst excesses of the former without unnecessarily shackling the latter.
I am personally not persuaded of the intellectual basis of the simple overall leverage ratio. It goes against the whole thrust of the idea that capital should be allocated against economic risk – after all, the Basel risk weights are a sophisticated leverage ratio already. I have not seen persuasive evidence that the banks of countries that had a leverage ratio in place have systematically outperformed those that did not.
Nonetheless, it had already been agreed, before Australian officials joined the Basel Committee, that such a device would be introduced as a ‘back-stop’ to prevent extreme leverage in instances where the Basel rules, for some reason, may not. Provided that it is suitably calibrated, the leverage ratio will probably not do any great harm. That is, however, a very important provision. Were it to be calibrated in a way that unnecessarily constrains the common or garden variety commercial bank, it could be unduly costly. So the calibration is important and in this regard it is critical that adequate time be allowed for the completion of the technical work in assessing the quantitative impact of this measure. That will take at least another year.
As far as the proposed counter-cyclical capital buffers are concerned, this is an appealing idea, based on the notion that it is precisely at the moment when capital appears to be abundant, profits high and the economy booming that true risk is approaching its peak. Requiring more capital to be put aside at that time, which can then support balance sheets after the cycle turns, sounds very desirable.
But we should not think it will be easy to achieve. The proposals appear to involve balancing a degree of mechanical linkage to particular variables – such as credit – with an appropriate degree of short-term flexibility. Those familiar with the old debates about rules versus discretion in monetary policy might notice an echo here. Based on experience in monetary policy, I am sceptical about the durability of hard rules, but all too familiar with how difficult it can be to deliver genuinely counter-cyclical policy. There is no reason to think it will really be easier when using prudential tools for ‘macro-prudential’ purposes. That is not to say that we can't devise a framework combining certain rule-like behaviour with a sensible degree of discretion, but it might take a while. In the case of monetary policy, it took a couple of decades or more.[2]
If one is inclined to place a good deal of importance on quality supervisory judgment – as I think we have to – as much may be achieved by adjustments to accounting arrangements for provisions as by a complex set of variable capital ratios. What is needed is to allow banks more easily to make more forward-looking provisions when either they or the supervisor thinks they should. This is an important area of work for regulators.
One should also be realistic that while using balance sheet regulation as a macro-prudential tool may have attractions, it is no panacea. Of course there may be occasions when the setting of monetary policy is about right for most of the economy but there is a desire to calm down some over-exuberant borrowing behaviour in a particular sector. In such cases, some kind of temporary regulatory measure may well be useful. But as those with any recollection of Australian experience of the 1960s and 1970s will know, if the fundamental problem is actually that financial conditions are just too easy – that is, interest rates are too low – balance sheet regulation won't ultimately constrain credit growth. Over time, private markets will find a way of doing the business outside of the regulated sector. Then the authorities face the question of whether or not to expand the scope of regulation to more sectors – just as we did in the 1970s. A possible outcome is that, the harder we regulate a set of institutions as a result of the last crisis, the more likely it becomes that the next crisis occurs in the hitherto unregulated part, perhaps even among institutions that do not yet exist. If the conditions are such that people want to take risk and gear up, they will find a way.
Of course, that is acceptable provided the relevant private parties can be allowed to fail without bringing down the core part of the system. Caveat emptor can apply outside of the regulated net if there are few spillovers. But if risk-taking activity goes on long enough, then sufficient leverage may well accumulate somewhere to make the ensuing deleveraging generally disruptive, placing policy-makers once again in a very awkward situation.
Too big to fail
And that brings us to the most difficult of all the issues, namely that of the too-big-to-fail institution. Here the term ‘big’ might mean a large balance sheet, or refer to interconnectedness or complexity. Or all three of the above. In some countries, the debate on this issue is now quite active. One potential response is a tax on size: much higher capital requirements for ‘systemic’ institutions so as to lower greatly the probability of a failure of a really large firm, either by making large firms much less risky, or giving them an incentive to no longer be large. Another approach would simply be much more intensive – and intrusive – supervision of such entities. Either of these measures could be complemented by the requirement that a large firm compile a ‘living will’, in which it writes its own break-up/wind-down plan ahead of time – in the process, hopefully, highlighting those bits of complexity that ought to be removed while it is still alive. Before regulators even got to any of those possibilities, they would have to grapple with the practical difficulties of setting the threshold as to what constitutes ‘systemic’. Expect furious lobbying by the finance sector on that.
Nor is this issue just domestic in nature. It goes to the heart of what it means to have a globalised financial system. In the absence of clearly articulated rules about burden sharing, a potential failure of one of these institutions is further complicated because the frameworks for global governance and crisis resolution have not kept up with the process of globalisation itself. It is very hard for them to do so. Even in a region such as Europe where there has been six decades of continuous effort in building collective structures, the resolution of problems at entities like Fortis has, by all accounts, been very difficult.
One response to that would be to unwind the globalisation of the financial institutions and go back to having local banks just doing local business. That seems absurdly costly, though – in general, capital flows have been a tremendous force for higher living standards over the long run. It would surely be a retrograde step to shut them off. A less radical response would be subsidiarisation – where foreign banks have a presence in the form of locally capitalised and governed structures, in which local authorities could intervene in the event of a shock in another country affecting the viability of the parent. That still entails some costs in terms of efficiency, albeit ones that countries might now be prepared to tolerate. To succeed in this approach, a country would need to have the capability and resources to ensure the viability of a local subsidiary of a failed major global institution, taking control of it if necessary. This would be at a time of tremendous damage to the relevant global brand. For many small countries this might be a big ask.
In the crisis itself, the too-big-to-fail issue presented simply as an imperative for a number of governments to prevent failures. But as the crisis recedes, and the global financial system is gradually nursed back to health, it is this issue that is going to leave the biggest lingering challenge. The Financial Stability Board will be directing particular attention to it over the coming year. It is not likely to be amenable to simple solutions, or easy ones. In the meantime, enormous moral hazard, perhaps greater than ever before, exists in the global financial system as a result of the actions – albeit essential ones in the circumstances – of 2008.
Conclusion
A year ago, I wished this audience a much less interesting 2009. That wish was partly fulfilled in that 2009 was less ‘interesting’ than 2008, though still not quite boring enough in my view.
As 2009 draws to a close, things in the global financial system look much less worrying than they did a year ago. With the sense of immediate crisis much reduced, regulators can devote more focus to the job of designing and implementing changes to regulatory frameworks – work that is better done outside a period of crisis anyway.
Realistically, the task is to reconfigure regulatory frameworks to lower the probability, and the cost, of future crises while assisting recovery from the recent one. That is every bit as difficult a challenge as getting through the immediate crisis itself. It will require very careful judgment to strike the right balance between costs and benefits of revised regulatory structures and practices, and due regard to the possibility of unintended consequences. It will also take a great deal of determination on the part of regulators to enforce arrangements adequately in future booms. And there is little doubt such booms will occur because, ultimately, the cycle of greed and fear itself cannot be regulated away. To assume that unrealistic optimism will not again, at some point, overwhelm the more sober instincts of investors, bankers, commentators and others would be a triumph of hope over experience.
But it can't be beyond us to achieve some worthwhile reforms in this area and 2010 is a year in which we can hope to make some progress.
I wish all of you a Merry Christmas and a prosperous and stable new year.
Endnotes
Specifically these relate to: securitisation exposures in the trading book; sponsorship of off-balance sheet vehicles; re-securitisation exposures; and pipeline and warehousing risks with regard to securitisation exposures. [1]
We ended up with a form of constrained discretion: a reasonably clear medium-term objective combined with short-term operational flexibility, and independence of the decision-maker from the day-to-day political process (with accountability to legislatures). The counterpart framework for counter-cyclical capital management will, if it is to be successful, require similar elements, including the independence of the relevant decision-maker – from both the political process and the markets and institutions affected by the decisions. [2]