What makes good prudential supervision?

John Laker, Chairman - American Chamber of Commerce Business Briefing, Melbourne

I am pleased to have this opportunity today to address the American Chamber of Commerce in Australia at one of its Business Briefings. This is the first time, I believe, that APRA has met with the Chamber and its guests and your invitation to a financial regulator has proven very timely, in view of the sweeping financial reforms that were signed into law in the United States only last month.

Three years after the global financial crisis erupted, the view from our two countries could not be more different. In the United States, the epicentre of the crisis, economic recovery may be losing steam, unemployment is still at double-digit rates, budgetary pressures remain intense and, with bank balance sheets still under repair, the banking system is only slowly regaining confidence to resume lending. Australia was at the periphery of the crisis but was not immune to its shock waves, hitting initially through the dislocation of global funding markets, then the global economic contraction and, more recently, through the market fall-out from European sovereign debt concerns. Nonetheless, the Australian economy is looking to above-average growth in the next two years, the unemployment rate starts with a five, the Government's budgetary position is the envy of other advanced economies and our financial system is in sound shape.

Post mortems on the global financial crisis point to much the same set of causal factors — the long period of low interest rates and abundant liquidity that drove the "search for yield" and led to excessive leverage, poor underwriting standards in the US subprime mortgage market, shortcomings in governance and risk management practices, distorted remuneration and other incentives, and regulatory frameworks that did not cover risks adequately and had porous boundaries.

In the immediate aftermath of the crisis, a desire for decisive and visible action led global policymakers to focus on regulatory frameworks. As I speak, proposals to strengthen global capital and liquidity regulations are being finalised. More recently, however, there has been a greater recognition that good financial outcomes require not just effective rules and ratios but also quality prudential supervision. By prudential supervision, I mean the direct oversight of financial institutions to ensure that they are not just complying with the regulations in a legalistic sense but are operating soundly and prudently in the spirit of the regulatory framework. As many now acknowledge, some advanced countries with similar financial systems, operating more or less under the same set of global regulations, were less affected than others in the crisis. Australia is one such country. While there may be a number of reasons for this, the International Monetary Fund (IMF) has recently offered one explanation — that prudential supervision was more effective in some countries than in others.

In other words, a regulatory framework cannot be a "set and forget" list of rules but one that needs to be implemented and actively enforced — ideally by a skilled, empowered and well-resourced prudential regulator. This view is explicit in the communiqué from the G20 Leaders Toronto Summit in June, which identified effective supervision of the banking system as the "second pillar" of the global reform agenda, the first pillar being a strong regulatory framework.

Against this background, I want to share with you today some insights into APRA's approach to prudential supervision, before and during the global financial crisis.

That will provide a lead into the question I have posed — "What makes "good" prudential supervision?" Unfortunately, that question cannot be met with a simple, checklist answer. Over APRA's short history, we have learnt what the Australian community expects from its prudential regulator; we have known highs and lows. In 2005/06, our performance was subject to thorough examination by a „jury of our peers‟ through the Financial Sector Assessment Program undertaken by the IMF and World Bank. What helps me today is that the IMF has recently published a paper, The Making of Good Supervision: Learning to Say "No"[1], which emphasises the importance of an active and hands-on approach to prudential supervision and identifies what the IMF believes are the key elements of good supervision. These key elements provide helpful benchmarks for comments on APRA's approach. To illustrate supervision in action, I would like to conclude with some remarks about the issues currently on APRA's supervisory radar.

APRA’s approach to prudential supervision

A sporting analogy might be a helpful starting point to explain APRA's somewhat arcane role. Australians do not share the American affinity for baseball nor do most Americans understand the mysteries of cricket. Our preferred football codes also differ markedly. So, for simplicity, let me refer to the world game our two countries both call "soccer".

APRA's regulatory role in setting the rules of the game can be likened to that of FIFA. We determine the governance, capital, liquidity and other minimum prudential requirements that regulated institutions must meet if they want to take the field. This can be thought of as determining the boundaries of play, the calibre of a club and the team it wishes to field, the depth of the bench, the offside rules and the offences for which a player may receive a yellow or red card. There are 17 laws in the official Laws of the Game for soccer; prudential regulators have a rather larger number! Having set the rules, our supervisory role goes beyond FIFA to that of a referee. We licence financial sector participants — we decide whether each team has the fitness, skills and experience to compete — and then we monitor teams continuously to ensure they are meeting prudential requirements and managing their affairs with appropriate prudence. In other words, we ensure that regulated institutions play within the letter and spirit of the rules and remain match fit.

The 2010 World Cup reminded us, yet again, that there can be good and bad referees. But what makes a good financial referee? Global policymakers have been grappling with this question. To what extent did the rule-makers set rules that gave inadequate attention to safety and allowed players to expose themselves (and the spectators) to the risk of serious injury? To what extent did the referees exercise inadequate control over the players, failing to curb behaviour that was dangerous to all concerned? Did some referees lack powers to control the game or did they just fail to use the whistle?

Unfortunately, I can only take the sporting analogy so far in explaining what we at APRA do. Supervising financial institutions is a complex task, with significant consequences when it is not done well. Let me explain in a little more detail what APRA supervisors do on a day-to-day basis.

APRA's approach to supervision has a clear goal. It is to ensure that our supervisory judgements about a regulated institution are accurate, timely and robust, and that the supervisory attention afforded to each institution is appropriate.

APRA needs to understand the strengths, weaknesses and major risks facing our regulated institutions. To achieve this, each regulated institution has assigned to it a Responsible Supervisor — for the largest institutions, this is in practice a team of supervisors — who is tasked with knowing all there is to know about that institution, identifying areas of potential risk and requiring appropriate rectification. Supervisors will typically build-up this knowledge by drawing on the expertise of APRA's large team of specialists (risk management experts, actuaries, accountants, lawyers, etc), but they are ultimately accountable for the assessment undertaken and the decisions arising from it.

The centrepiece of our supervisory cycle is the forward-looking supervision action plan that our supervisors develop and tailor for their respective institutions. Beyond some baseline requirements we have for all institutions — designed to ensure that every institution receives a minimum level of supervisory oversight — each plan sets out the key risks we see facing the institution and in response, over the next one to three years, the desired supervisory outcomes and the activities necessary to achieve them. Limited resourcing will necessarily restrict the supervisory activities that can be completed and, consistent with a risk-based approach, supervisors must prioritise activities to ensure the most important issues are dealt with.

Having determined the plan, our supervisors then move into action. This can range from off-site analysis of financial and other statistical data provided to APRA on a regular basis, to more detailed on-site reviews at which risk management and internal controls can be tested, to high-level strategic discussions with boards and senior management. The nature of our interactions will be determined by the supervisor‟s assessment of the risk profile of the institution. Where we sense all is not as it should be, the frequency and intensity of our reviews will be lifted.

APRA has had a formal risk assessment model for some years now, known as the Probability and Impact Ratings System (PAIRS), which ensures that the risk judgments that underpin supervisory action plans are rigorous and consistent. Every APRA-regulated institution has a PAIRS rating, which is constructed by applying supervisory judgements to a common set of risk factors. This allows APRA to utilise a unified framework and language for comparing institutions and risks, and helps us to make resourcing decisions.

Prudential supervision during the global financial crisis

APRA's modus operandi did not change significantly during the global financial crisis.

We did not need to make material changes to the regulatory framework. In banking, APRA had taken advantage of the good years before the crisis erupted to strengthen its governance, "fit and proper" and capital requirements; with the implementation of the Basel II Framework from the beginning of 2008, the regulatory framework was very largely complete. Hence, APRA came under no immediate pressure to address any gaps. Coordination arrangements between APRA and the other regulatory agencies in Australia, particularly the Reserve Bank of Australia, also proved their effectiveness during the crisis. As a consequence, there has not been the need for any significant remaking of laws and regulatory arrangements such as that embodied in the Dodd–Frank Wall Street Reform and Consumer Protection Act in the United States. APRA is always vigilant for signs that the regulatory framework needs to be updated to take account of market developments. However, other than our recent proposals for supervising conglomerate groups, we are not proposing any overhaul of the regulatory framework beyond that required to implement global capital and liquidity reforms.

We did not materially change our culture. As I will discuss further below, APRA is a robust and determined prudential regulator and our spine has certainly been stiffened since the early years of this decade. We seek an open, productive and cooperative relationship with regulated institutions built on mutual respect, but we require institutions to constantly demonstrate, and not just assert, that they are financially sound and prudently managed. When we have doubts, we act. This approach served us well leading into the crisis because it gave us a good understanding of each institution‟s strengths and frailties, and we were quickly able to target vulnerabilities.

We did not materially change the sorts of supervisory activities we undertake. The supervision cycle I described earlier is well embedded in APRA. We have reshaped and refined how we employ our various supervisory tools but we have not fundamentally adjusted our toolkit over the past few years. What we did do during the crisis was make more use of some tools, and less of others, consistent with our objective of being risk-based and flexible.

What we did materially change, however, was our tempo. We stepped up the intensity of our supervision markedly during the crisis. To return to the sporting analogy, we became much less inclined to just "let the game flow" and much more focussed on the match fitness of each participant. We sought more frequent and/or more comprehensive information to understand the different impacts of the crisis and to satisfy ourselves that regulated institutions were employing strategies in line with their capabilities. For example, we sought more frequent (sometimes daily) liquidity reporting from our larger banking institutions during the height of the crisis. We engaged more frequently with boards and senior management to ensure they were on top of any difficulties they faced. We established teams of supervisors and specialist risk staff to monitor and coordinate responses to specific risks and provide a cross-industry perspective. We undertook a range of stress-tests across our regulated industries, culminating in a major macroeconomic stress-test in banking, to evaluate whether institutions held adequate capital to cope with "severe but plausible" crisis scenarios.

The making of ‘good’ supervision

Over recent years, APRA has come to be acknowledged as an example of good supervision in practice. The Financial Sector Assessment Program for Australia I referred to earlier found that APRA's prudential framework was principles-based and implementation was of a generally high standard, while APRA's supervisory activities embodied many best international practices and the overall quality of supervision was good. A subsequent review by the IMF in 2009 of the implementation of the Basel II Framework in Australia concluded: APRA operates a sound supervisory system that meets Basel II requirements and builds on the robust regulatory and supervisory process already in place. There is sufficient evidence of its effectiveness in promoting a well-capitalised banking system, both on an ongoing basis and in response to specific events, such as the recent global financial crisis.

Note the use of terms like "good" and "sound" — qualities that are often in the eye of the beholder. To give more meaning to these epithets, the IMF has drawn on its experience of over 150 Financial Sector Assessment Programs to provide guidance on supervision in its recent paper, The Making of Good Supervision: Learning to Say "No".

This paper emphasises that prudential supervisors must have both the ability and the willingness to act to redress instances of excessive risk-taking. The paper also identifies the following key elements of good supervision — that it is "intrusive, sceptical, proactive, comprehensive, adaptive, and conclusive". These words are, of course, an echo of the urgings of the HIH Royal Commission that APRA develop a "sceptical, questioning and, where necessary, aggressive approach" to prudential supervision. We heard that clarion call some years ago and we have worked hard in response. We haven't needed to hear it again!

Let me offer some comments on APRA's approach to supervision against the IMF's key elements.

To the IMF, an intrusive supervisor is not a hands-off or distant observer but rather a "presence that is felt continuously". Regulated institutions certainly feel APRA's presence, particularly because of our strong emphasis on on-site supervision. It seems hard to believe that there has been debate internationally about whether on-site supervision is even necessary, and some supervisors do outsource all on-site reviews to external auditors or consultants. Not APRA. We expect institutions to have appropriately resourced internal and external audit functions and we review audit reports on a routine basis. However, when it comes to judgments about governance arrangements or the calibre of risk management systems, the buck stops with us.

APRA"s on-site visits are not multi-week affairs with dozens of staff but rather short, sharp and targeted discussions. They take place only after supervisors have reviewed a wide range of relevant documentation and internal institution reports. Typically, our findings are sent to the board. Outcomes range from requests for more information, to requests for remediation plans, or requirements to take particular actions to comply with a prudential standard. Recommendations to improve internal risk reporting and the articulation of risk appetite are common findings from our on-site visits. We place a heavy emphasis on the quality of internal reporting that the institution, and importantly its board, is using to monitor its own risks.

We do not take a compliance-based approach and there are few checklists involved in either our on-site or off-site supervision activities. We are able to take this approach because APRA has a core cadre of experienced frontline supervisors and risk specialists who have conducted scores of reviews and know what to look for.

Aside from on-site supervision, our impression is that APRA's prudential framework involves a broader range of notifications and approvals than many other regulators. This provides opportunities for multiple points of "intrusiveness" with regulated institutions. For example, APRA generally reviews most capital instruments issued and institutions need formal approval from APRA to buy-back shares or to call capital instruments. They also need to notify APRA of material outsourcing arrangements, new board directors and changes to internal risk models. This has fostered a culture where institutions routinely consult APRA supervisors and technical experts when developing a new product or considering a complex transaction. The process is very time consuming for us but the close level of liaison pays dividends in improving our understanding of institutions‟ business.

Ahead of the global financial crisis, we did engage in numerous rounds of meetings and queries on appropriate new product approval processes and the correct regulatory capital treatment for a range of more complex financial products. In some cases, supervisors and our technical experts required amendments to transactions in order to meet capital or securitisation requirements. This early involvement by APRA did not remove all temptation to enter into higher-risk transactions but would probably have thrown some "sand in the gears" that may have limited the extent of this dabbling by Australian institutions.

To the IMF, a sceptical but proactive supervisor must question an industry's direction or actions even in good times and not act only after troubles have surfaced. This is a demanding requirement. Identifying emerging risks in advance of the broader industry is particularly difficult, even for the most insightful and forward-looking supervisor. No supervisory agency we are aware of, ourselves included, has claimed that it saw the global financial crisis coming in its full severity, although many of us were warning about the underpricing of risk in those heady days before the crisis.

APRA can take satisfaction, nonetheless, from flashing the orange warning light on credit standards in housing lending when those standards came under pressure during Australia‟s housing market boom in the early years of this decade. APRA's pioneering stress-testing in this area, its surveys on valuation practices and debt serviceability and the on-site reviews of its specialist credit risk teams had the cumulative effect of emphasising to boards and management of banking institutions that housing lending had downside risks.

Comprehensive supervision, according to the IMF, is vigilant about risks emerging at the edge of the regulatory perimeter, including risks in unregulated subsidiaries. As an integrated regulator, APRA can take a broader view of the main activities of our large banking and insurance conglomerates. This has given us a more comprehensive picture of the risks to the overall institution than perhaps some other supervisory agencies can gain. We have not had the situation, for example, of banks owning largely unregulated securities subsidiaries with material exposures outside the bank capital adequacy framework and the visibility of the supervisor.

APRA has had longstanding restrictions on banks moving assets off their balance sheet into special-purpose vehicles and out from under our capital requirements. Further, APRA does not permit banks to move assets back onto their balance sheet — once moved off, they must stay there. During the crisis, APRA received a number of requests for exemptions from this strict separation rule, which we did not grant. As a consequence, while off-balance sheet vehicles are not uncommon in Australia, they are also not a very effective means of hiding problems from us.

Good supervision must be adaptive to meet the challenge of a constantly evolving financial sector and stay in tune with best industry practice. APRA's use of a consistent risk assessment model enables peer comparisons that provide valuable information about outliers in terms of business model and risk management practices. Subject to the usual confidences, APRA will share peer group information with boards and management of regulated institutions to educate them on where they stand relative to their peers. This proved particularly useful during the more acute phases of the global financial crisis when APRA identified a range of short-term liquidity management practices, some much less conservative than others.

The IMF sees supervision as conclusive if outstanding prudential issues are followed through to their satisfactory resolution. This, too, can be a demanding requirement. Taking timely and decisive action when a prudential concern is uncovered may seem straightforward but can be difficult in practice. Institutions may want to challenge APRA on the enforcement of prudential requirements and argue for special treatment. Rules inevitably have grey areas that may invite testing. At times, institutions may not imbue a particular issue with the same significance as the supervisor.

Because APRA is a relatively small and non-hierarchical organisation, supervisors are able to get feedback quickly from risk specialists and from their senior management including, if needed, from the APRA Members; in trickier situations; they do not need to wait for a committee meeting or memo to be signed off through multiple levels. They can get clear and immediate management endorsement to take a strong position on an issue. Further, APRA's Members and senior management can and do take a supervisor‟s concerns directly to an institution‟s board and CEO.

An important aspect of taking matters to their conclusion is the exercise of supervisory discretion. More and more, APRA prudential standards incorporate specific discretions that can be applied by supervisors, based on their judgement about the institution‟s ability to comply with a particular requirement and the possible consequences if it were allowed to comply in an alternative way. Built into the exercise of the more significant discretions is an internal process that requires a consistency check and consultation with our risk specialists. APRA also uses capital requirements where necessary to back up our risk concerns. We will readily increase an institution's minimum capital requirements if we judge that its risk profile has deteriorated. We also use capital as a tool to influence risk-taking behaviour across an industry, for example, when the risk-weight on low-doc housing loans was increased in 2004.

For completeness, I would mention two other factors that have contributed to APRA's effectiveness. One is its clarity of purpose. APRA has a simple and clear mandate — to protect the depositors, policyholders and superannuation fund members of the institutions we regulate. We believe this "single purpose" focus has benefited us significantly, in providing our staff with a clear purpose and mission and in avoiding the need for compromises when multiple policy objectives come into conflict.

Another factor is consistency of implementation. Inconsistency in applying prudential requirements can tempt institutions to push for lenient treatment and, in the extreme, can undermine the credibility of the supervisory process. We take this issue seriously in APRA. In the case of the large complex institutions, for example, our frontline supervisors and specialist risk staff are all in Head Office. This facilitates the sharing of information across institutions that is needed for proactive and consistent supervision since supervisors are able to interact with one another on a daily basis. Compared with overseas peers, APRA maintains relatively small supervisory teams, with senior-level representation on each, and critical issues are immediately escalated. Team members do not become overly specialised or have 'forest vs. trees' problems in seeing the overall risk picture rather than just one small element.

Issues on APRA’s supervisory radar

The IMF's key elements of good supervision, particularly under the headings of "comprehensive" and "adaptive", emphasise the need for supervisors to be vigilant against the build-up of systemic risks within an industry or across industries. APRA has now established a formal structure and dedicated resourcing for this task. The goal is to identify "thematic" or macro-prudential risks and develop an appropriate APRA response that mitigates the risks, prioritises the use of our scarce resources and facilitates consistent treatment across institutions. Embedding the knowledge gained is also critical. The essential features of our approach are risk "registers" for each of our regulated industries, in which risks are graded by their potential severity, and the assignment of risk owners whose task it is to scope the activities required in response, determine resources and timeframes and see the activities through to conclusion.

Our risk registers address risks that are judged to be elevated and industry-wide. "Normal" or idiosyncratic risks facing an institution remain the concern of the Responsible Supervisor. Let me illustrate the sort of issues on the registers.

Though our authorised deposit-taking institutions (ADIs) bore the initial brunt of the global financial crisis, they have emerged profitable and well-capitalised. Our major supervisory focus during the crisis has been on ADI funding and liquidity, which was badly disrupted in the latter months of 2008. Though conditions have subsequently improved, we have maintained this focus because sentiment in offshore wholesale funding markets, on which our larger ADIs are dependent, remains fragile. APRA has implemented more granular and frequent reporting of liquidity and funding positions to improve our "line of sight" on issues affecting individual ADIs and the industry as a whole. APRA has also heightened its scrutiny of credit quality and the adequacy of provisioning. Again, though loan losses appear to have peaked, APRA is continuing its thematic work on issues such as ADI commercial property exposures, an important source of impaired assets, and ADI housing lending, against the background of higher mortgage rates.

APRA's thematic work in the ADI industry also includes operational risk issues arising from significant mergers and acquisitions, and from core banking system upgrades.

The life insurance industry was buffeted by the volatility and deterioration in domestic and global equity markets during the global financial crisis, which cut significantly into profits. For that reason, APRA's supervisory focus has been on the capital strength of the industry and its capacity to withstand adverse financial market movements. That capacity has been subject to detailed stress-testing by APRA and the institutions alike. The recovery in equity markets from crisis lows, which has helped to restore profitability, has enabled APRA to ease back on its supervisory intensity but recent renewed volatility will keep us on alert. That said, improvements in capital management and reporting have improved the industry's ability to absorb further shocks.

In addition to capital management, thematic issues in life insurance include the operational and pricing risks arising from increasing product complexity, the industry's capacity to provide group risk insurance to the larger superannuation funds at sustainable prices, and the rationalisation of legacy products. It is not unknown, for example, for a life insurer to have acquired up to eight or more administration systems through mergers and acquisitions!

The recovery in equity markets over much of 2009/10 has also helped the superannuation industry to consolidate, after two years of negative returns. APRA's supervisory focus during the crisis has not been on investment outcomes as such but on the risk management practices of trustees. One particular ongoing issue is the management of liquidity risk, where we have identified scope for improvement in areas such as managing liquidity at the investment option level and comprehensive liquidity stress-testing. Another ongoing issue is the valuation of unlisted assets, where the absence of updated market prices can lead to inappropriate valuations and inequities between incoming and exiting members.

The general insurance industry has been left relatively unscathed by the global financial crisis due to conservative investment strategies and the resilience of the Australian economy. The industry remains profitable and well capitalised, although the boost to profits over recent years from reserve releases is masking some underlying deterioration in profitability. Our thematic focus currently includes stress-testing within the industry, where we have identified room for improvement in the breadth and severity of the tests applied.

Concluding comments

Global regulatory reforms are currently attracting intense industry and media interest but I have wanted to turn a spotlight today onto the work underway, but receiving much less publicity, to ensure that supervisory agencies are fully equipped for their role. Recognising that more needs to be done here, the G20 Leaders have asked the Financial Stability Board (FSB) to develop recommendations to strengthen prudential oversight and supervision, by October this year. The FSB is to look at the mandate, capacity and resourcing of supervisors, as well as their powers to intervene early in addressing emerging risks. This is an ambitious task given the range of practices and outcomes observed across our regulatory peers, and will be only the start of a more open dialogue about the quality of supervision in action, as opposed to the length of supervisory rule books.

At least that dialogue has started. We are sometimes asked in APRA whether we have been inundated with requests from other supervisory agencies to learn more about Australia‟s experience during the global financial crisis. The short answer is that we have not — agencies tend to look inwards firstly in the search for improvements. Under the auspices of the FSB, however, senior supervisors from a number of jurisdictions have been meeting to share practical experiences — successes and challenges — from the crisis, in pursuit of the answer to the question: "What makes "good" prudential supervision?" We are taking part in these workshops. And we look forward to the FSB's recommendations, which will set benchmarks to guide the further development of APRA's approach to supervision. As the crisis has brought forcefully home, no supervisory agency will ever reach the top of the learning curve!

Other issues are reinsurance counterparty risk and potential exposures of insurers to recent corporate failures.