Supervisory lessons from the global financial crisis

John Laker, Chairman - AB+F Leaders Lecture Lunch 2010, Sydney

Just over two years have passed since those fateful days in October 2008, following the failure of Lehman Brothers, when the global financial system appeared to be looking over the precipice. The view down was daunting — a potential freezing of global funding markets, a collapse in consumer and investor confidence, severe global economic dislocation. As we now know, the extraordinary actions taken around that time by governments and central banks, including in Australia, took the system back from the edge. The worst fears about the impact of the global financial crisis — very real though they were then — did not materialise. The extremes of risk aversion have eased, wholesale funding is flowing and the global economy is back to trend growth overall, although a number of North Atlantic economies continue to languish under the burden of their earlier emergency rescue packages.

Today, the vista that faces the Australian banking system is very different. Ahead lies the prospect of above-trend economic growth in Australia boosted by a powerful terms-of-trade effect. This promises a very sound operating environment for our banking institutions. The key word is ‘promises’, since uncertainties persist. Recent developments in the Euro zone and on the Korean Peninsula illustrate yet again the fragility of sentiment in global financial markets. Households and some business sectors in Australia have emerged from the crisis in still cautious mood, while other businesses have been able to meet their funding needs without recourse to the domestic banking system.

The period ahead will therefore require careful navigation but it will reward those banking institutions with measured strategic ambitions and strong funding and capital positions, over those simply waiting for the ‘salad days’ of high volume growth in the early and mid-2000s to return. Those days have gone, and it is critical that the institutional lessons of the global financial crisis are not quickly forgotten.

That said, our banking institutions are clearly moving out of defensive mode. This is quite understandable. So too is APRA. As the prudential supervisor, we need to make sure that we do not have a Maginôt Line, but flexible defences that address the different types of risk that can accompany a strongly growing economy — risks of over-confidence, of poor strategic decision-making, of eyes ‘off the ball’ on credit standards and risk management generally. In repositioning our defences, we have been reviewing the supervisory lessons we have learned from the crisis. We have been assessing those aspects of our supervisory approach that proved most effective and those that we want to reinforce as we move ahead.

In my remarks today, I would like to share some of these broad supervisory lessons with you. This is a companion piece to a speech I gave in August this year to the American Chamber of Commerce on What makes „good‟ prudential supervision? Today, my focus is on the banking system — on authorised deposit-taking institutions (ADIs) as we know them — which was the regulated industry most exposed to the crisis. However, the lessons are much broader in their application. I will also offer some comments on some current risk issues in the ADI industry that are receiving closer attention from APRA. Not surprisingly, these issues go to the heart of a prudential supervisor’s focus on banking institutions — namely, credit quality and funding.

The global focus on good supervision

APRA is not alone in its introspection! Over the past year, global policymakers have been looking more squarely at the roles and performance of prudential supervisors during the global financial crisis. There is a clear recognition that global regulatory reforms, on their own, will not be sufficient to ensure a stronger global banking system; as the G-20 Leaders stated in Seoul last month, ‘... the new financial regulatory framework must be complemented with more effective oversight and supervision.’ Some banking systems in advanced economies were more buffeted than others during the crisis, even though they operated under the same broad set of global banking regulations. And some banking institutions failed even though they were reported to be, and were assessed by their supervisors to be, well capitalised and very liquid.

In May this year, the International Monetary Fund (IMF) released a paper, The Making of Good Supervision: Learning to Say "No", which emphasised the importance of an active and hands-on approach to prudential supervision. The IMF identified two pillars that support good supervision: the ability to act and the willingness to act.

The ability to act — in law and in practice — means having the authority to be intrusive and to challenge board and management judgment, the skill to adapt to innovation and the tenacity to follow an issue through to its resolution. As the IMF sees it, the ability of a prudential supervisor to act must be built on robust legal authority, adequate supervisory resources, a clear supervisory strategy and effective working relationships with other agencies.

The willingness to act manifests itself in timely and effective intervention, in intruding on poor decision-making and in questioning common wisdom, at times in the face of industry or media criticism. The consequence, quite often, is being unpopular. Willingness to act requires a prudential supervisor to have a clear and unambiguous mandate, operational independence coupled with accountability to the public, skilled staff and a constructive relationship with boards and industry that avoids ‘regulatory capture’.

The IMF paper also identifies six key elements of good supervision — that it is ‘intrusive, sceptical, proactive, comprehensive, adaptive and conclusive’. My August speech offered some comments on APRA’s approach to supervision against these key elements, which are an echo of the HIH Royal Commission’s urgings.

Last month, the Financial Stability Board (FSB) released its paper, Intensity and Effectiveness of SIFI Supervision, which provides a series of recommendations for enhancing prudential supervision drawing on lessons from the crisis. The recommendations are primarily aimed at making systemically important financial institutions (SIFIs) less susceptible to failure but the FSB sees implications for the supervision of banking institutions more generally. The FSB’s approach has benefitted from the input of senior supervisors in a number of jurisdictions, including APRA, who have been meeting to share practical experiences — positive and negative — from the crisis.

The FSB’s recommendations build on the themes in the IMF paper. There are 32 recommendations, which inter alia emphasise that prudential supervisors must have a clear mandate geared to active early intervention, operational independence, adequate supervisory powers and resources, a group-wide approach to supervision and frequent communication between institutions and their supervisors. The FSB has also identified several areas where improvements in supervisory techniques are being made, or should be made, in different jurisdictions.

General supervisory lessons from the crisis

Against this emerging global focus on the quality of prudential supervision, let me comment on some of the general lessons APRA has learned from our experience of the crisis.

APRA is considered to have come through the crisis with an enhanced standing, in the eyes of industry and, we would hope, the community more generally. Certainly, industry perceives us as a ‘strong’ supervisor and has readily acknowledged that the Australian banking system was well regulated and supervised before and during the crisis.

And herein lies the first lesson. As we see it, our most meaningful contribution to the resilience of Australian banking institutions during the crisis came from our efforts to promote their financial health prior to the crisis. These efforts reflect the priority we have given to building a robust and comprehensive prudential framework for Australia. The tough decisions taken in the good times, such as to adopt more conservative ADI capital requirements than most of our overseas peers and to impose meaningful governance requirements, meant that our supervisors were able to pursue prudent risk outcomes with a strong prudential framework to back them up. APRA’s efforts also reflect the constant oversight of and pressure on our institutions that has been the hallmark of our forward-looking approach to supervision in recent years. Individual supervisors were also supported by APRA-wide initiatives, particularly our pioneering stress-testing and other analyses of ADI housing lending portfolios, which provided important industry perspectives.

This preventative role before the crisis struck shows how supervision can act as a countercyclical force. As the IMF has noted, ‘[p]rudential supervision is most valuable when it is least valued; restricting reckless banks during a boom is seldom appreciated but may be the single most useful step a supervisor can take in reducing failures.’ This is a core lesson that APRA and all prudential supervisors must take to heart.

The second lesson. Once a severe financial crisis takes hold, prudential supervisors like APRA generally do not have the ability to dramatically change the course of events. The stability of the Australian banking system during the crisis was largely due to the work of others — the fiscal and monetary stimulus provided by the Government and the Reserve Bank of Australia (RBA) respectively, and the Government guarantee of deposits and wholesale funding — and to more exogenous factors, such as the shock absorber provided by the sharp deterioration in the exchange rate and, of course, the quick rebound in the Chinese and other Asian economies. The fundamental strength of the Australian banking system also played an important part.

APRA’s role once the crisis began was to be on top of the situation, provide information and advice to the Government and the RBA, particularly during the darkest days of the crisis in September and October 2008, and turn up the pressure on our institutions, across all our regulated industries, to ensure they recognised the severity of the situation and took necessary protective measures to maintain their financial viability. We did this job well, I believe. Our effectiveness was helped by the constructive working relationships we had developed with our institutions prior to the crisis, which enabled us to increase the intensity of our supervision, request more information — particularly on liquidity and asset quality — and give clear direction to institutions when needed. We were also able to build on our earlier stress-testing work to gauge the adequacy of capital in our regulated industries. There were spot fires to put out. After all, the crisis did expose weaknesses in some institutions’ governance arrangements, the quality of risk management, business models or capital positions. However, the spot fires were contained and APRA did not have to contemplate the type of large-scale rescue operations that a number of our overseas peers had to mount.

A third lesson, reinforcing the messages of the IMF and FSB, is the importance of adequate staffing. Although resourcing has always been tight, APRA has been able to work through a substantial prudential policy agenda and build up our supervisory staff levels relative to the number of institutions we supervise. As a consequence, APRA has been able to absorb a higher supervisory workload and, at the same time, understand our institutions better and be in more regular contact with them, than was the case earlier in the decade. Even so, the supervisory resources APRA dedicates to the supervision of large and complex ADIs is at the lower end of the scale compared to many of our peers. We prefer an effective strike force of experienced supervisors and risk specialists over a large standing army of examiners; the FSB notes that some prudential supervisors have over 100 staff per large and complex financial institution.

APRA’s current staffing levels are supported by the special four-year funding we received from the Government to deal with the global financial crisis. That funding comes to an end after 2011/12 and we will begin a dialogue with the Government and with industry next year about our longer-term staffing needs, drawing on our crisis experience.

One final general lesson. The crisis has confirmed, in our minds certainly, the importance of a clear and simple mandate, and of operational independence. The FSB’s recommendations place great weight on these requirements but we are doing no soul-searching here — arrangements in Australia have worked very well. APRA has a ‘single purpose’ mission to ensure that, under all reasonable circumstances, regulated institutions meet their financial promises to depositors, policyholders and superannuation fund members. We are not distracted from that mission, or conflicted, by what the FSB sees as ‘highly politically sensitive issues’ such as consumer protection and anti-money laundering work. As for operational independence, our relationship with Government over many years has been cooperative and effective and we look forward to continuing on that basis. When it comes to decisions on individual institutions or on the establishment of prudential standards, the call is APRA’s and we readily accept our accountability for those decisions.

Supervisory priorities looking ahead

Let me turn now to some aspects of our supervisory approach to banking institutions that we are seeking to reinforce, particularly with prospects for the industry firming. Initiatives along these lines are flagged by the FSB in its coverage of ‘improved techniques’ although they have been on our agenda well before the FSB reported.

The first aspect concerns our engagement with boards. Boards have primary responsibility for ensuring the prudent management of ADIs (and other regulated institutions, of course) and our prudential standards make boards accountable for a range of prudential matters. We also seek to empower boards by giving them specific responsibilities, and the necessary information, to ensure that they can perform the role expected of them. APRA has increased its engagement with ADI boards in recent years but, nonetheless, we are keen to have more face-to-face contact with the boards of the largest institutions. As we have learned, direct access to the board, and an effective relationship built on mutual trust and respect in good times, can be crucial in managing situations when times turn sour. More face-to-face contact will help to strengthen relationships and, at the same time, allow us to properly test board capabilities and director contributions.

A particular focus of our enhanced engagement with boards will be risk appetite. At the heart of a good risk management framework is a clearly articulated statement of the board’s appetite for risk. From this can flow the myriad of policies, procedures, limits, reporting and other internal control and assurance mechanisms that make up the entirety of a risk management framework in a financial institution. In contrast to the general and life insurance industries, APRA’s prudential standards for ADIs do not yet require a formal board risk appetite statement but a risk management standard containing this requirement is being developed.

Looking ahead to that standard, APRA has been reviewing the risk appetite statements of many of our regulated institutions to help us identify good practice. Not surprisingly, we have found a wide range of practices — some risk appetite statements are quite short while others go for many pages; some are very quantitative while others are more qualitative in nature; some are enterprise-wide in their scope while other are narrowly focussed on business lines or risk types. APRA does not have a particular template in mind. Rather, as a principles-based prudential supervisor, our interest is in how well the risk appetite statement fulfils its purpose, in conveying clearly and concisely the Board’s preferences and risk tolerances to the rest of the institution.

A second aspect of our supervisory approach that we want to reinforce relates to our capabilities for assessing business models and product lines. An important lesson from the crisis for our supervisors and regulated institutions (particularly at the smaller end) is the need to understand what ‘Plan B’ might be in the event that an institution’s business model becomes unsustainable. A forward-looking supervisor needs to understand the circumstances (on either the asset or liability side of the balance sheet) that make an institution non-viable, and what would follow. Is a merger possible? If so, with whom? Can the business model be changed? If so, how? Is liquidation the only option? The crisis has confirmed that APRA has strong skills in identifying real or potential operational problems. Our intention now is to build up our skills in understanding industry dynamics and identifying strategic problems and their solution. Stress-testing of the premises on which business models are based will form an important part of this build-up.

In this context, a particular challenge for APRA — for all prudential supervisors, for that matter — will be how to identify quickly the next generation of risky products and businesses. Prudential supervisors can never expect to be ahead of the market in terms of new product development and innovation; however, close and continuous supervision will help supervisors to at least keep abreast of developments. The crisis has taught that supervisors need to do more than review new products in a one-by-one static manner, with the focus on appropriate capital treatment; they need to monitor more actively the growth and performance of these products so as to identify any aggregate trends that raise prudential concerns.

A similar set of issues arises in the case of business lines. Supervisors by their nature tend to focus on areas where there are signs of trouble. One obvious warning sign is a poorly performing business line, where there is a danger that management will attempt to improve profitability by assuming more risk. Supervisors are generally alert to such dangers. However, the crisis has highlighted that apparently well-performing business lines can send misleading positive signs because underlying risks are being underestimated. Fast growing loan books are one example; another example — fortunately only limited in Australia — was the apparently high risk-adjusted returns on structured credit portfolios that proved in the end to be illusory, since risks were not well understood or measured.

Strengthening our analytical capabilities is a goal we take very seriously. We are, for one thing, improving the quality and usefulness of prudential data provided to our supervisors for risk analysis. For another, we have introduced a comprehensive and improved financial analysis training program, using specialist trainers, to enhance the skills of supervisors in identifying and assessing risk issues. Crucially, we are committed to keeping the skills and knowledge of our specialist market, credit, insurance and operational risk teams up-to-date with a constantly evolving marketplace.

A third aspect of our supervisory approach that we are reinforcing is our industry-wide analysis of risks, sometimes called ‘thematic’ or ‘horizontal’ reviews. One of the most powerful tools available to a supervisor is a strong ‘compare and contrast’ function, which enables them to readily and with confidence identify areas in which the institution they supervise is an outlier. Without an industry-wide perspective, there is no doubt that supervisors can fall into the trap of not seeing the wood for the trees.

To help avoid that trap, APRA has within the last couple of years established an Industry Analysis Team of industry specialists who are responsible for the preparation and updating of industry risk ‘registers’ for each of our regulated industries. These registers identify the main ‘thematic’ risks and supervisory issues for each industry, graded by their potential severity. A risk owner with the appropriate expertise is then assigned to each of the high risks identified to recommend and coordinate an appropriate APRA response.

APRA is keen to strengthen the capacity of its Industry Analysis Team and enhance our industry-wide perspectives. Aside from its involvement in the risk registers and macro-economic stress-testing, the Team is developing industry ‘dashboards’ that will help supervisors quickly identify outliers and exceptions. A liquidity and funding dashboard has already proven its value in our assessment of how ADIs have managed pressures in this area since the crisis began.

The FSB has recommended that prudential supervisors should place greater emphasis on the role played by thematic or horizontal reviews. We agree. The FSB acknowledges that thematic reviews provide useful relative rankings of institutions and, certainly, we have found that nothing focuses the mind of a board more than being told by APRA that their institution lags its peers in a particular aspect of risk management or exposure. The FSB also recognises that thematic reviews can help to determine if industry practice as a whole is strong enough to address the risks embedded in a particular business activity. Our focus on credit standards in housing lending earlier in this decade is an example of this broader role.

Before I illustrate some of our current ‘thematic’ work in the ADI industry, I want to comment briefly on a fourth aspect of our supervisory approach that has come under the spotlight during the crisis. That aspect relates to parental support. Prior to the crisis, APRA’s supervisory approach to foreign-owned entities tended to treat the parent as a source of support and comfort — that is, a positive factor in our risk assessment — rather than a source of risk to the local operation. For that reason, APRA did not generally conduct detailed analysis of the financial health of foreign parents, relying instead on the home supervisor, and we have had a deliberately ‘lighter touch’ supervisory methodology for ADI branches, as reflected in their lower rate of APRA levies.

The crisis has challenged this supervisory philosophy. In several cases, and not just in banking, we encountered basically sound Australian businesses that were destabilised by problems emanating from their foreign parent. The foreign parent was, in fact, a source of stress. As a consequence, our supervisory philosophy needs to view parent entities much more neutrally, assessing each case on its merits. We also need to ensure that we have strong relationships and clear lines of communication with foreign supervisors that can be readily tapped when an international financial group is in distress. Supervisory colleges are important in this respect. While some in industry have questioned the value of college meeting themselves, there is no doubt that supervisory relationships being established through these colleges are important networks to have in place for the next crisis.

Current risk issues in the ADI industry

As I mentioned at the outset, the current risk issues in the ADI industry that are receiving closer attention are the ‘bread and butter’ issues of credit quality and funding. This was not, of course, the order of priority when the global financial crisis struck. Liquidity and funding issues became all-consuming for a period. Although our concerns on this front have subsequently eased, recent market disturbances explain why these issues still rank high on our risk register.

Despite forecasts of major loan losses arising out of the crisis, ADI credit quality has proven to be significantly more robust than in most other advanced economies, and than in Australia’s recession of the early 1990s. One key prudential indicator is the non-performing loan ratio, which remains at less than two per cent for banks and is lower for building societies and credit unions. To put this in perspective, the figure for banks compares with a peak of well over 6 per cent reached in our 1990s recession. And a look at the Emerald Isle shows the damage that poor credit quality can wreak: the non-performing loan ratios for the large Irish banks in mid-year ranged from 5.5 per cent to 48.0 per cent!

Over much of the crisis period, APRA’s focus on ADI credit quality has not been on housing lending. Australia had its ‘wake-up call’ in this area a few years earlier and housing lending portfolios have continued to perform well. Rather, the focus has been on business lending, particularly commercial property lending. As an echo of the 1990s recession, commercial property exposures have been the main source of banks’ impaired assets. Traditionally, commercial property lending has been an avenue for rapid asset growth by ‘new’ lenders. Barriers to entry are typically low and a branch network is not critical; large volumes can be lent in single transactions; and the exposures, being secured, can offer what appear to be attractive risk-adjusted returns. Before the crisis, competitive pressures saw reductions in covenant strength, more opaque borrower structures and a willingness to finance riskier tranches of property development. Obviously, some fingers got quite burnt in this area.

Our thematic review of credit quality in commercial property lending is aimed at strengthening risk management practices. A team of credit risk specialists has been assessing:

• the risk appetite and business strategy of each ADI active in this line of business;
• underwriting standards and the quality of the origination process;
• portfolio limits on such lending and the rationale for any recent changes to limits;
• the quality of internal management reports on the performance of the portfolio; and
• the rigour of stress-testing applied to the portfolio, particularly by geographic area.

APRA also incorporated a ‘severe but plausible’ commercial property stress in its macroeconomic stress test of the ADI industry earlier this year.

Corporate and SME lending has also performed less well than housing lending during the crisis. A significant source of credit losses were the exposures that some ADIs had to certain high-profile corporations whose business models were no longer viable; however, the resilience of the Australian economy ensured that additional credit cycle effects were more muted than had been earlier expected. The focus of our thematic work here has involved close monitoring of ADI ‘watchlists’, large exposures and industry concentrations, as well as assessments of:

• the effectiveness of credit risk grading systems as early warning systems;
• the quality of internal management reports in providing timely and comprehensive analysis of credits migrating to troublesome status;
• the adequacy of frameworks for aggregating and limiting exposures to groups of related entities; and
• as with commercial property lending, the rigour of stress-testing applied to corporate and SME exposures.

Over the past year, and notwithstanding the good performance during the crisis, we have also been paying closer attention to credit standards in housing lending. As a forward-looking supervisor, we need to remain alert to competitive pressures on credit standards and there were some earlier signs of aggressive lending at high loan-to-valuation ratios. Mortgage interest rates have also been on the rise from their generation-low levels during the crisis. Our thematic work in this area starts, as with the other loan portfolios, with each ADI’s risk appetite and business strategy for the housing lending portfolio and goes on to assess, amongst other things:

• minimum origination standards and the interest rate buffer built into debt serviceability calculations;
• how lending practices are modified if confronted with rapidly escalating property prices in certain regions; and
• how overrides of underwriting standards are monitored by the ADI over time.

Some final words on our continued close oversight of the management of liquidity and funding by ADIs. Prior to the crisis, prudential supervisors around the globe tended to look at liquidity risks through the lens of a ‘name crisis’ scenario; the prospect that apparently deep and liquid global funding markets could freeze was remote, if not unimaginable. This reflected something of the ‘wood for the trees’ issue I mentioned earlier — supervisory analysis tended to focus on what could go wrong with an institution, rather than on what could go wrong in the system that could impact on the institution.

Clearly, a valuable lesson has been learned, and APRA’s processes are being revamped to ensure that this lesson is hard-coded into our supervisory processes. Our thematic work in this area now focuses much more on the stability and diversity of funding sources, cash flow analysis and stress-testing. The Basel liquidity reforms to be announced within the next few days, which will include enhanced reporting arrangements, will also ensure that liquidity risk management by ADIs will be much more rigorous in future.

Concluding comments

Let me sum up briefly. With the darkest days of the global financial crisis now, we hope, well behind us, we in APRA have been taking stock of our performance during what has been a period of unprecedented global turmoil. There is always a danger in making judgments with the benefit of 20/20 hindsight. That said, I believe that APRA’s supervisory processes, supported by a robust prudential framework, helped to ensure that our banking institutions were in generally good shape before the crisis, and provided a basis for rapid responses and more detailed engagement once the crisis hit. As a risk-based supervisor, we would, of course, never claim that we had — or could — identify every vulnerability in our regulated institutions.

For us, the most important lessons from the crisis are that we should continue and, when necessary, reinforce our sceptical and questioning culture, maintain the intensity while increasing the breadth of our supervision, and promote more forward-looking analysis of the strengths and frailties of our institutions. All this, in the face of inevitable staffing pressures and temptations in industry to declare victory and move on. As the IMF has emphasised, ‘[s]upervisors are expected to stand out from the rest of society and not be affected by the collective myopia and consequent underestimation of risks associated with the good times.’ If good times do indeed lie ahead for Australia, you know where APRA will be: remaining most valuable when least valued!