APRA Chair John Lonsdale’s speech to AFR Banking Summit

Severe but plausible: Taking a wider view of risk
 

Good morning.

I’m pleased to be speaking to the AFR Banking Summit again this year. There are always many topics for a prudential regulator to discuss about the banking system – the challenge is to choose a couple. So I have decided to talk further today about financial system safety and the work that must be constantly undertaken to ensure the system is as safe and resilient as we can make it.

Only weeks before this event last year, global financial markets were in turmoil after several American banks failed and European giant Credit Suisse only narrowly avoided the same fate. Amid fears the crisis could impact Australia, the summit presented an opportunity to re-emphasise the strength of the Australian banking system, including some of the ways it is safer than many overseas jurisdictions. Fortunately, we avoided real contagion impacts, in part due to more than a decade of strengthening capital requirements and introducing new liquidity and interest rate risk measures. However, there are still lessons for us.

Today, we remain in an environment of heightened financial risk: inflation remains high, interest rates have risen further in the past 12 months; global growth remains uneven with specific risks emerging in both the Chinese property sector and commercial real estate; and geopolitical uncertainty has further escalated with renewed fighting in the Middle East.

One of the primary tools APRA uses to ensure banks are prepared for a range of “severe but plausible” downside scenarios is our annual authorised deposit-taking institution (ADI) stress test. The results of our 2023 ADI stress test, which I will run through shortly, indicated once again the resilience of our banking system in the face of a major economic shock. Although reassuring, the reality is that the risk of contagion from international or domestic events doesn’t stop at banks. Such is the interconnectedness of the modern financial system that a crisis in banking would impact superannuation and insurance, and vice versa. Going further, last November’s Optus outage demonstrated how even problems in non-financial services industries can spill over into the financial system.

While industry-specific stress tests will remain an important part of our supervisory toolkit, it’s clear a system-wide perspective is also needed to better grasp these linkages and potential impacts. With that in mind, APRA has begun planning our first financial system stress test. But it’s not only APRA that needs to take a wider view of risk. Banks, too, have more work to do in expanding the scope of their stress testing capabilities to incorporate a broader range of factors to ensure they’re prepared to respond to financial or operational threats, wherever they arise.

Setting the scene

While stress testing is a theoretical exercise and the scenarios are not a forecast or prediction, they are always developed by APRA with one eye on the global and domestic environment of the day. For the 2023 ADI stress test, APRA presented banks with a hypothetical scenario featuring a deep and prolonged global economic downturn, accompanied by a sustained period of high inflation and interest rates. The downturn was triggered by a supply shock in global commodities markets that further intensified inflationary pressures, leading to weaker global growth and heightened risk aversion within international capital markets.

In Australia, our scenario saw inflation sharply rise to a peak of 8.6 per cent and remain well above the Reserve Bank of Australia’s (RBA’s) target range. To combat higher inflation, the RBA further increased the official cash rate. The cumulative effect of these factors saw gross domestic product fall by 4 per cent, unemployment rise to 10 per cent and house prices fall by 35 per cent over three years. To up the ante further, we incorporated a temporary closure of international funding markets, a minimum three-notch downgrade for bank debt ratings, ongoing maturities of the RBA’s Term Funding Facility and intense competition for deposit funding.

This stress test was the first exercise to test the resilience of banks under APRA’s new bank capital framework that came into effect last January, and which was designed to be stronger and more flexible than its predecessor in responding to stress scenarios. Overall, the results were encouraging. Of the 11 large banks chosen to undertake the stress test, all had sufficient capital to withstand the severe downturn and support an economic recovery. In terms of specific details:

• The industry common equity tier 1 (CET1) capital ratio sharply fell by up to 3.3 percentage points, towards a low point of 9 per cent and well into capital buffers. 
• The banks incurred significant credit losses, while both profits and dividends fell significantly. 
• The industry aggregate liquidity coverage ratio fell sharply by 36 percentage points, however the banks were able to continue to meet their liquidity obligations.

Once banks took mitigating actions, their capital levels were restored above buffers and back towards targets. Assisting this was banks’ ability, with APRA’s hypothetical blessing, to release their 1.0 per cent counter-cyclical capital buffer – a macroprudential option not available under the previous capital framework when the default buffer was set at zero.

Capital raisings continue to be the key mitigant taken by banks, which collectively raised $37 billion of equity during the severe downturn. This increased CET1 ratios in aggregate by approximately 150 basis points and also provided a key source of additional liquidity for many banks during the stress. However, there is always uncertainty around the feasibility of this action due to a lack of precedent, given that Australian banks did not experience significant capital stress during either the global financial crisis or the pandemic. Moreover, liquidity stress can emerge in rapid and unexpected ways should public confidence be shaken. We urge banks to focus on understanding the different aspects of liquidity risks they are exposed to and incorporate them within their own stress testing exercise.

Looking around

In recent weeks, APRA has written to the participating banks providing feedback on their 2023 stress test results. This feedback not only covers what each bank’s results were compared to its peers, but also how well we assessed their capability to undertake the test.

Broadly speaking, we have observed ongoing improvement in banks’ internal stress testing capabilities, however, there remains room to do better. These areas include governance, modelling and data, and stress test scenarios.

In respect of governance, stress tests are an extensive exercise, covering a wide spectrum of risks across the entire bank. We therefore want to see a wide range of stakeholders across the banks (including the lines of business, the risk teams, senior management and the board) working together and really challenging each other to understand what their stress test results mean.

It’s not only results of the stress test at face value that matter, but also how banks understand the assumptions in their modelling and their limitations. Modelling stress scenarios are inherently uncertain and they also require progressive development. We saw some approaches to modelling that either did not highlight their limitations, or were too optimistic in their assumptions, or used models that didn’t differentiate between key sources of risk within banks’ portfolios. Our message here is that underlying investment in data and internal processes is fundamental to good stress testing.

As part of their internal risk and capital management, we also expect banks to do their own stress tests using their own scenarios. What we have seen is banks using the scenarios presented by APRA as a heavy source of inspiration. As flattering as that might be, we see potential for banks to widen the lens they take on the potential source of future risks, especially spillover shocks emanating from other sectors.

In part, this is a product of the interconnectedness I mentioned earlier. Consider the commercial property market: banks lend to it, super funds invest in it and insurers underwrite it. Many workers, however, despite their employers’ best efforts, don’t want to use it as much as they once did and would prefer to work partly or fully from home. Should we see a major correction in commercial property valuations, all three industries – banking, super and insurance – would be impacted. Climate risk is another inter-related area. The declining affordability and accessibility of property insurance in many parts of Australia, for example, isn’t only bad news for those communities. It also impacts the ability of households and businesses to get credit, to rebuild after a disaster or to repay loans, and might also impact super though requests for early releases on compassionate grounds.

But another major factor contributing to the need for banks’ stress testing to take a more system-wide view is the rise and rise of superannuation as an economic force. The banking industry remains the cornerstone of Australia’s financial system, but that status is increasingly being challenged by superannuation. Over the past decade, the value of assets managed by the superannuation sector has grown at almost double the rate of banking – 8.8 per cent a year compared to 4.8 per cent. If this trend continues superannuation assets will exceed the size of the banking sector in time. But to be clear, superannuation doesn’t need to be bigger to have an influence over financial stability.

System-wide stress

The size and complexity of the Australian financial system has grown significantly in recent years, as has the interconnectedness between industries. APRA’s regulated industries now have more than $9 trillion in total assets.

It is in this context that APRA is planning to launch our first financial system stress test in 2025. The opportunity to develop a stress testing activity and framework with the primary lens on the broader financial system was recognised in APRA’s 2023-24 Corporate Plan. It also aligns with the recommendations of the Financial Regulator Assessment Authority’s review of APRA in the middle of last year. The intention behind the new test is to sharpen APRA’s response to systemic risks by deepening our understanding of the transmission mechanisms of shocks across the financial system. We hope to gain insight into the impacts of spillover and amplification risks between industries and identify possible “blind spots” in our supervisory regime. Learnings from the process would also inform our future stress testing program, including similar future exploratory and industry-specific exercises.

We are considering both the design of this new system-wide stress test and how it might be rolled out. Our expectation is that the test would involve selected large entities and be conducted in stages, with a year for the design and a year for the exercise itself. Later this year, we will be engaging with stakeholders throughout the financial system on the design of the test, focusing on exploring systemic risk hypotheses and potential scenarios. This process will involve our fellow agencies on the Council of Financial Regulators, as well as entities and industry associations, so we look forward to your input.

This type of system stress test is at the cutting edge of regulatory best practice globally. While the US Federal Reserve has begun exploring broader market risk shocks as part of its bank stress testing program, it’s our peers at the Bank of England who have travelled furthest down this path to date. Last year, the Bank launched its first system-wide exploratory stress test exercise including banks, insurers, pension funds as well as other financial market participants. The exercise is scheduled to conclude later this year.

APRA has had several meetings with the Bank to understand its approach to the exercise and how it went about designing it, which took them the best part of a year. Their thinking, successes and lessons learned will all contribute towards the approach APRA ultimately takes in designing its own system-wide stress test. At this stage, our intention is to provide an update on our thinking before the end of the year ahead of rolling out the test in 2025.

Monitoring the blind side

Before that, however, will be the 2024 ADI stress test. Over the next month, APRA will continue to engage with participating banks on this year’s stress test, which will be conducted, as usual, from mid-May to the end of July.

Operational resilience has risen prominently as a focus for APRA. Where last year’s international banking turmoil highlighted the speed at which a liquidity crisis could spread, fuelled by online banking and internet innuendo, recent cyber incidents is a reminder that the financial sector is heavily dependent on non-financial industries to operate.

APRA released prudential standard CPS 230 Operational Risk Management last July to ensure banks, insurers and superannuation trustees could better manage operational risks and respond to business disruptions. It doesn’t make a bank responsible for a telecommunications or electricity outage. But it does mean we expect these types of scenarios to have been identified against defined critical operations and plans put in place to respond so that impacts on business continuity and customers are minimised.

By mid-year, we will be releasing the final version of the accompanying prudential practice guide, along with a response paper to the consultation. We will be including some clear expectations for implementation; what we expect to be in place from day one and where we expect practices may develop and evolve over time. In updating the final version of the guidance, APRA has been informed by insights from the cross-industry preparedness meetings we have been undertaking since October last year. Through this work, several key themes have emerged when we look at those entities making the strongest progress in preparing for the new standard.

First, entities that can shift their mindset to focus on critical operations are best able to understand the purpose of CPS 230. The new standard is about understanding and protecting the areas that have the most impact should they fail. Second, entities that are taking a logical sequence to understanding critical operations then connecting to material service providers, before progressing to defining tolerances, can manage the change in an orderly way. And thirdly, entities that are focusing on the capability needed to deliver the regulatory change for CPS 230 and sustain it are also thinking about oversight of service providers under the new standard. We’ve also received feedback from entities using pilot programs that these provide great insight into the capability needed to support the new standard.

Eyes open

Whether it’s a bank shock in the US spreading to Europe, as we saw last year, or a virus spreading globally, an increasingly interconnected world requires greater awareness of linkages and potential exposures across the financial system. That includes connections with parties that may not provide financial services, but which are essential for those services to be provided.

In developing a system-wide stress test, APRA is seeking to contribute to greater understanding of these linkages and how they could spill over or amplify a potential future shock. But this is not a task for regulators alone. Banks, as well as insurers and superannuation trustees, need to develop their stress testing capabilities further by taking a broader view of where severe but plausible risks could emerge.

Hindsight may be 20/20, but as risk managers with responsibility for protecting the community and national prosperity, our foresight also needs to be sharp. That includes our peripheral vision, because the next crisis may not develop in our direct line of sight.