'Bagehot and the Lender of Last Resort – 150 Years On' - Speech by Brad Jones

Well, good morning. Thank you for those very kind remarks. This event is a highlight on the academic conference circuit, so I'm very thankful to the Institute of Global Finance for the invitation to offer some remarks today. This year marks the 150th anniversary of Walter Bagehot's Lombard Street, where he spelled out the terms by which central banks should act as the lender of last resort to help stabilize the financial system in times of stress. As the banking turmoil offshore earlier this year demonstrated, policymakers continue to wrestle with the issue 150 years later. I'll begin today with a brief history of the origins of emergency central bank lending internationally and in Australia, discuss how the application of emergency liquidity operations has evolved around the world in more recent times, and finally I'll conclude with a review of how these wider developments have been reflected in aspects of the Reserve Bank's own framework. Let's begin with some history. Maintaining financial stability has been a cornerstone responsibility for central banks for as long as they've been in existence. The reason for this is twofold: one, financial stability is a public good; two, central banks occupy a unique position at the heart of the financial system, which reflects their ability to create safe liquid assets on demand. But how central banks have discharged this responsibility, including as lender of last resort, has evolved and been subject to much debate right up to present times. The earliest direct reference to the lender of last resort concept was traced to Sir Francis Baring following a wave of bank runs in England during their war with France in 1797. Henry Thornton progressed these ideas further in 1802 by noting the tradeoff between encouraging moral hazard on the one hand and allowing systemic risk to go unaddressed on the other. But it was not until 1873, some years after the collapse of London's biggest bill broker, that Bagehot offered his famous counsel: to avert panic, central banks should lend early and freely to solvent firms against good collateral and at high rates. Central to Bagehot's axiom was the idea that confidence could be enshrined and an unnecessary credit contraction averted by a conditional commitment from the central bank to provide liquidity insurance in a crisis. By the end of the 19th century, the Bank of England had become well practiced in last resort lending. On the other side of the Atlantic, the establishment of the Federal Reserve in the shadows of the First World War was a direct response to the liquidity crisis that engulfed the American financial system in 1907. Up until such time, it had been left to private financiers like JP Morgan to coordinate ad hoc responses to financial panics. A century on, the Global Financial Crisis saw the Federal Reserve invoke emergency lending powers, allowing it to lend to non-bank institutions like Bear Stearns and AIG for the first time since the Great Depression. If the Federal Reserve System was in large part born out of the 1907 liquidity crisis in the United States, then the lineage of today's Reserve Bank of Australia can be traced to the two major financial upheavals of the 1890s and the 1930s. Indeed, the 1890s crisis saw more than half of Australia's trading banks forced to suspend payment, one-third of whom never reopened. Partly as a result, the 1911 Act establishing the Commonwealth Bank, what would later become the Reserve Bank, was a compromise between those pressing for a central bank with expansive powers over private banks and those seeking to nationalize the banking sector altogether. In the aftermath of the Great Depression, the 1937 Royal Commission into the Monetary and Banking Systems highlighted the need for a central bank with last resort lending powers. But in contrast to a number of other countries, last resort lending has been conducted sparingly in Australia since Federation. Last resort loans have been extended only to the Primary Producers Bank in 1931 and to three banks supporting a liquid building societies in 1974 and 1979. The Reserve Bank also provided a liquidity facility to support the orderly takeover of the Bank of Adelaide in 1979. It's beyond the scope of these remarks to delve too deeply into why last resort lending has been deployed sparingly in Australia, but I'll briefly touch on three elements. One is that the Australian financial system entered the largest international crisis of the past century, the Great Depression and the Global Financial Crisis, with fewer vulnerabilities than elsewhere. But this resilience was forged out of the earlier stresses of the 1890s and the early 1990s and the regulatory responses that followed. Take the 1890s crisis, which left in its wake a lasting cautionary impression on a generation of bankers and regulators. Just three banks failed in Australia during the Great Depression compared to more than 9,000 in the United States. Decades later, the distress experienced by some lenders in the period from the late '70s to the early 1990s, coupled with substantial regulatory reforms, led to a restructuring of the Australian financial system that contributed to avoiding the worst of the credit quality problems experienced internationally during the GFC. Stepping back, it's notable that no Australian bank depositor has lost money since Federation, with the single exception of depositors at one small bank in the 1930s who lost just 1% of the value of their deposits. The second explanation for why last resort lending has rarely been used in Australia can be traced to the numerous instances where private sector liquidity support has been extended as an alternative to direct official support. Examples can be found during the Great Depression, 1970s and '90s recessions, and again during the Global Financial Crisis. A related lesson is that the private sector is more likely to find solutions to its own problems when prospective suitors have a firm understanding of the risks they are taking on. As the Lehman Brothers episode illustrated, complex business models and opaque risks on the balance sheet of troubled institutions can serve as a strong deterrent to potential buyers. A third explanation relevant to recent decades relates to the bank's willingness to flexibly deploy its open market operations during episodes of system-wide liquidity stress as a means of forestalling larger problems. I'll return to this later. Let me now discuss the traditional framework for last resort lending to banks as a public good. In the normal course of events, liquidity operations of central banks are prosaic affairs. Dealing rooms stand ready to supply enough liquidity against high quality collateral to ensure the demand for cash can be met each day at the target cash rate. But last resort lending by central banks is fundamentally different, involving the dire circumstance where stability of the financial system is in question. For most of history, this has meant lending to private banks. This reflects two things: the pivotal intermediation role banks play in the economy and the liquidity risk inherent in a business model where illiquid long-term loans are funded with runnable deposits. We have long accepted as a society that liquidity and maturity transformation of this sort is economically and socially valuable. After all, requiring banks to fully self-insure against liquidity risk by holding only the most liquid assets would undermine their ability to extend credit and see credit risk concentrated in less regulated institutions. However, while depositors and money markets might do a decent job of discriminating sound banks from less sound ones in normal times, liquidity stress at one institution can rapidly spread in a panic. And because banks are highly leveraged, there need only be a question over the value of some assets to affect perceptions of solvency and set off a deposit run. If the withdrawal of liquidity turns indiscriminate, overwhelming pressure can bear down on institutions that were healthy just a short time earlier. A self-fulfilling panic can ensue as banks are left to raise liquidity through asset fire sales, further depressing asset values and threatening the solvency of previously sound institutions. This scenario has all the hallmarks of a market failure. Left unaddressed, it risks harming the broader financial system and the economy at large. And so enter the central bank. Only it can act as the backstop provider of emergency liquidity insurance to fundamentally sound institutions in bad states of the world. In such circumstances, last resort lending can spare society from the severe disruptions associated with bank failures. Let me now turn to how last resort liquidity operations have evolved internationally in more recent years. Bagehot's initial prescription might seem straightforward enough, but on closer inspection it provides only the most general counsel. Since 2008, central banks have had to confront a range of issues that were not addressed or foreseen in Bagehot's time. Some of these have presented significant governance and operational challenges for central banks. Others have connected to deeper questions about the limits to public institutions insuring society against bad states of the world and whether assistance to individual institutions risks undermining the principles of capitalism. While there is no uniform approach to tackling these issues, it is possible to trace out some areas where a degree of commonality has emerged across central banks internationally over recent years. In my assessment, the most consequential development has been the formalization and more open communication of the framework for last resort operations. Prior to the GFC, arrangements for last resort lending were often hastily compiled without the benefit of formal frameworks or playbooks, nor were they well communicated. In fact, less than half of the central banks in advanced economies had publicly released statements on their emergency lending policies, some of which were deliberately vague while retaining elements of flexibility. Clearer frameworks have since become the standard. Moreover, the previous regime of constructive ambiguity has given way to constructive clarity in public communications. This reflects a few considerations. It better facilitates the recovery and resolution plans now required by prudential regulators and which were mostly absent prior to the GFC. It acknowledges that the previous ad hoc approach was unhelpful for decision makers and could exacerbate market uncertainty. And the issue of moral hazard has become somewhat less vexed, reflecting tighter prudential liquidity requirements and severe, well-telegraphed consequences for the leadership of mismanaged institutions that require public support. Another feature of the evolution in last resort operations has been to encompass a more expansive set of institutions, collateral, and asset markets. For a start, most central banks now incorporate liquidity facilities for non-banks, including financial market infrastructures. The Federal Reserve and Bank of England have gone further in establishing sector-wide liquidity facilities for investment vehicles like money market funds in the case of the US, and pension and insurance funds in the UK. This recognizes the key intermediation role these entities now play in the US and UK economies and the country-specific features of their operations that can leave them vulnerable to liquidity stress. Collateral management practices have also evolved. A wider set of collateral is now accepted in liquidity operations, and to assist institutions with recovery planning, more information is publicly available on valuation haircuts. Furthermore, central banks have displayed a greater preparedness to expand their operations beyond secured loans to include intervening in outright terms as a buyer or market maker of last resort during periods of dysfunction in key markets. This recognizes the potential for market dysfunction to destabilize the financial system and threaten the attainment of central banks' monetary policy goals. Previously, these types of interventions were rarely observed. But one aspect of last resort lending that has seen less change relates to the stigma institutions tend to associate with emergency borrowing from the central bank. Some fear a Northern Rock episode where knowledge of their intent to access emergency liquidity leaked and made a bad situation worse. Continuous disclosure requirements for banks can be a related complication. In an effort by central banks to meet their public accountability obligations on the one hand while not making a liquidity run worse, some now disclose emergency bilateral lending only on an aggregated and significantly lagged basis. But balancing the various tradeoffs remains a challenge for stressed institutions and central banks alike. And though a lot has changed in recent years, one guiding principle on last resort lending has remained unimpeachable, and that is central banks should not lend to insolvent institutions. This is close to an iron law in central banking, and there are good reasons why the doctrine has stood the test of time since at least the days of Bagehot. First, the decision to support a fundamentally unsound institution is a distributional one most appropriately made by elected officials. Second, it would risk overstepping the legal authority of the central bank. Third, even just the prospect of solvency support from a central bank could reopen moral hazard risks that have recently been tamed somewhat by regulatory reforms. And fourth, it would likely worsen the stigma problem. For instance, if it was accepted that the central bank will lend to insolvent institutions, those that are fundamentally solvent but experiencing a temporary liquidity shortage will be disincentivized to request central bank support if they feared such information could leak into the public domain. Let me now turn to the contemporary framework for emergency liquidity operations in Australia, which has been refined in a number of respects since the Global Financial Crisis. The first point to make here is that in periods of generalized stress, the Reserve Bank always stands ready to quickly and significantly boost system-wide liquidity through open market operations with eligible counterparties. There is considerable flexibility in these operations. They can be introduced at short notice and conducted on a weekly, daily, or intraday basis. Lending maturity terms can be readily adjusted. Requirements on collateral eligibility can be adjusted, including to include a broader pool of securities. And the amount of liquidity provided at each operation can be expanded. The elastic supply of liquidity in periods of market-wide stress can help to forestall systemic risk by dampening the cost of liquidity and reducing uncertainty about its availability. This liquidity can in turn flow from eligible counterparties to other financial institutions that cannot borrow directly from the Reserve Bank. This said, where liquidity pressures are confined to an individual institution or in parts of the financial system where open market operations are less well suited, the bank can consider providing what we call exceptional liquidity assistance, or ELA, directly to eligible institutions. In recent years, the bank has continued to refine its ELA framework and publicly communicated its key elements to assist financial institutions in their recovery and resolution planning. Detailed information is readily available on the bank's website in relation to eligible counterparties as well as eligible collateral for open market operations, which would also be acceptable for ELA, and valuation haircuts. Last month, APRA also published its proposed prudential guidance in relation to ELA with the aim of ensuring authorized deposit-taking institutions, or ADIs, in Australia have the right systems and information in place to enable them to apply for ELA in short order. And the bank maintains a detailed ELA execution playbook that sets out the processes to be followed by CFR agencies. As the Reserve Bank's public guidance in 2021 set out, in the rare circumstance where an eligible counterparty experiences acute liquidity difficulties but is solvent, the bank may provide ELA if it is judged to be in the public interest. Provision and terms of ELA are at the absolute discretion of the Reserve Bank, but a baseline expectation is that ELA would be provided on a secured basis via repo for a short period of time. Entities would also be expected to do three things: have informed their regulator immediately of any liquidity concerns and their intention to request ELA prior to approaching the Reserve Bank; have already made reasonable efforts to access private sector sources of liquidity; and present evidence of their solvency, including an attestation of positive net worth. ELA can also be considered by the bank where APRA deems temporary bridging finance necessary to facilitate the orderly resolution of a solvent prudentially regulated entity. Consistent with the international experience, the scope of the Reserve Bank's liquidity assistance operations has also been broadened and clarified. During the GFC, for instance, the bank substantially widened the list of collateral eligible for open market operations. This included self-securitizations, which are structured pools of mortgages that enable ADIs to transform illiquid loans into cash via repos with the bank. Self-securitization became a key source of collateral for ADIs accessing the term funding facility during the pandemic and could be considered for use in ELA operations. And while the historical focus of liquidity assistance in Australia has been on system-wide support for ADIs, ELA could be considered for a domestic clearing and settlement facility under broadly similar conditions and arrangements for ADIs. This includes during recovery or resolution when extraordinary measures were being taken to ensure the facility did not pose a systemic threat to financial stability. In addition to lending to eligible institutions on a secured basis, the Reserve Bank has long been prepared to support financial system stability by intervening directly in financial markets. But as elsewhere, in the years before the GFC this was perhaps most evident in foreign exchange markets. More recently, it has been seen in outright purchases of government securities during periods of extreme market dysfunction. As mentioned earlier, this preparedness to intervene directly in government securities markets to restore orderly functioning has become standard practice among central banks. It recognizes that dysfunction in these markets directly affects key areas of central bank policy, namely monetary policy implementation, monetary policy transmission, and financial stability. That said, the Reserve Bank has yet to see the case for lending directly to investment funds in the way that has recently occurred in the US and the UK after major dislocations there. The main reason is that historically, at least, the Australian financial system has been less directly exposed to the risk of systemically important liquidity mismatches associated with these funds. For instance, unlike the US, we don't have a large money market fund industry that is interconnected with the banking system. Australian superannuation funds don't have runnable liabilities in the traditional sense, at least, and relative to their UK counterparts are much more restricted in their capacity to borrow, have larger cash holdings, and most do not offer guaranteed returns to members. Nevertheless, given the industry is an important source of funding for banks and the events of 2020 showed superannuation funds can experience liquidity draining events like margin calls on currency hedges, it is important that their approach to liquidity risk management continues to strengthen in line with APRA's recently revised guidance. Let me conclude. One overarching lesson from the global disruptions of recent years is that financial system participants need to conduct their affairs with the expectation that large liquidity shocks will occur. Put simply, the next major shock is a matter of when, not if. There's considerable work underway to guide regulators in responding to the liquidity stresses affecting the global financial system over recent years. This said, I don't think it's too premature to offer some preliminary reflections. First, severe liquidity stress can emerge in a financial system even in an environment of abundant central bank reserves. The distribution of liquidity across institutions matters, not just the aggregate level. Second, as we saw in March, bank deposit runs can now occur far more rapidly than envisaged when the Basel III requirements were put in place. I don't see this strengthening the case for the straitjacket of narrow banking, but it does point to ADIs probing whether they have prepositioned sufficient eligible collateral with the central bank that could be exchanged for cash at short notice as a form of crisis insurance. This contingent collateral helps to ensure ADIs are operationally ready in peacetime and reduces their need to engage in fire sales of government securities to raise liquidity in periods of stress. To the extent that contingent collateral comprises mostly self-securitized assets rather than securities, it is also less distortive. The asset mix of ADIs would continue to reflect their core economic function, which is to extend loans. In Australia, prudential guidance on self-securitized assets as a form of prepositioned or contingent collateral was strengthened last year, while other countries are now giving similar thought to the issue. Third, left unaddressed, systemic risk can emerge from liquidity stress experienced by bank and non-bank lenders that are not considered systemically important by traditional metrics. I spoke about this recently so won't labor the point here, other than to say that the concept of what is systemic needs careful consideration. Fourth, as an increasing share of activity in the global financial system is conducted outside of the traditional banking industry, central banks will be prompted to revisit their framework for emergency lending to non-banks and outright buy or sell operations in key markets. Ensuring these operations operate as effective backstops and don't interact with monetary policy in an unhelpful way will similarly require ongoing consideration. This latter issue is a bigger challenge when severe liquidity shocks require the central bank to expand its balance sheet at the same time that it is tightening monetary policy. Finally, while prevention is always preferable to cure and there should be no illusion that the primary obligation of industry is to manage its own risk prudently, recent events internationally have highlighted the importance of clear response frameworks and public communication by policymakers in periods of stress. Executed well, this can help to stop a financial panic in its tracks. Rest assured that the Reserve Bank and CFR agencies continue to strengthen crisis preparedness arrangements to ensure that the Australian financial system remains resilient far into the future. Thank you, and I welcome your questions.

Thank you for the question. Your first question was on the rise of electronic banks, is that right? The way that regulators would typically respond to this is that institutions that are conducting the same activities have the same risks should be regulated in the same way. I think what your question actually pertains to, the link between your first and second question, is how is innovation in the financial system and the banking system, have we got the right regulatory guardrails in place? The way that countries internationally have thought about this is twofold. As I said, if it's conducting similar sorts of activities, it should be regulated firmly within the existing regulatory perimeter for the conventional banking industry. And what we've seen there is a very significant uplift both in terms of capital and liquidity requirements, governance in general. The Basel III requirements, they apply to the largest institutions. What we've seen is different countries approach differently the question of whether those Basel III requirements are effectively a cap or a floor for their own large banks. In Australia, the way that APRA has approached it is to build a regulatory regime that is super-equivalent to Basel III. So we take Basel III almost as a minima and add layers of protections on top of that in terms of capital and liquidity. There's also a big question at the moment globally about what sort of regulatory regime is appropriate for banks not captured by Basel III, the smaller institutions. There you've seen a wide variety of approaches. For instance, we've seen in the United States that a lot of the banks that recently got into trouble, in fact all the banks that recently got into trouble there, were not captured by the Basel III requirements. That is prompting a lot of introspection and reflection about how we should be thinking about systemically important thresholds. I don't think there's a clear answer yet, but I can certainly tell you at all the international meetings I attend, this question of what is a systemically important institution is very much in play. Thank you.

I don't want to be drawn into commenting on fiscal policy here. If I could maybe just reframe your question slightly into the context of this discussion, which is one thing that we've learned, we certainly observed in March, was the importance of having all the key regulators and key public stakeholders including the government acting in a powerful, coordinated, clear way. Hence my remarks at the end there about the importance of very clear frameworks, very clear public communication by all the important agencies, which in a crisis inevitably is going to involve the treasury as well. I think that lesson was learned during the Global Financial Crisis and we relearned it in March again.

The IMF, the BIS, and individual central banks have looked at this issue. The consensus internationally and here in Australia is that the crypto market is not of sufficient scale and doesn't have sufficient interconnections with the traditional banking industry to pose systemic risks. If that market were to develop very rapidly and interconnections with traditional finance were to take hold, then I think you would see a much more vigorous response. What you are seeing, where you are seeing a very vigorous response that's true domestically and also particularly internationally, is to ensure that whatever activities are going on in the financial system are consistent with the law and regulatory requirements. Our colleagues at ASIC have been very strong and very clear about how they view those risks and activities and the obligation of all participants in the financial system to be operating within the law and within the regulatory guardrails that have been set.

One thing we've observed in the last decade or so is that the market share for credit for the so-called big four has fallen about 10 percentage points over the last decade or so. So your opening proposition is absolutely true, we do have quite a concentrated banking system, but the degree of that concentration has actually diminished somewhat from a very high level to a slightly less high level over the last 10 years. That suggests to us that we're seeing some competition at work. There are a couple of banks in particular that have made very significant inroads into market share more generally. We are seeing innovation in the financial system in Australia. The key from a public policy perspective is to ensure that we've got the policy settings that encourage competition and innovation, but not in a way that threatens financial stability. That's the constant tradeoff we wrestle with every day at the bank, including in our work in the payment system, for instance. But it's certainly true for other regulators in Australia as well. They are very aware of the need to make sure that our regulatory settings are facilitating and encouraging competition where it's in the national interest, but in a way that doesn't undermine financial stability. I can show you that every meeting of the Council of Financial Regulators we're talking about these issues.

As you would expect, the bank closely monitors developments in the housing market, both price developments and non-price developments. There's certainly no sense at the bank of wanting to target specific levels of house prices. That's not the way the bank thinks about it, certainly not the way that APRA would think about it either. Where we from a financial stability perspective are most focused here is first of all on household resilience. So to the extent that households have a lot of debt in Australia, we spend a lot of time at the bank, an increasing amount of time, looking at very large databases, lots of disaggregated data, loan level data, to better understand how resilient Australian borrowers would be to different types of shocks. We feature that work every six months in the semiannual Financial Stability Review. That's a key work program for us. I would also add that APRA is looking at this issue through the prism of the ability of lenders to absorb an increase in non-performing loans, whether it comes from household borrowers or from the corporate sector. So we're looking at households and businesses. We're also looking at the resilience of the financial sector to absorb shocks, including an increase in non-performing loans. That's really where our focus from a financial stability perspective is, not so much on the level of or recent change in house prices.

I lost the first part of your question on employability.

When you say employability, what do you mean by that?

It's a general caution. You are aware that the unemployment rate is very low.

I think maybe the way to cut through there a bit is, I think possibly what you're circling around there is questions of housing affordability. This is a critical issue of national significance. There's a big intergenerational component to this issue. The Reserve Bank and financial regulators have fairly limited tools, in fact very limited tools, in that regard. Ultimately, what we're going to need to do as a country, as a society, is to look at a whole range of supply side measures and demand side policies to ensure that we're setting ourselves up for the future in a way that housing affordability was certainly less of an issue than it is today. Right now it's obviously a very binding constraint on upward mobility, and there are a whole host of not just economic but social consequences from having unaffordable housing. That's why the government in particular is very focused on this issue. They've recently announced some measures, and as a fellow Australian, I think it's in all of our interest to get to a better place.

You're right, that is a separate dedicated speech, and a number of my colleagues including the governor quite recently set out in great detail how the bank is thinking about that. Because it's kind of off topic, I might refer you to those other speeches. At a very high level, we're looking at it through multiple dimensions: the impact on the economy, we're looking at it through the impact on the financial system including financial system stability, and the bank is also engaging in a very active way with domestic other domestic agencies and international agencies, for instance in the areas of developing frameworks to help incentivize investment in renewable energy and so on. So there's a multifaceted approach the bank is taking to that. I'd suggest that if you have a look at the Reserve Bank website, you'll see a number of speeches over recent years that set out in detail how the bank is thinking about those various dimensions.