Thank you for coming to the Reserve Bank of Australia (RBA). I would like to acknowledge the Gadigal People of the Eora Nation, the traditional owners of the land we meet on. I pay my respects to their Elders past and present and extend that respect to all First Nations peoples present.

Today I’ll discuss the Monetary Policy Board’s (MPB’s) new Framework for Additional Monetary Policy Tools at low interest rates. We have just published this alongside my speech.¹

The framework helps to prepare us for any future episode in which the cash rate is very low. It draws on our recent experience, the experience of our peers and the broader literature. We also tapped into the helpful guidance of three external experts: Loretta Mester, Malcolm Edey and Begoña Domínguez.²

The framework is a guide to making decisions about additional monetary policy tools. It sets out how we – the MPB and RBA staff – would approach the design, the use of, and the exit from, additional tools. It highlights the key judgements and decisions the MPB would need to make. The framework is not prescriptive. It can’t be, because future shocks are unlikely to look like the pandemic or earlier crises overseas.

Although today’s focus is on additional tools, the cash rate target remains our primary and preferred instrument for achieving low inflation and full employment. It is well understood and highly effective. Indeed, during the pandemic the largest contribution we made to supporting the economy came from lowering the cash rate target to historically low levels and keeping it there for some time. Additional tools can play an important role during extraordinary times and provide some extra support, but they are more complex and carry greater risks. Their effectiveness is also uncertain and depends on the context and the nature of the shock.

I’ll start by introducing the framework, before turning to some observations from our previous use of additional tools.

The Framework for Additional Monetary Policy Tools at low interest rates

The new framework stems from the RBA Review and the commitment to explain how the MPB would approach additional tools.³ It also reflects changes to the RBA’s governance and risk management, most notably the creation of separate Governance and Monetary Policy boards.⁴

The framework has four guiding principles designed to support an effective and tailored policy response, while taking careful account of the trade-offs and risks:

• First, any tool or package should contribute to the RBA’s monetary policy objectives and be consistent with financial stability.
• Second, the expected benefits should be reasonable relative to the potential costs.
• Third, tools should be ready to use and sufficiently flexible.
• Finally, the RBA should account for broader public sector policies and the consolidated public sector balance sheet, while maintaining operational independence.

The fourth principle is worth spelling out. Talking with other agencies will help us to make sure our actions fit well with theirs, while preserving our independence and theirs. There are two reasons. First, some additional tools transfer risks from the economy to the consolidated public sector balance sheet – the net assets and liabilities of all government agencies combined. And those potential costs should be identified clearly. Second, when responding to a shock, it is important to consider the overall effect of different tools across agencies. This includes how other agencies might respond and whether there are important implications of the RBA’s tools for their work. So, the framework emphasises consultation with the Treasury, the Australian Prudential Regulation Authority (APRA) and other relevant agencies. This approach preserves central bank independence and allows the RBA to act decisively in pursuit of our mandate as needed. But it will also ensure a common understanding of the nature of the shock to enable the most effective policy response.

A future episode

How would we approach a future episode when interest rates are low, and more stimulus is still required? At its core, the answer is straightforward. Rely on the cash rate first and foremost. Use additional tools if necessary.

In Australia and elsewhere, additional tools have generally been used once the policy rate is already constrained. But they could also be used alongside cuts in the cash rate towards zero to front-load support and reduce later reliance on them.

More broadly, when interest rates are already low, the MPB may have less tolerance for inflation falling below the 2 to 3 per cent target. In those circumstances, it may consider responding earlier and more decisively to disinflationary shocks by pre-emptively lowering the cash rate target.

Such a response may reduce the need to rely on alternative tools. But they can still provide valuable support in extraordinary times. Let me highlight three ways in which we would approach their use.

First, the purpose must be clear and communicated clearly. We would explain why particular tools had been chosen, recognising that their effectiveness depends on the context. A tool that works well in some circumstances may be much less effective in others, depending on the nature of the shock and conditions in the economy and financial markets.

We would also distinguish clearly between tools used for monetary policy purposes and those used to support market functioning or financial stability. In practice, the motives can overlap, and if they do, we would clearly say so; ambiguity reduces effectiveness and makes exit harder.⁵ Tools aimed at stimulating aggregate demand are typically larger and longer lasting, with greater implications for the consolidated balance sheet and entailing a wider range of risks. So, we need to be deliberate and clear in how we would use them.

Second, decision-making needs to be dynamic. Tools should be assessed and reassessed over their full lifecycle, across a range of scenarios, and subject to robust review and challenge. This reflects two realities: benefits often come early while costs build over time, and we are unlikely to calibrate the response exactly right in real time. Hence, as with the cash rate target, settings for alternative monetary policy tools will need to be revisited as conditions evolve.

Third, exit strategies need to be considered from the outset. There can be a trade-off between effectiveness and flexibility. A strong commitment may make a tool more stimulatory, but also much harder to unwind if the economy evolves differently from expected. We would set out how the exit would be determined as part of the design of any tool or package.

Lessons from the pandemic

As I said earlier, the framework draws on lessons from our recent experience during the pandemic.⁶ The pandemic showed that a package of additional tools can provide valuable support in extraordinary times, but success depends on the context. It also highlighted important challenges in design, calibration, communication, exit and broader risk management of the tools. These lessons are reflected in the approach we have set out:

• Be clear about the purpose of each tool.
• Reassess decisions over time as conditions evolve.
• Consider the full lifecycle of a tool at the outset, including risks and exit strategies.

To make these lessons more concrete, it is helpful to consider the use of additional monetary policy tools during the pandemic, starting with the Term Funding Facility.⁷

Term Funding Facility

The Term Funding Facility helped calm Australian funding markets during a period of severe global stress and supported the pass-through of low policy rates. It lowered funding costs for lenders, and hence it contributed to much lower rates for borrowers.

At the same time, the facility underscored the trade-off between support for the economy and policy flexibility. While the low fixed-rate and fixed term of the facility reinforced other parts of the policy package, it provided no flexibility on the amount of stimulus it provided. It also introduced interest rate risk for the RBA and the consolidated public sector balance sheet.⁸ A variable-rate facility would have involved less financial risk, though it also could have been somewhat less stimulatory. That option was not readily available at the time, but now it is.

This experience showed the value of testing a wider range of scenarios, including faster than expected recoveries. With more weight on upside economic scenarios we might have reached different decisions, including on the extension to the Term Funding Facility in September 2020.

It also highlighted the importance of the close consultation we had at the time with APRA and our engagement with the banks, particularly as the facility matured. A key system-wide risk is that a lot of banks may need to replace this type of funding at the same time, which could create pressures in private funding markets.

In sum, term funding facilities can ease financial conditions and support the economy, but outcomes depend importantly on their design and the broader context.

Government bond purchase program

Government bond purchase programs have been particularly effective when markets have been impaired. Bond purchase programs can support market functioning and reduce liquidity premia, thereby restoring the transmission of monetary policy. That was part of the logic for the RBA’s early bond purchases in 2020.⁹

However, the effect on aggregate demand of a longer-lived bond purchase program is probably weaker here than in some peer economies. While the pandemic-era program provided some support, estimated effects for Australia were at the lower end of international evidence. This probably reflects both the broader policy package in place at the time and structural features of our financial system. Namely, borrowing is largely bank-based and most of it is linked to shorter-term rates, so lower long-term yields are less relevant to overall borrowing costs. Bond purchases could also put downward pressure on the Australian dollar, but the extent of this is uncertain and depends in part on the policies of other central banks.

Large-scale bond purchase programs also expose the RBA’s balance sheet – and therefore the consolidated balance sheet – to financial risks because they involve funding fixed-rate assets with floating-rate liabilities.¹⁰ If policy rates need to rise significantly, those risks become (mark-to-market) losses. These effects should be considered as part of a broader assessment of the impact on the consolidated balance sheet, including the extent to which stronger economic activity supports tax revenues.

Taken together, experience suggests that in Australia, government bond purchases are most effective as a targeted tool to restore market functioning in periods of stress. In some circumstances, they can also be effective at easing financial conditions and supporting economic activity, but their use should be guided by the limits and associated risks of such purchases.¹¹

Forward guidance with commitment

Forward guidance in its stronger form involves committing to keep policy easier than otherwise, either until specific economic outcomes are achieved or until a specific date. When that commitment is clear, credible and well-understood, it can contribute to easier financial conditions. However, this strong form of forward guidance – particularly a time-based one – is vulnerable to changing economic conditions, under which the best response in the future may differ from the initial commitment.

The pandemic showed how hard it can be to communicate conditionality and how reputational costs can arise as circumstances change. In our case, this was made harder by the interaction with the yield target and the Term Funding Facility, which were both firmly anchored in time-based commitments.

The implication is that forward guidance with commitment can support the economy, but it is difficult to implement in practice and is generally advisable for it to be explicitly conditioned on economic outcomes, rather than tied to a fixed date.

That experience also reinforced two broader lessons. First, communication in exceptional times is more effective when there is a clear and well-understood reaction function in ‘normal’ times. Second, communication needs to be embedded in policy design from the outset, particularly when multiple tools are deployed together. The challenge is not just how forward guidance is communicated, but how it is designed, how it interacts with other policy tools and how it is understood publicly.

Australian Government bond yield target

A yield target can help to support markets when they are under strain and contribute to lower funding costs. Indeed, the three-year yield target helped achieve exactly this early on, alongside the broader pandemic policy package.

But a yield target is not well suited to a highly uncertain environment where the outlook can change significantly. As our experience showed, it shares many of the same challenges as forward guidance with the added complication of material risks to the balance sheet for the RBA. There can also be risks to market functioning, especially if the exit is disorderly.

Overall, yield targets require particular caution. They may have a role in exceptional circumstances, but only with a clear understanding of the risks, careful scenario analysis and a credible exit strategy from the outset.

In short, the most important support during the pandemic came from lowering the cash rate to historically low levels and keeping it there. Additional tools can reinforce that support, but their effects – beyond addressing severe market strains – are likely to be more marginal and their risks need to be managed carefully.

Other tools in the framework

The framework also considers negative interest rates and purchases of foreign exchange assets. The RBA did not use these tools during the pandemic, but they have been discussed publicly in the past and have been used in peer economies. Their inclusion in the framework reflects the need for a considered view of all the available options and their limits.

Negative interest rates could, in principle, provide some additional stimulus. But over time they can impair market functioning, add to financial stability vulnerabilities and weaken monetary policy transmission. They also pose significant operational and communication challenges and have proven deeply unpopular where they have been used. Even so, there may be circumstances where they are useful. In particular, the experience in some other small open economies suggests they can help resist appreciation pressures, especially when larger trading partners have adopted negative policy rates.¹²

Foreign exchange asset purchases could support aggregate demand by putting sustained downward pressure on the Australian dollar, or by resisting appreciation. That is quite different from short-term intervention to address disorderly market conditions.¹³ But the Australian dollar is a very liquid currency trading in deep markets. So, to have a meaningful macroeconomic effect, the RBA would probably need to undertake very large purchases of foreign exchange and take on substantial financial risk to the balance sheet.

While these two tools remain in the toolkit, I expect they would only be considered in truly exceptional circumstances.

Next steps

The MPB’s framework for Additional Monetary Policy Tools at low interest rates completes a Review recommendation. More importantly, it puts us in a stronger position to respond in the rare circumstances when the cash rate is very low and additional support may be needed.

To reiterate though, the cash rate target remains our primary and preferred instrument for achieving our inflation and employment objectives. But there may be times when conventional policy space is constrained and additional support is needed. In those cases, the new framework makes us better prepared to respond.

But there is more to do. An important next step is to run fire-drills involving the Monetary Policy Board, the Governance Board and RBA staff. These drills will test the framework and our internal readiness with making decisions under time pressure and with incomplete information.

Also, the framework is not static. It will be updated as we learn from those exercises, from further research and from international experience, and our own experience with any future use of these tools. Other tools could also be considered over time. And while the next crisis will not look like the last, we will be better prepared to respond to it.

Endnotes

* I thank Matthew Boge, Michelle Lewis, Laura Nunn and Callum Shaw for leading the work to develop this new framework. I would also like to thank numerous RBA staff for their helpful comments and suggestions. The views here are my own and not necessarily those of the MPB.

1 The full framework is available at ‘Framework for Additional Monetary Policy Tools at Low Interest Rates’.

2 The staff’s assessment of individual tools is available at RBA, ‘Staff’s Detailed Assessment of Additional Monetary Policy Tools - Supplementary document to the AMPT Framework’. The series of reviews focused on the RBA’s pandemic experience are available at RBA, ‘Reviews of the Monetary Policies Adopted in Response to COVID-19’. While the review by external experts helped to inform the development of the framework, the framework itself is owned by the Monetary Policy Board and reflects its approach to additional monetary policy tools.

3 See Australian Government (2023), ‘Review of the Reserve Bank of Australia’, Final Report, March and the Treasurer and the Monetary Policy Board (2025), ‘Statement on the Conduct of Monetary Policy’, July.

4 See RBA (2026), ‘Risk and Compliance Management Framework’ and RBA (2025), ‘Memorandum of Understanding among the Monetary Policy Board, Payments System Board, Governance Board and Executive’.

5 For a case study of experiences in the United States and the United Kingdom that illustrates this point, see Kashyap (2024) ‘Monetary Policy Implications of Market Maker of Last Resort Operations’, Panel Discussion, Jackson Hole Symposium.

6 For a summary of the RBA’s response to the pandemic, see RBA (2025), ‘Supporting the Economy and Financial System in Response to COVID-19’.

7 The names of the tools used during the COVID-19 pandemic do not always align with those in the new framework.

8 A fixed-rate term lending facility involves the RBA lending to banks at a fixed-rate, while paying a variable rate on the Exchange Settlement (ES) balances used to fund that lending. If policy rates rise before the loans are repaid, the interest the RBA pays on the associated ES balances increases, while the interest it receives on the lending remains fixed. This creates interest rate risk for the RBA with the cost determined by the path of policy rates. In the case of the TFF, stronger-than-expected economic conditions and higher policy rates resulted in a financial cost to the RBA of around $9 billion.

9 From March to May 2020, the RBA purchased around $40 billion in Australian Government bonds in the secondary market to support the smooth functioning of that market. For more information, see Finley R, C Seibold and M Xiang (2020), ‘Government Bond Market Functioning and COVID-19’, RBA Bulletin, July.

10 For the RBA, purchased bonds pay a fixed return, while interest paid on the Exchange Settlement (ES) balances created to pay for the bonds varies with monetary policy settings. As interest rates increase there is a financial cost to the RBA. The bond purchase program has cost the RBA $30 billion to date.

11 Riksbank Governor Erik Thedéen made similar arguments about the future use of a government bond purchase program, see Thedéen, E. (2026), ‘Large-scale asset purchases can impair the functioning of the market and involve risks to independence’, Speech, Sveriges Riksbank, June.

12 When implementing negative policy rates, both the Swiss National Bank and Danmarks Nationalbank noted the need to respond to currency appreciation and capital inflow pressures. Negative interest rates are not operationally ready in Australia.

13 Purchases in the spot market to address dysfunction in the Australian dollar would be captured under the MPB’s Policy on Financial Market Intervention to Address Market Dysfunction.