Macroprudential Policy Framework

Non-technical summary

This document outlines how the Reserve Bank of New Zealand – Te Pūtea Matua (the Reserve Bank or we) uses macroprudential policy to mitigate systemic risk and support financial stability. In this context, systemic risk relates to disruptions that affect all lenders, such as an economic downturn, rather than risks limited to individual institutions.

Macroprudential policy refers to tools we use to target these system-wide risks. This complements microprudential policy, which focuses on the soundness of individual deposit takers, for example by ensuring they have sufficient capital and liquidity to withstand adverse events.

Macroprudential policy is designed to mitigate negative feedback loops between the financial system and the economy. It contains excessive risk-taking by lenders during boom periods to improve the resilience of borrowers in an economic downturn. It also mitigates the risk that deposit takers reduce lending in an economic downturn, which could exacerbate the downturn by reducing household and business spending.

Without macroprudential policy, it is more likely that banks’ lending behaviour exacerbates downturns in the economy, like we saw in the 2007-09 global financial crisis.

Borrower-based measures address systemic risks related to the housing market

Borrower-based measures include loan-to-value ratio (LVR) and debt-to-income (DTI) restrictions.

LVR restrictions limit how much residential mortgage lending banks can provide to borrowers whose loan is large relative to the value of the property(s) that the loan is secured on. This essentially requires most borrowers to have a minimum deposit or level of equity. LVR restrictions reduce the potential losses for lenders and borrowers by reducing the extent of negative equity faced by borrowers in a housing market downturn. They are designed to be at their long-run settings most of the time.¹  However, we expect to temporarily tighten settings from this long-run level in situations when systemic risk increases materially.

Similarly, DTI restrictions limit mortgage lending with high DTI ratios. By linking borrowing capacity to income, these restrictions reduce the probability that borrowers default on their loans in a period of stress. They are designed to act like guardrails against excessive risk-taking, meaning they only constrain lending in housing booms and periods of low interest rates. As such, we do not intend to adjust DTI settings regularly. However, we may consider whether the settings remain appropriate from time to time.

These measures support financial stability by reducing the number of loans in default and, relatedly, the scale of losses faced by borrowers and lenders, particularly in the face of an adverse event. In turn, this also helps to reduce the risk of a downturn in the broader economy.

Our LVR and DTI restrictions allow for a percentage of lending above set thresholds, giving lenders discretion to offer some high-LVR or high-DTI loans to otherwise creditworthy borrowers. We do not regulate how lenders use this discretion when making individual loans.

Capital and liquidity tools target banks’ balance sheets

These macroprudential tools include the counter-cyclical capital buffer (CCyB), sectoral capital requirement (SCR), and adjustments to the core funding ratio (CFR) requirement. They involve adjusting requirements for banks through the financial cycle to support financial stability.

The CCyB allows us to temporarily reduce banks’ capital requirements in a significant economic downturn to prevent these requirements from unduly constraining access to credit. Similarly, we could temporarily lower the CFR requirement when wholesale funding markets tighten to reduce the risk of a broader reduction in access to credit.

By contrast, we could use the SCR to increase capital buffers for loans within a specific sector if there is a build-up of systemic risk in that sector.

We will review policy settings at least annually

Annual reviews will assess whether housing-related risks, such as unsustainably high house prices and rising shares of high-risk lending (as a percentage of total lending), warrant tighter policy settings. Our focus in these regular reviews will tend to be on LVR settings. In periods of systemic stress, we may need to run ad hoc reviews as conditions worsen rather than waiting for our annual review.

Reviews of long-run settings will be less frequent (around every 3 to 5 years) and will consider structural changes in the economy, behavioural responses to policies, and evidence of unintended consequences.

We intend to consult publicly when considering new tools or a material tightening of settings. However, we generally will not consult on urgent changes to settings for existing tools and non-urgent easing of settings.

While decision making on macroprudential policies is delegated to the Financial Policy Committee (FPC) by the Reserve Bank Board, we will continue to inform the Minister of Finance and Treasury prior to macroprudential policy announcements.

The build-up of systemic risk can vary over time, which relates to the ‘boom-bust’ nature of financial cycles. During ‘booms’, credit growth and asset prices can become unsustainable. This increases the risk of a ‘bust’, potentially resulting in significant losses for banks, businesses, and households. During a bust, banks’ ability and willingness to continue lending to the economy can reduce sharply, potentially leading to negative feedback loops where the downturn in the financial cycle and the slowdown in the economy become mutually reinforcing.

One of the reasons that boom-bust cycles can develop is because the financial system is procyclical. Rising asset prices and narrowing credit spreads can lead to borrowers and lenders underestimating risks in an upturn, encouraging looser lending standards and more lending. This dynamic can continue to amplify the economic and financial cycle until a downturn occurs and the process works in reverse. A slowdown in economic activity can result in falling asset prices, increasing defaults and rising risk perceptions, leading to banks restricting their lending to the economy. The reduction in credit availability can cause further declines in economic activity and asset prices, creating negative feedback loops. This dynamic can have significant and lasting impacts on the economy, partly determined by the prudence of bank lending standards before the systemic stress occurs.

Evidence shows there is a strong relationship between banks’ lending standards during the good times and the sensitivity of household consumption to financial stress. Floden (2014) finds that countries with high household DTIs in 2007 suffered a larger reduction in consumption between 2007 and 2012, after controlling for other factors. Research on New Zealand data has similarly shown household leverage reinforces the housing wealth effect in a housing bust (de Roiste, Fasianos, Kirkby, & Yao, 2019).

The housing market is the main source of potential systemic risk to New Zealand’s financial system. Residential mortgage lending comprises more than half of banks’ assets.5 While we do not target house prices as an end goal of macroprudential policy, they are one of the key indicators we use to inform policy settings. For example, strong house price growth and household credit growth are both linked to the likelihood and severity of housing market downturns (Thornley, 2016). If house prices rise significantly above levels implied by economic fundamentals, the risk of a sharp correction in prices also increases. This could potentially result in large losses for banks, households cutting back spending, and adverse impacts on the wider economy. As a result, macroprudential policies that aim to mitigate systemic risk related to the housing market may have the intermediate objective of constraining excessive mortgage lending and consequently slowing house price growth.

Figure 1 illustrates how losses in the housing market can also cause banks to reduce lending and trigger a negative feedback loop that can further amplify an economic downturn.

When macroprudential policy was introduced, we expected these measures to be in place temporarily while risks were elevated and removed when risks abated. However, our approach to setting macroprudential policy has evolved over time. While we still intend to adjust macroprudential policy settings in response to changes in systemic risk, we now have long-run settings for our macroprudential tools that we expect will remain in place most of the time. These long-run settings apply when risks are not accumulating rapidly and balance the benefits of promoting resilience against the potential costs from restricting access to credit.

We still intend to tighten some policy settings when risks are elevated to promote a higher level of resilience when it is needed most. Tightening macroprudential policy settings in a boom period can lean against the procyclical nature of the financial system and dampen the amplitude of financial cycles. Macroprudential policy also helps build resilience so that, when a bust does happen, the financial system is less likely to amplify the downturn. These benefits for financial stability are generally only observable over the longer-term and when systemic risk crystallises.⁶  Table 2 notes some reasons why systemic risk can vary over the financial cycle and how macroprudential policy can mitigate these risks to support financial system resilience.

3. Legal and regulatory framework

Macroprudential policy is one way in which the Reserve Bank uses its statutory powers to meet statutory objectives and purposes. In December 2028, once the relevant provisions of the Deposit Takers Act 2023 (DTA) are in effect, macroprudential policy will be applied under the DTA.⁷

The DTA creates a single, modern regulatory regime for all financial institutions in the business of ‘borrowing and lending money’ in New Zealand. This will include banks and non-bank deposit takers.⁸

The Deposit Takers Standards (the standards) will replace our existing prudential requirements that are currently contained in several different sets of documents (such as Banking Supervision documents and Banking Prudential Requirements) and given effect through banks’ conditions of registration.⁹  The standards will set the rules that deposit takers must meet and will be in the form of secondary legislation. Standards may be issued if we are satisfied that they are necessary or desirable to achieve one or more of the purposes of the DTA, after having regard to a range of statutory principles and the Financial Policy Remit issued by the Minister of Finance.¹⁰

Whilst the DTA does not explicitly refer to macroprudential policy, it allows us to use various prudential tools to achieve macroprudential policy objectives – for instance, to avoid or mitigate adverse effects of the risks to the stability of the financial system and risks from the financial system that may damage the broader economy.¹¹

The various macroprudential policy tools will be set out in a range of standards:¹²

• The Lending Standard provides for LVR and DTI restrictions (that is, borrower-based measures).¹³
• The Capital Standard provides for the CCyB and SCR (that is, capital-based measures).
  The Liquidity Standard provides for adjustments to the CFR (that is, liquidity-based measures).

These tools are discussed in more detail in the next section.

Scope

Under the DTA’s Proportionality Framework, the calibration of prudential requirements on locally incorporated deposit takers will take into account their relative size and complexity, placing them into three groups.¹⁴

At this stage, we intend to apply our macroprudential policy tools only to Group 1 and 2 deposit takers. None of the tools will apply to Group 3 deposit takers as they are small and do not currently pose systemic risk to financial stability. However, we intend to monitor this and will change our approach as needed (this is discussed in the Review and Evaluation section).

We do not intend to apply our macroprudential policy tools to branches of overseas deposit takers given that the business they conduct in New Zealand will be restricted to corporate and institutional clients. This means that branches of overseas deposit takers will not be involved in residential mortgage lending, which is the main source of potential systemic risk to New Zealand’s financial system. Moreover, branches are not subject to quantitative capital or liquidity requirements in New Zealand.

It is noted that non-deposit taking lenders are generally outside the scope of prudential requirements. However, the DTA enables us to apply the Lending Standard to these lenders, subject to approval from the Minister of Finance.¹⁵  We do not currently intend to do this but will monitor the lending of non-deposit taking lenders.

Calibration

LVR restrictions for residential mortgage lending are set with a ‘threshold’ and a ‘speed limit’. The threshold specifies the LVR where loans are considered high-LVR. The speed limit restricts the proportion of a banks’ new residential mortgage lending that can have a high LVR outside of certain exemptions (see Box A).

There are different LVR settings for owner-occupier and investor loans. Investor loans typically have a lower threshold and lower speed limit for high-LVR lending compared to owner-occupier loans. Tighter restrictions for investors reflect the higher risks associated with this type of lending.¹⁷

Investor behaviour is more likely to amplify financial and housing market cycles. Compared to owner-occupiers, investors are more likely to respond to a house price downturn by selling some of their investment properties to reduce leverage. This increases the risk of creating widespread distressed house sales, which could further increase the number of investors that fall into negative equity and increase potential losses for banks. Tighter LVR restrictions for investors means there is a lower proportion of high-LVR investor loans, helping to mitigate this risk.

LVR restrictions may also become less binding on investors over time as these borrowers can leverage existing equity to acquire more properties, potentially amplifying house price cycles. Tighter LVR restrictions for investors limit highly leveraged borrowing and may help to dampen house price cycles to reduce the risk of a substantial downturn over time.

Maintaining LVR restrictions through the financial cycle helps to mitigate procyclicality and borrower vulnerability to house price declines. Therefore, we have a long-run setting for LVR restrictions, which is the default level we expect them to be set at most of the time. The long-run high-LVR threshold is currently set at 80% for owner-occupiers and 70% for investors, with speed limits of 25% for owner-occupiers and 10% for investors.¹⁸  These settings aim to support financial stability by building household resilience over time without unduly restricting access to credit.

While long-run LVR settings generally improve borrower resilience over time, we may still tighten (or possibly loosen) settings from the long-run level in some situations to manage systemic risk. When the Lending Standard comes into force, it will specify a relatively wide range of possible settings for the high-LVR threshold and lending speed limit, giving us flexibility to respond to evolving financial stability risks.¹⁹  If it was necessary to adjust LVR restrictions, we would typically look to move them back to the long-run level once risks had normalised. The Decision-Making Process section discusses how we assess risks that may motivate a change in policy settings.

Debt-to-income restrictions

DTI restrictions limit the amount of residential mortgage lending that banks can provide to borrowers with a DTI above a certain threshold. A borrower’s DTI is calculated by dividing their total debt by total annual income, subject to certain exclusions and assumptions specified by the Reserve Bank. We introduced DTI restrictions in 2024.

Transmission to financial stability

LVR restrictions target one aspect of systemic risk related to the housing market. However, another key component is a borrower’s capacity to service a loan, which affects their probability of default. DTI restrictions address this second dimension of housing-related systemic risk, complementing LVR restrictions.

In a boom, borrowers could contribute to the cyclicality of the financial cycle by borrowing large amounts relative to their income and bidding up house prices. Borrowers can also contribute to cyclicality in a downturn if they come under financial stress and default on their mortgage payments, potentially forcing them to sell. Systemic ‘waves’ of mortgage defaults could lead to many borrowers having to sell within a short period of time, forcing sales at ‘fire sale’ prices. This can result in significant losses to deposit takers and a reduction in credit availability, creating further economic and financial instability through negative feedback loops.²⁰

DTI restrictions limit the amount of high-DTI mortgage lending by banks which, over time, reduces the number of highly indebted households. Lower debt levels relative to income supports household resilience in a downturn and therefore reduces the probability that borrowers will face stress or default on their mortgages. This reduces the probability of a systemic wave of mortgage defaults in a downturn, supporting financial and economic stability.

DTI restrictions can also help to moderate the cyclicality of the housing market. The restrictions tie mortgage borrowing capacity directly to borrower income, which fluctuate much less than house prices on average. Limits on high-DTI lending can help to moderate house price growth during booms, reducing both the probability and potential magnitude of housing market corrections. This reduces the probability of negative feedback effects from a housing market downturn on financial markets and the broader economy.

Calibration

Like LVR restrictions, DTI restrictions are also set with a threshold and speed limit. The threshold specifies the DTI ratio where loans are considered high-DTI. The speed limit restricts the proportion of a banks’ new residential mortgage lending that can have a high DTI ratio.

DTI restrictions apply to both owner-occupiers and investors, but not necessarily at the same level for these borrower types. Setting different DTI restrictions allows us to calibrate them proportionately to the risks that each group represent.

For owner-occupiers, DTI restrictions lower the probability of borrowers entering financial stress and defaulting on their mortgages, therefore reducing the probability of a systemic wave of mortgage defaults. In a downturn, owner-occupiers are unlikely to sell their property as it is their primary residence and are more likely to reduce spending and potentially default on their mortgages as their financial situation deteriorates. This dynamic becomes more pronounced as the downturn persists and unemployment rises and incomes fall for some borrowers.

For investors, DTI restrictions aim to reduce cyclicality in the housing market that can be created if investors respond to a downturn by selling some of their investment properties, further pushing house prices down. However, financial stress also generally occurs at a higher level for investors compared to owner-occupiers. Investors tend to have more income available to service loans as they receive rental income in addition to labour income, with a smaller share of this additional income needed to cover essential expenses. Therefore, investors are typically better placed to service high-DTI loans, which means that the high-DTI threshold for investors can be set higher relative to owner-occupiers to achieve similar financial stability benefits.

DTI restrictions are counter-cyclical in nature as the proportion of high-DTI lending tends to increase in boom periods, such as when house prices are rising and low interest rates support borrowing capacity (and vice versa). This means that DTI restrictions can be set at a level where they constrain lending only when conditions mean high-DTI lending would normally be elevated. They would have limited impact at other times, without needing to be adjusted. This is known as the ‘guardrail’ approach.

The figure below illustrates how DTI restrictions act as guardrails. When DTI restrictions were introduced in 2024, the threshold was set at 6 for owner-occupiers and 7 for investors, with a 20% speed limit for both borrower types. These are also the long-run DTI settings. DTI restrictions were not binding at the time as the housing market was relatively subdued. However, they would have been binding during the house price boom in 2021, had they been in place then.

Counter-cyclical capital buffer

The CCyB is a capital buffer that forms part of deposit takers’ overall capital requirements. The key difference between the CCyB and other components of the capital stack is that the CCyB is designed to be adjusted during the financial cycle in response to changes in systemic risk. At this stage, it is intended that the CCyB will take effect from December 2028.²³

Transmission to financial stability

The CCyB is designed to be partly or fully reduced (on a temporary basis) as the financial system enters a period of stress and banks’ capital ratios deteriorate. This reduces the risk of regulatory capital requirements constraining credit availability in a significant economic downturn. For example, when the CCyB is reduced, banks can use this ‘released’ capital to help absorb losses without breaching regulatory capital requirements. This encourages banks to continue lending to creditworthy borrowers without needing to raise additional capital. Therefore, easing the CCyB requirement in a downturn reduces the likelihood of the banking system amplifying the downturn through a broader tightening in credit conditions.

Another potential role of the CCyB is to mitigate procyclicality by imposing higher capital requirements when systemic risks are elevated. This helps to build the financial system’s resilience to a future downturn and can help constrain excess credit growth. However, it is unlikely that we would use the CCyB for this purpose given we have more targeted tools to address the main sources of systemic risk in New Zealand.

Calibration

We will use an ‘early-set’ approach for the CCyB, where the buffer is set before systemic risks begin to emerge. The CCyB will be set at a positive long-run rate during normal times and is only intended to be reduced in periods of systemic stress. The CCyB would then be raised back to the long-run level as soon as is appropriate following the systemic stress event. The next section discusses our processes for deciding to reduce or restore the CCyB in more detail.

As part of the 2025 Review of key capital settings, we decided to set the long-run CCyB rate at 1% of risk-weighted assets.²⁴  It is unlikely that we would raise the CCyB above the long-run rate.

The early-set approach to the CCyB is important for ensuring the additional capital requirements are in place when risks begin to emerge. It takes time for banks to increase their capital ratios as the CCyB increases, but decisions to reduce the CCyB can come into effect immediately. Requiring banks to quickly increase capital ratios as systemic risk begins to rise could unduly constrain credit availability if banks cannot obtain new capital quickly enough. It could also mean that banks do not have enough time to sufficiently increase their capital buffers before a downturn occurs. Having a positive CCyB in place through the financial cycle increases the likelihood that it is effective at supporting credit availability in a downturn and reduces the risk of unintended consequences.

Sectoral capital requirement

The SCR is an additional capital buffer that we can apply to specific sectors or segments of sectors where we consider that excessive credit growth is leading to a build-up of systemic risk. The SCR could either be applied through overlays to sectoral risk weights or as an additional capital requirement that is calibrated as a proportion of deposit takers’ risk-weighted assets for a particular sector. It would be removed when sectoral risks normalise, or as the economy enters a downturn. We have not used the SCR to date.

Transmission to financial stability

The SCR works similarly to the CCyB but targets systemic risks from a specific sector. As with the CCyB, the SCR increases the amount of capital that banks have to absorb losses from a particular sector in a downturn. The benefit of a SCR overlay would be immediate, in contrast to borrower-based tools that take time to build resilience.

In addition, the SCR could make lending to specific sectors relatively less attractive as banks need to hold relatively more capital against these exposures. Banks might decide to moderate their lending to the targeted sector if faced with a higher cost of funding, mitigating the build-up in systemic risk that is coming from excessive credit growth in that sector. Banks could also pass on these higher funding costs to borrowers, helping to reduce demand for credit in the specific sector.

Calibration

While the SCR is part of our macroprudential toolkit, we have not yet used it, and we do not currently intend to activate it in the future.²⁵  As such, the SCR is currently set at 0%. However, we retain the flexibility to activate the SCR if necessary. We would consult on the use of this instrument before it is implemented.

Adjustments to the core funding ratio

The CFR imposes a minimum requirement on the percentage of a banks’ overall lending that is funded by retail deposits, long-term wholesale funding or capital. These funding sources are relatively more stable than short-term wholesale funding. A bank’s CFR is calculated by dividing its stable funding by its total lending.

Maintaining a strong requirement for stable funding is important for reducing banks’ vulnerability to funding market disruptions and ensuring they can fund their lending on an ongoing basis. Banks are usually required to maintain a CFR of at least 75%. However, this requirement could be adjusted through the financial cycle in response to emerging systemic risks or a deterioration in market conditions. For example, we reduced the CFR requirement between 2020 and 2022 during the period of uncertainty from the COVID-19 pandemic.

Transmission to financial stability

When systemic risk is elevated, increasing the CFR could improve financial system resilience by further increasing the stability of bank funding to offset any increase in the risk of a funding market disruption. A higher CFR requirement could also restrain excessive credit growth if banks choose to increase their CFRs by moderating their lending, helping to mitigate procyclicality.

Easing the CFR requirement can support financial stability if there is a significant deterioration in external funding market conditions, and there is a risk that banks would otherwise need to reduce lending to continue to meet this requirement. This would support credit availability through the period of systemic stress. The CFR requirement could then be restored to its usual level once funding market conditions had stabilised to rebuild banks’ resilience to disruptions.

Calibration

Similar to the SCR, we only intend to adjust the CFR in exceptional circumstances, such as during the COVID-19 pandemic. For most of the time, the CFR will be set at 75% as per the Reserve Bank’s prudential liquidity requirements.

5. Interactions between policy tools

Figure 4 summarises the different ways in which LVR restrictions, DTI restrictions and the CCyB impact financial stability. As these tools target different aspects of systemic risk, we think holistically about how tools work together to best achieve our statutory purposes. International experience suggests that using a combination of tools is more effective than using them individually due to synergies between tools (Lo Duca, et al., 2023)

Interactions between LVR and DTI restrictions

There are separate thresholds and speed limits for each of the LVR and DTI restrictions. For example, if a borrower is above both the LVR and DTI thresholds, the lending counts towards the speed limits for both policies. If the borrower is above the LVR but not the DTI threshold, it will count towards the LVR speed limit but not the DTI speed limit (and vice versa).

LVR and DTI restrictions typically affect different groups of mortgage borrowers, as lenders are generally reluctant to approve loans that have both a high DTI ratio and a high LVR (McDonald, 2025). High-LVR loans tend to have lower DTI ratios due to stricter affordability assessments. Conversely, high-DTI loans are usually only approved for borrowers with sufficient equity and lower LVRs. Therefore, keeping the restrictions separate provides flexibility to target risks independently through the cycle.

LVR and DTI restrictions also complement each other as they tend to bind in different situations. The LVRs of existing homeowners fall as house prices increase causing LVR restrictions to become less binding. In contrast, DTI restrictions link borrowing to incomes which can help to maintain more stable lending conditions even when house prices rise.

A benefit of having both DTI and LVR restrictions is that together the tools can be set at a looser level than either tool individually to achieve the same or better financial stability benefits, while also reducing unintended consequences. For example, with only LVR restrictions in place, tighter LVR settings may be needed to also capture higher-risk high-DTI loans. However, these restrictions would also capture some low-DTI loans, which may unduly restrict access to credit for these borrowers. By having DTI restrictions in place as well, high-DTI loans can be restricted directly, allowing for slightly looser LVR settings.

We would also expect to tighten LVR settings from their long-run level less often with DTI restrictions in place. The guardrail approach to setting DTI restrictions means they will automatically bind in housing upswings, helping to stabilise credit growth and reducing the need for more frequent adjustments to LVR settings.

Interactions with the CCyB

The CCyB complements LVR and DTI restrictions by allowing a temporary loosening of capital requirements in economic downturns. LVR and DTI restrictions are more effective at preventing a build-up of systemic risk before a downturn occurs. Easing these restrictions in a downturn to support credit availability could result in additional systemic risk by allowing for a greater proportion of highly indebted borrowers. Instead, easing the CCyB would more directly target the supply of credit, particularly if capital requirements were deemed to be constraining lending. We discuss this further in the Decision-Making Process section.

We would consider tightening settings as risks increase

Based on our assessment of these four vulnerabilities, we consider if overall risks are building sufficiently to warrant tighter settings. We expect to generally focus on LVR restrictions (and where they sit relative to our long-run assumption). However, at times, we may also consider adjusting DTI restrictions, especially as we learn more about their impact. If risks were elevated in a specific sector, we could consider activating the SCR, but we expect that this would be relatively unlikely.

Conditions for tightening settings

The risk of a house price correction and the riskiness of new lending are key indicators. Two notable situations where tighter settings could be warranted are:

• If house prices were well above our sustainable estimates, like they were in 2007 and 2021. Based on our current estimates, house prices were 28% above the mean of our sustainable indicators in 2007 and 30% above in 2021. While we would not necessarily wait for house prices to get this far above sustainable levels before responding (partly because estimates at the time may not be as large), this gives an appreciation for how much house prices can temporarily exceed our sustainable estimate. Given this context, a deviation of less than 10% above the mean is unlikely to be a major concern on its own.
• If household credit growth was high, banks were using most of their speed limits for high-LVR and high-DTI lending, and there was a build-up of lending just below the high-LVR and high-DTI thresholds. Annual mortgage credit growth exceeded 10% from 2003 until 2008, a period of rapid growth in household debt. Since then, it has only exceeded 10% briefly in 2021. We would consider tighter settings if similar situations were expected to develop. Our response would depend on the extent this growth was underpinned by high-LVR and/or high-DTI lending.

In any situation where we considered tightening, we would also consider our assessment of the vulnerabilities amongst indebted households and the resilience of lenders. Unless we see one of the situations above (or aspects of them), we expect vulnerabilities amongst existing household borrowers or developments in the resilience of lenders to rarely warrant changes in settings on their own.³³

Restoring LVR restrictions to their long-run levels

When risks are no longer particularly elevated, we would look to return LVR settings back to their long-run level. We would use the same assessment criteria and indicators as we did for tightening, including for example whether house prices are significantly above what is sustainable.

Easing settings may be gradual. For example, in 2023 and 2024 we eased LVR settings in steps. For owner occupiers, we increased the speed limit applied to new lending above an LVR of 80% from 10% to 15% in 2023. The following year we increased the speed limit again to 20%. The exception to this gradual approach is in an economic downturn, when we may return LVR settings to their long-run level more quickly, but this would depend on the situation.

Easing the CCyB is our preferred response to an economic downturn

An easing of macroprudential settings may be beneficial to support the provision of credit during a significant economic downturn. Easing the CCyB is best suited to this because it supports lending to both households and businesses, meaning it has broader benefits than easing LVR or DTI restrictions. We could also consider easing the CFR requirement if there was a disruption in funding markets.

We are wary about easing LVR or DTI restrictions during a period of systemic stress to looser settings than the long-run level. These tools are designed to prevent a build-up of systemic risk before a downturn occurs, but are less effective in response to downturns to support the economy. There may be a lack of demand for higher risk lending and banks may not want to lend in this situation. We also may not want to promote risky lending in case conditions change quickly. As highlighted by our experience easing LVR restrictions at the start of the COVID-19 pandemic, the housing market can be unpredictable, and it takes time to reinstate LVR restrictions.³⁴

One situation where we would consider easing LVR or DTI restrictions is if they were tighter than their long-run level, to bring them back to their long-run level. If easier settings than the long-run level were thought necessary, we would prefer to ease the restrictions rather than fully removing them. Another option is to remove restrictions for a specific period, which we would state publicly, then we could extend this if the downturn persists.

Conditions for reducing the CCyB

The purpose of easing the CCyB in a downturn is to reduce the likelihood that banks’ capital positions constrain lending, which could make the downturn worse. Given this, easing the CCyB would generally be appropriate if both:

• Capital ratios across the banking system as a whole were declining or at a heightened risk of declining materially (that is, to the point where capital buffers no longer exceeded the required levels).³⁵  To assess this in a timely manner, we would be forward looking and likely review settings between our annual reviews. This means in the early stages of a recession we would likely use projections of economic activity along with credit risk models and previous stress test results to estimate (approximately) how much capital could fall.³⁶  We could make use of the Reserve Bank’s Monetary Policy Statement projections. We would also need to consider other policy responses, such as restrictions on dividends (which we put in place during the COVID-19 pandemic).
• Credit conditions had materially tightened or were expected to do so. Tighter credit conditions may be reflected in more stringent lending policies by banks or a reduction in lending appetite across sectors. In addition to industry outreach, we would use our Credit Conditions Survey to get information from banks on expected credit availability and lending policies. If needed, we could run additional rounds of this survey outside of the usual six-monthly rhythm. This would give us the banks’ perspectives of any changes or expected changes in the supply of credit. We would also monitor credit growth across sectors, although this can be lagging and would also be influenced by demand for credit.

Once we decide to lower the CCyB, we would reduce it immediately. We would reduce it by either 0.5 or 1 percentage point depending on the severity of the situation.

As an example, it is likely we would have reduced the CCyB during the early stages of the COVID-19 pandemic, had it been in place then. There was broad agreement across the public and private sector that the economic impact would be severe. Given this, we would likely have assessed that capital requirements would become binding and expected a worsening in credit conditions if those scenarios had played out as predicted. Many countries that had the CCyB already in place (such as in Europe) eased it at the start of 2020 to support lending to households and businesses.

Restoring the CCyB to its long-run level

We intend to return the CCyB to its long-run level (1%) at an early opportunity after the worst of the period of systemic stress has abated. We would want to be confident the economy is recovering - for example, by having seen some increase in economic activity and by considering projections for economic activity.

We would only restore the buffer if we were confident it would not prolong the downturn – for example, by adversely affecting access to credit. For this to be the case, banks would need to have the ability to raise capital or have enough capital already to meet the increase. We would assess this by considering bank capital ratios at the time, as well as their profitability and market conditions for raising capital.

We would announce increases in the CCyB ahead of time to give banks time to raise capital. For example, we would likely give banks around a year to increase it back to the 1% level if starting from zero. The rate of increase will depend on conditions at hand, and we could adjust in smaller increments (such as 0.25 or 0.5 percentage points) if necessary to avoid prolonging the downturn.

If the CFR requirement was eased, we would only restore it to its usual level once we were confident that banks could again source stable long-term funding on reasonable terms. We would also need to be confident that there was already sufficient credit availability for credit-worthy borrowers and that increasing the CFR requirement would not negatively impact this. As with the CCyB, we would give banks time to transition back to the usual requirement.

See Appendix C for more information on how we expect to set macroprudential policy in different parts of the cycle and how this may interact with monetary policy settings.

7. Consultation and stakeholder engagement

Macroprudential policy is designed to promote financial stability, but as highlighted in box A it can also have unintended impacts. Part of the purpose of publishing this framework document is to be transparent about how we use macroprudential tools and how we consider these trade-offs. Another way we ensure transparency is through public consultation, cost-benefit analysis and stakeholder engagement.

Public consultation on macroprudential policy

Public consultation includes publishing an explanation of the policy decision and giving people outside the Reserve Bank an opportunity to submit their views. This is good for transparency and for getting feedback.

However, there is a trade-off to consider. For some policy changes, quick adjustments can help to ensure settings are right for the situation at the time. Producing the material and allowing time for public consultations takes time (and resources both for us and our stakeholders).

Delays with adjustments to settings can make it harder to meet the policy objectives. For example, if housing-related risks build quickly then tighter LVR settings can help to constrain high-risk lending and prevent a cohort of risky borrowers, which in some situations could subsequently contribute to housing volatility and banking sector losses.

We saw this around the end of 2020 and the start of 2021 when it took around 6 months to reinstate LVR restrictions. These restrictions had been removed during the onset of the pandemic. By the end of 2020, the housing market was booming and risky lending had increased.³⁷  The delay in reinstating restrictions meant more high-risk lending occurred than we would have liked.

A quick response is not always important. Changes to our assumption of long-run LVR and DTI settings can be done without urgency. A public consultation would be important if a material tightening in one of these tools was being considered. This would give us an opportunity to explain and get feedback on the benefits of tighter policy settings and the trade-off with any unintended impacts. A public consultation is less beneficial for immaterial changes and when we ease settings.

Therefore, our view is that we will publicly consult when introducing new tools and for non-urgent tightening of settings.

To implement changes to LVR and DTI settings, previously we have made changes to banks’ conditions of registration and in the future will need to make changes to banks’ licence conditions. For these changes we consult with banks for a short period. This is largely an operational process, focusing on how the change is implemented rather than policy decision itself. If we were tightening policy, we would allow time for banks to manage their existing pre-approvals before the changes take effect. We would also engage with banks prior to reinstating the CCyB to collect information about any unintended consequences from increasing their capital requirements.

Regulatory impact assessment

When considering changes in macroprudential policy settings, we consider if the benefits of that change outweigh the costs and other potential unintended consequences. This includes considering if other regulatory responses could better meet our objectives in the situation at hand.

We carry out and publish regulatory impact assessments in line with the requirements as per our legislation.³⁸  This provides further transparency around our decisions, including in cases where the urgency of the situation means we do not consult publicly before implementing a change in settings.

Engagement with the Minister of Finance and Treasury

While decision making on macroprudential policies is delegated to the FPC by the Reserve Bank Board, we inform the Minister of Finance and Treasury prior to macroprudential policy announcements. This is in line with the MoU with the Minister of Finance and helps support co-ordination between our response and any broader government response to the situation at hand. We intend to continue informing the Minister of Finance in this way when the DTA standards come into effect in December 2028.

Our view is that after having introduced DTI restrictions in 2024, we have all the macroprudential tools we need for now. However, if over time we form the view that there would be benefits from implementing a new tool, we would need to amend our regulatory requirements and would consult the public and other authorities accordingly. Moreover, if we wanted to apply lending restrictions to a different sector other than housing, this would also require approval from the Minister of Finance.