Region

Jurisdiction

Regulator

2026-02-26

2025 Review of Deposit Taker Capital Settings: Summary of Submissions and Policy Decisions

The Reserve Bank of New Zealand has eased prudential capital requirements for deposit takers following a targeted 2025 review that recalibrated its risk appetite to support economic prosperity. The policy changes include the removal of Additional Tier 1 capital instruments, the introduction of loss absorbing capacity requirements for the largest institutions, and the implementation of more granular standardised risk weights. These revised settings are scheduled for phased implementation starting in October 2026, with full compliance under the new Capital Standard required by December 2028.

New Zealand

Reserve Bank of New Zealand

IN CONFIDENCE IN CONFIDENCE 2025 Review of deposit taker capital settings Summary of Submissions and Policy Decisions 27 February 2026

2025 Capital Review Summary of Submissions and Policy Decisions 1 Introduction Deposit taker capital requirements set out the minimum amount and type of capital deposit takers must hold. These requirements are designed to promote and protect the stability of the financial system, ensuring the safety and soundness of deposit takers and promoting public confidence in the system. Capital requirements that are set too low risk deposit takers being unable to absorb unexpected losses when they arise and therefore failing – which may impose a direct fiscal cost on New Zealand taxpayers, as well as reducing the long-run prosperity and wellbeing of New Zealanders. Conversely, capital requirements that are too high can reduce credit availability and increase costs unnecessarily. In 2019, we completed a multi-year review of New Zealand’s capital framework (‘the 2019 Review’). This resulted in the decision to significantly increase the quantity and improve the quality of capital that banks are required to operate with to improve the resilience of the banking system. Since then, we’ve begun the process of moving to a new suite of prudential standards under the Deposit Takers Act 2023 (DTA). 1 As well as operating under an updated legislative framework, a new Financial Policy Remit (FPR) was issued in 2024, which places a greater focus on efficiency and competition. Additionally, respondents to recent inquiries have expressed concerns that our bank capital settings may be unreasonably conservative – undermining competition and development of the New Zealand economy. As a good financial system steward, it is important that we consider recent developments and any new evidence presented as part of recent inquiries. Therefore, in March 2025 we announced a targeted review to test whether we have got our capital settings right.2 We consulted on options for change between August and October 2025.3 We want capital settings that support a sustainable economy that ultimately promotes the prosperity and wellbeing of all New Zealanders. Following consideration of new evidence, consultation paper submissions and independent reports, we have recalibrated our risk appetite, leading to an easing of requirements overall. Our new settings also place a slightly greater emphasis on our ability to resolve an entity in distress, which allows a relaxation of going concern capital requirements. Our Decision Document, 4 published on 17 December 2025, contains an explanation of the decisions, including an overview of submissions, high-level implementation plan, risk appetite framework, and an overview of our cost benefit analysis. A summary of the policy decisions can be found in Appendix A of the Decision Document. This document provides more detail on the decisions. Chapter 1 provides an overview and more details on transitional arrangements. Chapter 2 contains a detailed response to submissions received. Please note that this document focuses on responses to matters in scope of the 2025 review of key capital settings for deposit takers (‘the 2025 Review’) and we have not responded to all comments that are out of scope.

1 Reserve Bank of New Zealand. (2025). 2017-2019 Capital Review. https://www.rbnz.govt.nz/regulation-andsupervision/oversight-ofbanks/how-we-regulate-and-supervise-banks/our-policy-work-for-bank-oversight/capital-review 2 2025 Review of key capital settings: Terms of Reference. https://www.rbnz.govt.nz/-/media/project/sites/rbnz/files/regulation-andsupervision/banks/capital-review/2025/2025-review-of-key-capital-settings-terms-of-reference.pdf 3 Policy proposals for feedback. https://consultations.rbnz.govt.nz/prudential-policy/review-of-key-capitalsettings/user_uploads/consultation-paper-review-of-key-capital-settings.pdf 4 Decision document of 2025 Review of key capital settings. https://www.rbnz.govt.nz/-/media/project/sites/rbnz/files/regulation-andsupervision/banks/capital-review/2025/20251212-decision-document-formatted-version.pdf

2025 Capital Review Summary of Submissions and Policy Decisions 2 Appendix A includes a glossary of technical and defined terms.

2025 Capital Review Summary of Submissions and Policy Decisions 3 Contents Introduction ___________________________________________________________________________________ 1 1.1 Overview_________________________________________________________________________________ 5 1.2 Transitional arrangements ______________________________________________________________ 6 2.1 Topics consulted on ____________________________________________________________________ 14 2.1.1 Assessment criteria and risk appetite 14 2.1.2 Context 18 2.1.3 Capital stack options 20 2.1.4 Additional Tier 1 capital 25 2.1.5 Standardised risk weights 31 2.1.6 Counter-cyclical capital buffer (CCyB) 58 2.2 Other topics raised _____________________________________________________________________60 2.2.1 Design of loss absorbing capacity (LAC) 61 2.2.2 Alignment with Australia 64 2.2.3 Output floor and scalar 65 2.2.4 Leverage ratio 66 2.2.5 Cost Benefit Analysis 67 2.2.6 DCS Levy 71 Appendix A: Glossary of technical terms ____________________________________________________73 Appendix B: Consolidated questions ________________________________________________________ 81

2025 Capital Review Summary of Submissions and Policy Decisions 4 1 Chapter 1 Capital settings for deposit takers Overview and Transitional Arrangements

2025 Capital Review Summary of Submissions and Policy Decisions 5 1.1 Overview The 2025 Review began in late March 2025 with a targeted Terms of Reference to enable decisions by the end of 2025 We consulted on the Capital Standard, as part of consultation on the four core standards of the DTA, from May to August 2024. On 27 March 2025, the Reserve Bank committed to undertaking a reassessment of regulatory capital settings for deposit takers. The 2025 Review’s Terms of Reference5 set out the purpose of the review of key capital settings, which is to assess whether the Reserve Bank’s prudential capital requirements for deposit takers are set at the appropriate level to support a stable financial system – one where resilient financial markets, institutions and infrastructures enable a productive and sustainable economy and ultimately promote the prosperity and wellbeing of all New Zealanders. To achieve the purpose by the end of 2025, the scope of the review was targeted to some specific issues. We have also attempted to address out-of-scope issues where practicable within the delivery deadline for the 2025 Review. We received 43 submissions in response to our consultation paper We consulted on options for revised capital settings for deposit takers from 25 August to 3 October 2025. Proposals consulted on included:  Introducing more granular risk weights.  Removing Additional Tier 1 (AT1) capital and partially replacing it with a mixture of Common Equity Tier 1 (CET1) and Tier 2 capital.  Two options – one without loss absorbing capacity (LAC) instruments (Option 1) and one with LAC (Option 2), with both options expected to result in lower average funding costs for deposit takers than under the outcomes that resulted from the 2019 Review. We also engaged with stakeholders during the consultation period and as we analysed submissions. This included a public webinar, facilitating deposit takers’ meetings with the independent international experts, and bilateral meetings with deposit takers and other stakeholders. The 2025 Review included two avenues of independent challenge to our advice The Terms of Reference set out the role of the Treasury and independent international experts in the 2025 Review:  The Treasury played a role in challenging the Reserve Bank’s analysis. We worked closely with the Treasury to update them on our thinking as part of the 2025 Review, and to consider their feedback on our analysis. The Treasury was supportive of our approach to the 2025 Review and process we followed.

5 2025 Review of key capital settings. https://www.rbnz.govt.nz/-/media/project/sites/rbnz/files/regulation-andsupervision/banks/capital-review/2025/2025-review-of-key-capital-settings-terms-of-reference.pdf

2025 Capital Review Summary of Submissions and Policy Decisions 6  Three international experts reviewed our options and recommendations to provide rigorous and independent perspectives as part of the decision-making process. Their final reports can be found on our website. 6 We decided to revise our capital settings Our new settings reflect a revised risk appetite and place a slightly greater emphasis on our tools and ability to resolve an entity in distress, allowing a relaxation of going concern capital requirements (as per Option 2 in our August consultation paper). The key features of revised settings are:  For the largest deposit takers (Group 17 ), a reduction in Tier 1 capital ratio requirements relative to the 2028 settings under the 2019 Review decisions, as well as a reduction in comparison with existing capital requirements, alongside the introduction of a requirement for additional LAC instruments.  For Group 2, a reduction in both Tier 1 and total capital ratio requirements relative to the 2028 settings under the 2019 Review decisions.  For Group 3, a reduction in Tier 1 capital ratio requirements relative to the DTA capital standard policy consultation (offset by an equal increase in Tier 2 capital).  More granular standardised risk weights, including larger reductions for some categories than we proposed in August 2025, reflecting our assessment of the evidence provided in submissions.  Removal of AT1 capital instruments. A summary of the policy decisions can be found in Appendix A of our Decision Document. 8 1.2 Transitional arrangements The decisions will apply to current regulated entities (banks and non-bank deposit takers (NBDTs)) and future licenced entities (deposit takers). For the decisions that are intended to apply before the end of 2028 (when the Capital Standard will be fully in force): a. We will revise the Banking Prudential Requirements (BPRs) to reflect revised risk weights and update the Conditions of Registration for registered banks to implement all changes in requirements. The BPRs will need a ‘one-off' change for new risk weights, while other changes to Conditions of Registration will be relatively minor annual adjustments (see the transition paths for each group of deposit takers below for more detail). 9

6 https://www.rbnz.govt.nz/regulation-and-supervision/oversight-of-banks/how-we-regulate-and-supervise-banks/our-policy-workfor-bank-oversight/2025-review-of-key-capital-settings 7 Under the RBNZ’s Proportionality Framework for developing standards under the Deposit Takers Act: https://www.rbnz.govt.nz/- /media/project/sites/rbnz/files/regulation-and-supervision/dta-and-dcs/the-proportionality-framework-under-the-dta.pdf 8 https://www.rbnz.govt.nz/-/media/project/sites/rbnz/files/regulation-and-supervision/banks/capital-review/2025/20251212-decisiondocument-formatted-version.pdf 9 We also intend to revise banks’ reporting and disclosure obligations to align with the changes in the BPR.

2025 Capital Review Summary of Submissions and Policy Decisions 7 b. We intend to implement the changes in risk weights for NBDTs through changes to their conditions of licence in 2026. Currently, risk weights for NBDTs are set out in NBDT regulations (a legislative instrument). In the consultation on capital setting proposals between August and October 2025, we stated that we intended to recommend to the Minister of Finance to amend the NBDT regulations. Now that we have confirmed the changes that need to be made, we can confirm that they can be implemented through licence conditions. Using licence conditions reduces delivery risks, including failing to meet proposed implementation timelines. We will consult with relevant entities individually through the implementation process. The updated risk weights are expected to significantly reduce the amount of capital certain NBDTs would be required to hold. To promote a sound financial system, we will also offset some of this reduction by increasing NBDTs’ minimum total capital ratio from 8% to 9% (current requirements will continue to apply to credit rating exempt NBDTs). This aligns with the proposed minimum set out in the options section. Looking forward, we intend to action revised settings through the Capital Standard. This will be done as part of tranche 3 of the broader DTA implementation timeline. 10 A policy consultation on more detailed aspects of LAC requirements will take place between June and September 2026, followed by an exposure draft consultation in February 2027. At this stage it is likely that LAC requirements will be included in the Capital Standard when the standard is issued on 31 May 2027 (rather than being included in a standalone “bail-in” standard, for example). Our intention is for the Capital Standard to commence on 1 December 2028 when a phased transition to new LAC requirements will begin. Table 1: High-level implementation plan Date Milestone H1 2026 February 2026: Full documentation published including summary of submissions with our detailed responses, an updated cost benefit analysis and full implementation schedule. March/April 2026: Short consultation on updated drafting of (BPRs) to give effect to changes for banks ahead of the Capital Standard coming into effect in late 2028. H1 2026: Implementing ‘fast-tracked’ changes to NBDT requirements ahead of the Capital Standard coming into effect in late 2028. June-September 2026: Consultation on the Capital Standard exposure draft and guidance (rules for all deposit takers from late 2028). Policy consultation on crisis preparedness standard and LAC requirements alongside the publication of indicative resolution strategies. H2 2026 1 October 2026: Target date for updated BPRs to come into force, covering new risk weights and first annual step changes in capital ratios. No further AT1 capital instruments can be issued after this date. 1 October 2026: Target date for updated NBDT requirements to come into force, covering new risk weights and change in capital ratios.

10 DTA implementation timeline https://www.rbnz.govt.nz/regulation-and-supervision/deposit-takers-act/dta-timeline

2025 Capital Review Summary of Submissions and Policy Decisions 8 Date Milestone Late-2026: Consultation on any changes relating to the review of key additional topics raised in feedback, such as reverse mortgages and commercial property risk weights. 2027 February-April 2027: Consultation on an exposure draft of LAC requirements. May 2027: Consultation on exposure draft of Crisis Preparedness Standard. 31 May 2027: Final Capital Standard (including the LAC requirements) issued. 1 October 2027: Annual step change in bank capital ratios. 2028 1 December 2028: Capital Standard (including the LAC requirements) commences with phased-implementation, annual step change in capital ratios. Late 2028: Crisis Preparedness Standard issued. 2029 and beyond Transition to new capital ratio requirements and LAC requirements. Mid-2029: Publish the Statement of Approach to Resolution. The tables below show indicative transition paths for all deposit taker groups. Indicative transition path for Group 1 deposit takers In October 2026 we intend to increase the existing prudential capital buffer (PCB) of 5.5% of riskweighted assets (RWAs) by 0.5pp to the steady-state level of 6%. Once the DTA is fully in force, LAC will be introduced and gradually increased. We intend to publicly consult on three implementation options for LAC as part of the policy consultation on LAC requirements in mid2026. Depending on the speed of the phase-in, the transitional period would be complete in either:  December 2033 (slow implementation);  December 2031 (moderate implementation); or  December 2030 (fast implementation). To give Group 1 deposit takers a sense of what their possible implementation pathway looks like, the table below shows an indicative transition path for the moderate implementation plan. Table 2: Indicative transition path (moderate implementation) for Group 1 deposit takers Minimum CET1 Minimum Tier 1 (of which max AT1) Minimum total (of which max Tier 2) Prudential Capital Buffer (PCB) – all CET1 LAC Total capital + LAC Dec 2025 4.5 7 (2.5) 9 (2) 5.5 0 14.5 Oct 2026 4.5 7 (2.5) 9 (2) 6 0 15

2025 Capital Review Summary of Submissions and Policy Decisions 9 Minimum CET1 Minimum Tier 1 (of which max AT1) Minimum total (of which max Tier 2) Prudential Capital Buffer (PCB) – all CET1 LAC Total capital + LAC Oct 2027 4.5 7 (2.5) 9 (2) 6 0 15 Dec 2028 6 6 (2.5) 9 (3) 6 1 16 Dec 2029 6 6 (1.675) 9 (3) 6 2 17 Dec 2030 6 6 (0.825) 9 (3) 6 4 19 Dec 2031 6 6 (0) 9 (3) 6 6 21 Note 1: Once the DTA takes effect in December 2028, the maximum amount of AT1 (issued under the rules currently in place) that a deposit taker can count towards Tier 1 requirements will gradually decline. The exact process for this will be set out in the exposure draft of the Capital Standard. Note 2: A Counter Cyclical Buffer (CCyB) of 1% is planned to come into effect from December 2028. The 6% Prudential Capital Buffer (PCB) includes the 1% long-run CCyB from December 2028. Indicative transition path for Group 2 deposit takers We will gradually increase the PCB, to its steady state level of 5% of RWAs, compared with 3.5% currently, in advance of the DTA commencing in 2028. The table below shows an indicative transition path for Group 2 deposit takers. We do not intend to change the transition path as a result of the policy consultation on LAC requirements in 2026 (given LAC requirements will not apply to Group 2 deposit takers). Table 3: Indicative transition path for Group 2 deposit takers Minimum CET1 Minimum Tier 1 (of which max AT1) Minimum total (of which max Tier 2) Prudential Capital Buffer (PCB) – all CET1 LAC Total capital + LAC Dec 2025 4.5 7 (2.5) 9 (2) 3.5 0 12.5 Oct 2026 4.5 7 (2.5) 9 (2) 3.5 0 12.5 Oct 2027 4.5 7 (2.5) 9 (2) 4 0 13 Dec 2028 6 6 (2.5) 9 (3) 5 0 14 Dec 2029 6 6 (1.675) 9 (3) 5 0 14

2025 Capital Review Summary of Submissions and Policy Decisions 10 Minimum CET1 Minimum Tier 1 (of which max AT1) Minimum total (of which max Tier 2) Prudential Capital Buffer (PCB) – all CET1 LAC Total capital + LAC Dec 2030 6 6 (0.825) 9 (3) 5 0 14 Dec 2031 6 6 (0) 9 (3) 5 0 14 Note 1: Once the DTA takes effect in December 2028, the maximum amount of AT1 (issued under the rules currently in place) that a deposit taker can count towards Tier 1 requirements will gradually decline. The exact process for this will be set out in the exposure draft of the Capital Standard. Note 2: The long-run Counter Cyclical Buffer (CCyB) of 1% is planned to come into effect from December 2028. The 5% Prudential Capital Buffer (PCB) includes the 1% CCyB from December 2028. Indicative transition path for Group 3 deposit takers We will allow a slower transition to higher capital ratios for Group 3 deposit takers, while bringing forward new risk weight changes and allowing “perpetual preference shares” (PPS) to count as Tier 2 capital for Group 3. This gives smaller deposit takers more options for raising Tier 2 capital if desired. The table below shows indicative transition paths for Group 3 (Mutuals) deposit takers and Group 3 (Non-mutuals) deposit takers, respectively. We are targeting October 2026 to bring forward the new risk weight changes and for the step changes in gross capital to begin. This timing is subject to changes in relevant requirements and will be confirmed in due course. We do not intend to change the transition path for Group 3 deposit takers as a result of the policy consultation on LAC requirements in 2026 (given LAC requirements will not apply to Group 3 deposit takers). Table 4: Indicative transition path for Group 3 (Mutuals) deposit takers Minimum CET1 Gross Capital / Minimum Tier 1 (of which max PPS) Minimum total (of which max Tier 2) Prudential Capital Buffer (PCB) – all CET1 LAC Total capital + LAC Dec 2025 - 8 (4) 8 (-) 0 0 8 Oct 2026 - 9 (4.5) 9 (-) 0 0 9 Oct 2027 - 9 (4.5) 9 (-) 0 0 9 Dec 2028 6 6 (3.5) 9 (3) 1 0 10 Dec 2029 6 6 (4) 9 (3) 2 0 11

2025 Capital Review Summary of Submissions and Policy Decisions 11 Minimum CET1 Gross Capital / Minimum Tier 1 (of which max PPS) Minimum total (of which max Tier 2) Prudential Capital Buffer (PCB) – all CET1 LAC Total capital + LAC Dec 2030 6 6 (2.5) 9 (3) 3 0 12 Dec 2031 6 6 (2.5) 9 (3) 4 0 13 Dec 2032 6 6 (0) 9 (3) 4 0 13 Note 1: Gross capital is a concept set out in the Deposit Takers (Credit Ratings, Capital Ratios, and Related Party Exposures) Regulations 2010. Before the DTA comes into force, NBDTs will continue to use this concept to calculate capital. Note 2: Group 3 deposit takers will be able to issue Tier 2 capital from December 2028 onwards. From that point, existing PPS can be counted as either Tier 1 or Tier 2 capital. The fluctuation shown in the level of PPS that can count as Tier 1 capital is the result of gradually increasing the PCB whilst concurrently reducing the amount of PPS that can be used to meet Tier 1 capital requirements. Note 3: Throughout the transition, NBDTs with credit rating exemptions will continue to be required to have higher levels of capital – at least 10% minimum for NBDTs with average consolidated liabilities less than $20 million, and at least 12% for NBDTs with average consolidated liabilities between $20 and $40 million. Table 5: Indicative transition path for Group 3 (Non-Mutuals) deposit takers Minimum CET1 Gross Capital / Minimum Tier 1 (of which max PPS) Minimum total (of which max Tier 2) Prudential Capital Buffer (PCB) – all CET1 LAC Total capital

LAC Dec 2025 - 8 (2) 8 (-) 0 0 8 Oct 2026 - 9 (2.25) 9 (-) 0 0 9 Oct 2027 - 9 (2.25) 9 (-) 0 0 9 Dec 2028 6 6 (3.5) 9 (3) 1 0 10 Dec 2029 6 6 (4) 9 (3) 2 0 11 Dec 2030 6 6 (2.5) 9 (3) 3 0 12 Dec 2031 6 6 (2.5) 9 (3) 4 0 13 Dec 2032 6 6 (0) 9 (3) 4 0 13

2025 Capital Review Summary of Submissions and Policy Decisions 12 Note 1: Gross capital is a concept set out in the Deposit Takers (Credit Ratings, Capital Ratios, and Related Party Exposures) Regulations 2010. Before the DTA comes into force, NBDTs will continue to use this concept to calculate capital. Note 2: Group 3 deposit takers will be able to issue Tier 2 capital from December 2028. From that point, existing PPS can be counted as either Tier 1 or Tier 2 capital. The fluctuation shown in the level of PPS that can count as Tier 1 capital is the result of gradually increasing the PCB whilst concurrently reducing the amount of PPS that can be used to meet Tier 1 capital requirements. Note 3: Throughout the transition, NBDTs with credit rating exemptions will continue to be required to have higher levels of capital. At least 10% minimum for NBDTs with average consolidated liabilities less than $20 million, and at least 12% for NBDTs with average consolidated liabilities between $20 and $40 million. Indicative transition plan for removing Additional Tier 1 (AT1) – all groups Our proposed approach for removing AT1 capital is set out below: 11  We will continue to recognise existing issued AT1 capital under the updated BPRs – though we will not allow any AT1 instruments issued after the updated BPRs come into force (expected to be 1 October 2026) to count as regulatory capital;  Under the Capital Standard, we will allow Group 1 and 2 deposit takers to count remaining AT1 towards their Tier 1 requirements on a reducing basis before complete derecognition;  We will allow Group 3 deposit takers to count a maximum of one-half of their PPS Tier 1 capital (with the rest as Tier 2 capital) for a period of time, with this ratio changing such that, eventually, PPS will only be able to count as Tier 2 capital; and  We will continue to apply our longstanding policy of considering applications for early redemption of AT1 where the terms of those instruments provide for this and the application demonstrates the conditions in our requirements are satisfied.

11 We intend to consult on the transition plan for AT1 as part of the Capital Standard consultation.

2025 Capital Review Summary of Submissions and Policy Decisions 13 2 Chapter 2 Capital settings for deposit takers Summary of Submissions and Responses

2025 Capital Review Summary of Submissions and Policy Decisions 14 Introduction This chapter summarises the feedback received on our August to October 2025 consultation paper on the 2025 Review and provides our responses. The first section covers the topics we consulted on and the second section covers other issues raised in consultation feedback. 2.1 Topics consulted on 2.1.1 Assessment criteria and risk appetite Assessment criteria In the consultation paper, we proposed assessment criteria to articulate the key factors for the review and to help us assess potential options for capital settings. This was based on our statutory parameters including the purposes and principles of the DTA. Financial stability criteria  Going concern loss absorbency: Maintain a sufficient prudential capital buffer above the regulatory minimum to absorb losses, protect and promote the stability of the financial system, and promote the safety and soundness of each deposit taker (links to DTA section 3(1) and 3(2)(a) purposes).  Crisis management: Enable a distressed deposit taker to be dealt with in an orderly manner, recognising the need for a credible resolution strategy for deposit takers to promote financial stability and avoid the use of public funds (links to DTA section 259 purposes). Other criteria  Proportionality: Take a proportionate approach to regulation and supervision (links to DTA section 4(a)(i) and (ii) principles).  Competition: Maintain competition within the deposit-taking sector, recognising the desirability of a diverse deposit-taking sector that provides financial products and services to a diverse range of New Zealanders (links to DTA section 3(2)(c) purpose, and section 4(a) and (b) principles).  Funding costs: Consider the impact on deposit takers’ weighted average funding costs, which in turn affect lending rates, recognising their importance for supporting the prosperity and wellbeing of New Zealanders (links to DTA section 3(1) and 3(2)(d) purposes).  Simplicity/achievability: Be practical to administer, easy to implement and avoid unnecessary compliance costs (links to DTA section 4(c) principle).  International alignment: Align with international standards where appropriate (links to DTA section 4(d) principle). We noted that we would assess options against these criteria, along with the results of our quantitative cost benefit analysis. The ‘right’ capital settings for New Zealand would depend on how we balanced different criteria.

2025 Capital Review Summary of Submissions and Policy Decisions 15 We sought feedback on our proposed assessment criteria during the consultation. Consultation feedback Respondents were generally supportive of the proposed assessment criteria. Most submissions emphasised the importance of specific criteria such as proportionality and competition, particularly in relation to smaller deposit takers, as well as international alignment. There were also some specific issues raised for us to consider within these criteria. Some feedback recommended amending the proposed criteria to strengthen specific considerations. Respondents sought greater international alignment, including with the Australian Prudential Regulation Authority (APRA) approach, by specifically considering differences in our framework compared to other jurisdictions. Respondents also sought greater emphasis on the funding cost criteria. Some respondents proposed additional criteria. One recommended including a criterion to evaluate options against the Finance and Expenditure Committee's (FEC) recommendations in the final report on the inquiry into banking competition. A few respondents noted the importance of efficiency and ensuring New Zealand’s overall capital requirements accurately reflect risk. Comment We acknowledge feedback that some criteria could be amended or additional criteria included. We developed the assessment criteria for the 2025 Review to reflect the range of factors we must consider when setting prudential requirements, in a way that is consistent with our purposes and principles under the DTA. Given the legislative basis for the assessment criteria, we did not think that it was appropriate to explicitly weight some factors more heavily than others. Our approach allowed us to appropriately balance competing criteria. We do not think that it is appropriate to fully align New Zealand’s capital requirements with international approaches, including APRA’s. This is because there are important differences in our regulatory frameworks, economies and banking systems that would mean full alignment would not meet our financial stability objectives and would be outside of the Reserve Bank Board’s risk appetite. This point is discussed further in the context of APRA’s requirements in section 2.2.2. We considered submissions to the FEC inquiry, and the recommendations made by FEC in developing the consultation paper and our final decisions. To ensure that New Zealand’s capital requirements are efficient, we assessed potential options against a range of factors included in the ‘other criteria’ as set out in our consultation paper, including proportionality, competition, funding costs, simplicity and international alignment and considered the results of our quantitative cost benefit analysis. We also considered whether capital requirements accurately reflect risk through other parts of this review, including by benchmarking New Zealand’s capital requirements against comparable jurisdictions and reassessing some risk weights. Response We did not change the assessment criteria following the consultation. The criteria we consulted on were consistent with our statutory parameters and allowed us to balance competing considerations. We did not think that the additional criteria suggested by some respondents needed to be specifically included as they were already considered in the review process.

2025 Capital Review Summary of Submissions and Policy Decisions 16 Risk appetite Capital requirements can impact the frequency and severity of banking system stress events in New Zealand, so how we calibrate requirements depends on our ‘risk appetite’. This is important for how we balance the benefits of higher capital, in terms of improved financial stability, against the potential costs, such as foregone economic output, when we set requirements. The 2019 Review decisions were calibrated so there was an extremely low likelihood of systemic crisis. We expressed our risk appetite using a tolerance for a “1-in-200-years” systemic crisis. This meant settings were calibrated so there was enough capital in the system to avoid being fully depleted to the point where there was no capital at all in a severe economic shock equivalent to what could be expected once every 200 years. In the consultation paper we stated that we now have a higher risk appetite than in 2019 due to the introduction of the DTA and strengthening of our broader prudential requirements. The Minister of Finance issued a new Financial Policy Remit (FPR) in December 2024 that emphasises the importance of improving competition in the financial sector and accepts that failures of individual institutions will happen occasionally. The options in the consultation paper would deliver a low (but not extremely low) likelihood of failure. We did not communicate this using a “1-in-X-years” approach and instead stated that we had focused on benchmarking our capital requirements against other countries. We sought feedback on the appropriate risk appetite for New Zealand’s capital settings. Consultation feedback We received 19 submissions that commented on risk appetite from deposit takers, industry groups, non-financial groups, and individuals. Feedback was largely supportive of our overall direction. Respondents agreed with removing the ‘1-in-X-years” description of our risk appetite and instead benchmarking New Zealand’s capital ratios against other countries. Deposit takers and industry groups agreed that a higher risk appetite compared to 2019 was appropriate, but some did not think that this was reflected in our proposals. Only one respondent disagreed with a higher risk appetite compared to 2019, citing financial stability concerns. Another respondent pointed out that we adopted the principle that our capital requirements should be conservative in 2019 to offset our relatively light touch approach to supervision and suggested that this point remains the same today. Feedback mostly focused on the need to clearly articulate our risk appetite. Respondents noted that while we stated we were focusing on how New Zealand’s capital requirements compare to other countries, we did not specify what our risk appetite is relative to other countries. One respondent recommended adopting a description like ‘unquestionably strong’ as is used by APRA. Comment Our risk appetite aims to balance the benefits to society of preventing or managing the failure of regulated entities with the costs to society of regulation. In the past, we have prioritised crisis prevention over crisis management by maintaining high CET1 requirements but taking a lighttouch supervisory approach, with limited resolution capacity. The introduction of the DTA has improved our supervisory and regulatory toolkit since the last capital review. This, along with the broader step up in our prudential regime, allows us to transition

2025 Capital Review Summary of Submissions and Policy Decisions 17 to a model that maintains a low (but not extremely low) likelihood of failure while likely reducing the cost of failure through credible recovery and resolution tools. We have carefully considered where to calibrate our rules to align with international norms where possible. There are many unique features of New Zealand that justify setting pillar 1 requirements slightly higher than our peer countries, but without the additional capital add-ons seen in many of these countries. These factors include our status as a small, open and high-risk economy, our highly concentrated banking sector with significant exposure to Australia at the D-SIB level, and our still relatively light-touch supervisory approach despite the DTA. Response The full risk appetite framework is available in Appendix C of the review Decision Document. 12 This sets out our general approach to prudential policy. We have decided to express our risk appetite as set out in our Statement of Prudential Policy. 13 Our policies are:  Simple – focusing on key risks to New Zealand and only adopting international requirements to the extent they are relevant to New Zealand’s financial system.  Strong – we set requirements that minimise the likelihood of failure and rely less on mitigating the cost of failure, should it occur.  Proportionate – requiring relatively higher requirements for large, systemically important entities.  Efficient – we operate a transparent framework, that seeks to minimise compliance costs and recognises the importance of the trans-Tasman regulatory framework. We have a low appetite for events that could materially damage financial stability. However, we do not operate a zero-failure regime. We have a moderate tolerance for risks that may lead to the failure of regulated entities where the impact is understood, manageable, and will not materially damage the financial system. This risk appetite reflects a slightly higher tolerance of failure of individual institutions compared to the 2019 Review, while still emphasising overall financial stability. The new capital requirements align with this by implementing slightly lower buffers than intended under the 2019 decisions and introducing additional LAC requirements to mitigate the risks for D-SIBs. We will continue to monitor our capital settings against this new risk appetite. We also intend to monitor how our settings compare internationally over time.

12 https://www.rbnz.govt.nz/-/media/project/sites/rbnz/files/regulation-and-supervision/banks/capital-review/2025/20251212-decisiondocument-formatted-version.pdf 13 Reserve Bank of New Zealand (2022), Statement of Prudential Policy, https://www.rbnz.govt.nz/- /media/project/sites/rbnz/files/regulation-and-supervision/statements-of-approaches/sopp-2022.pdf

2025 Capital Review Summary of Submissions and Policy Decisions 18 2.1.2 Context Changes since 2019 In the consultation paper we considered that cumulative policy changes since 2019 should lower risks in the New Zealand financial system and may support lower capital requirements. However, we argued New Zealand-specific risk factors were largely unchanged and risks in the broader macroeconomic environment had increased since 2019. We sought feedback on our assessment of these changes and whether there was other new evidence that we should be considering. Consultation feedback One-third of submissions commented on this analysis. Most of these submissions largely agreed with our assessment or expanded on the analysis. There were two specific policy changes that respondents suggested would have larger impacts on reducing risk and the appropriate level of capital than we stated in the consultation paper. Two respondents argued that the strengthening of the Credit Contracts and Consumer Finance Act (CCCFA) in 2021 would have a more-than-minimal impact as affordability assessments are more comprehensive than in 2019 and it has strengthened bank portfolio quality. Three respondents argued that enhanced stress testing would also have a more-than-minimal impact as the results show that banks are sufficiently capitalised to withstand extreme scenarios and suggested we could use enforcement powers or capital overlays in response to bad stress test results. Two respondents raised the potential for a stagflation scenario and mounting worldwide government indebtedness as additional macroeconomic concerns that we did not discuss in the consultation paper. It was suggested that these could increase the risk of an economic downturn and pose significant risks to banks. The international experts did not agree that cumulative policy and macroeconomic changes relative to the 2019 Review justified a change in capital settings. They broadly agreed that the macroeconomic environment has deteriorated in recent years, and that policy and legislative changes since 2019 would have unclear or minimal impacts on capital requirements. One expert thought we had underestimated the extent to which macroeconomic risks have increased and suggested changes could have severe impacts on the New Zealand economy that more than offset policy changes. Comment We agree with submissions that think that the impacts of policy changes may be slightly larger than what we consulted on, but not to the extent that some respondents have suggested. The additional macroeconomic issues raised support our view that risks in the broader environment have increased since 2019. Changes to the CCCFA in 2024 included a significant easing of requirements and allowed a return to 'principles-based' affordability checks. In our view, as a result, requirements for residential mortgage lending are not dramatically tighter than in 2019. CCCFA rules for personal lending are notably stricter than in 2019 and would improve loan quality in this area. In the consultation paper we noted that stress test results are limited to the specific scenario being tested, sensitive to underlying assumptions, and that it is difficult to capture the real-world complexities of a financial crisis. Our enhanced stress testing regime allows us to assess a wider

2025 Capital Review Summary of Submissions and Policy Decisions 19 range of specific risks, but the scenarios do not capture all possible risks facing the financial system. We therefore view stress test results and capital requirements as complements in our prudential framework, rather than substitutes. Response On balance, we do not think the broader context is materially different to our assessment in the consultation paper. The impacts of some policy changes since 2019 might be slightly larger than what we consulted on, but we do not think these impacts are as significant as some respondents suggested. Oliver Wyman report Alongside the consultation paper, we also published an independent report by Oliver Wyman comparing capital ratios for New Zealand’s five largest banks to major banks in a set of peer countries. The report found that New Zealand banks had relatively high amounts of CET1, but alternative metrics suggested that New Zealand was more typical of peer countries. We discussed the report’s findings in the consultation paper and asked respondents for any feedback on this analysis. Consultation feedback Around one-third of respondents commented on the analysis, including deposit takers, groups, and individuals. These respondents largely agreed with the findings of the report. Most noted that it highlights the general conservatism in New Zealand’s capital requirements or pointed to specific areas of conservatism identified in the report. Critiques were mostly targeted to specific methodological points:  Two deposit takers identified variations from Basel rules that Oliver Wyman did not adjust for in their report. These were third-party securitisation asset exposures and the effect that prohibiting convertibility features for Additional Tier 1 and Tier 2 instruments has on multinational banks.  One deposit taker suggested that Norway was not a useful comparator country because the sample only includes one Norwegian bank (which is highly capitalised).  One deposit taker suggested that capital coverage ratios were not a useful metric because they do not adjust for the riskiness of a banks’ lending.  One respondent argued that European countries were overrepresented in the sample and recommended including more countries with foreign-owned banks that have shown resilience in previous financial crises.  The same respondent pointed out that, on average, banks in comparator countries are much larger than those in New Zealand. Additional analysis from the respondent indicated that New Zealand bank capital ratios are more comparable to a wider group of global banks of similar size to the largest NZ banks.

2025 Capital Review Summary of Submissions and Policy Decisions 20 Comment The caveats raised by respondents were largely addressed either in the consultation paper or in the Oliver Wyman report. Oliver Wyman noted that adjustments are approximate because of the level of data that would be necessary to adjust capital ratios for all variations from Basel rules. The adjustments therefore focus on the most material areas of variations and key assumptions are outlined in the report. The set of comparator countries was chosen using an objective criterion based on economic and financial similarities with New Zealand. Some European countries were excluded to avoid overrepresentation and to ensure diversity in the sample. Alternative comparisons were included in the report to support the main findings. This included comparing New Zealand banks to a wider group of similarly sized global banks, which suggested that some international peer banks have higher capital ratios. Capital coverage ratios were also included to compare the level of capital that banks hold, but the disadvantages were noted. We drew out these alternative comparisons and caveats in our consultation paper. Response Our interpretation of the results of the Oliver Wyman report has not changed following the consultation. We generally agree with the caveats raised by respondents and considered these when making final recommendations. 2.1.3 Capital stack options In the consultation paper, we set out two options for Group 1 deposit takers and one for Groups 2 and 3. Our proposed capital settings continued to promote financial system stability and the safety and soundness of individual deposit takers. However, we noted that they are likely to reduce the cost of regulation compared to the 2019 framework. Option 1 used a similar form of capital to the 2019 framework but reduced the amount. It had an emphasis on having a large buffer of high-quality going concern (CET1) capital, which would promote deposit takers’ resilience to shocks and create headroom to recover the resilience of a deposit taker as it faces losses. Option 2, which was ultimately chosen, adjusted the form of capital by introducing additional Loss Absorbing Capacity (LAC) on top of the capital stack. It had a smaller Prudential Capital Buffer (PCB) for Group 1, compared with the 2019 framework. However, the LAC would be debt that could be bailed-in to recapitalise a distressed Group 1 deposit taker (See Section 2.2.1 for details). Under this option, tier 2 capital instruments for Group 1 deposit takers would also have the same write-down and conversion features as LAC instruments (“revised Tier 2”). These proposals are shown in the table below with the final decisions presented in the diagram below that.

2025 Capital Review Summary of Submissions and Policy Decisions 21 Table 6: Summary of proposed capital ratios (% of RWA) Proposed options for Group 1 (%) Proposed option for Groups 2 and 3 (%) Option 1 No LAC Option 2 LAC Group 2 Group 3 Minimum Tier 1* (CET1) 6 Minimum Total Of which max Tier 2 9 3 9 3 9 3 9 3 LAC - 6 - - Prudential Capital Buffer (all CET1) Of which CCyB Of which Domestical Systemically Important Bank (D-SIB) buffer 8 1 3 6 1 1 5 1 n/a 4** 0 n/a Totals Tier 1/CET1 14 12 11 10** Total + LAC 17 21 14 13 *Note: All options include the removal of AT1 capital as a recognised form of regulatory capital. This means all Tier 1 capital will be made up of CET1 capital. **Note: Group 3 deposit takers without a credit rating are subject to a 5 percent Prudential Capital Buffer.

2025 Capital Review Summary of Submissions and Policy Decisions 22 Figure 1: Final capital and LAC ratios (% of RWA) Consultation feedback Option 1 vs Option 2 Group 1 deposit takers supported the LAC option over the non-LAC option, though many noted that their support was conditional on LAC instruments and revised Tier 2 satisfying APRA rules (which is addressed separately in Section 2.2.1 below). Some Group 2 stakeholders argued that Option 2 would reduce the degree of proportionality in the framework relative to the status quo, limiting their ability to compete. They raised concerns about the use of internal LAC acting as a barrier to domestic banks moving to Group 1 and reducing the depth of the market for Group 2 Tier 2 instruments. The D-SIB buffer – designed to reflect the systematic risk associated with the largest deposit takers – was also 2% under the 2019 decisions but is only 1% in Option 2. In addition, the largest banks’ use of IRB models means they have lower risk weight outcomes, which in the past has tended to offset this D-SIB buffer. Our independent experts raised a number of concerns about the LAC option and two explicitly advised against this option. Their concerns include the need to confirm the design of LAC instruments, the risks that bail-in may not work as expected, the reduction in the depth and liquidity of markets for deposit takers’ Tier 2 capital instruments, and the fact that our crisis management framework is still under development. Closer alignment with APRA Respondents, particularly Group 1 deposit takers, argued the levels of capital in our proposals were still too high – with many arguing for alignment with the risk weights and capital ratios used in Australia. On the other hand, the independent experts – and some other respondents - questioned whether the reduction in the amount of capital compared to the 2019 Capital Review decision was warranted at all.

roup 1 roup 2 roup Tier 1 Tier 2 AC

2025 Capital Review Summary of Submissions and Policy Decisions 23 Levels for Groups 2 and 3 Many respondents were concerned that Option 2 lacked sufficient proportionality or differentiation in capital ratios, especially given Group 1 entities are systemically important. Group 3 deposit takers were also concerned about a loss of proportionality compared with 2024 Capital Standard proposals and argued for lower ratio requirements. They noted we proposed reductions in requirements for Groups 1 and 2 based on a higher risk appetite in the consultation paper but did not propose ratio reductions for Group 3. Response Choice of Option 2 We have adopted Option 2 as it aligns with our risk appetite and think that it best balances the costs and benefits for New Zealand. Final decisions required a careful assessment of all the evidence we received and heard on the consultation paper, including the views of the independent experts. Our cost benefit analysis (CBA) was one input into the decision-making process. Overall, our CBA suggests:  Option 2 – the approach we have adopted – has a larger net benefit than the 2028 settings under the 2019 Review decisions and Option 1. The expected net benefit of the chosen option is +0.12% of GDP (with a range of -0.02% to +0.19%) compared to the 2019 Review decisions.  Lending rates are expected to be lower under all options analysed as compared to the 2028 settings under the 2019 Review decisions. The chosen option is expected to reduce lending rates by 15bps as compared to the 2019 Review decisions.  The loss in GDP from a crisis is expected to be higher under all options analysed as compared to the 2028 settings under the 2019 Review decisions. The chosen option is expected to increase the cost of crises by 0.15% of GDP compared to the 2019 Review decisions. The CBA is however subject to relatively high levels of uncertainty, and so we considered a range of variations on these two options, with different mixes of capital to help inform the decisionmaking process (see the full CBA for details14). Internal LAC helps support our preferred crisis management approach to recapitalising a distressed Group 1 deposit taker, if needed. Our preferred crisis management approach for Group 1 deposit takers involves co-ordination among Australian and New Zealand authorities. Internal LAC has the potential to enhance the current “single point of entry” (SPE) approach for Group 1 deposit takers by providing a legal mechanism to ensure the down-streaming of new Tier 1 capital in a group-level recovery situation. This would likely support Trans-Tasman co-ordination in a recovery situation. However, there will always be significant uncertainty over how well LAC would work to stabilise and recapitalise a distressed deposit taker. We will therefore keep engaging with relevant authorities through existing co-operation mechanisms (e.g., Crisis Management Group and Trans-Tasman Banking Council) to maintain and enhance our approach to dealing with a

14 CBA: https://www.rbnz.govt.nz/-/media/project/sites/rbnz/files/regulation-and-supervision/banks/capital-review/2025/cost-benefitanalysis.pdf

2025 Capital Review Summary of Submissions and Policy Decisions 24 distressed deposit taker in accordance with the DTA’s purposes. 15 We will also need to use our wider DTA tools to ensure we have additional options available, so that we can meet our mandate as resolution authority. When domestic deposit takers grow and move to Group 1 or when roup 1 deposit takers’ ownership structures change (e.g. shifting to a non‑Australian parent), we can manage the circumstance through variations in that deposit taker’s conditions of licence incorporating the specific features of that deposit taker and our resolution strategies. Potential variations could include a higher level of Tier 2 capital (without conversion or write off features) or issuing LAC to a holding company in New Zealand. The consultation on the LAC design in 2026 will cover how LAC requirements will apply to a future Group 1 deposit taker that is not a wholly owned subsidiary of an Australian parent, to ensure there are not barriers to entry, growth or organisational evolution. The “Levels for Groups 2 and 3” section below also explains how we considered proportionality and competition. Closer alignment with APRA Aligning capital requirements with those used in Australia would leave New Zealand’s capital requirements lower relative to peer countries in the Oliver Wyman report than is appropriate – given our assessment that New Zealand is riskier than many of these countries. Further, this option performs worse in our CBA. In addition, as this option would lead to larger falls in capital for Group 1 than Group 2, it would reduce the gap between Group 1, and Group 2 and 3 deposit takers, undermining competition and proportionality. Australia’s capital framework is based on the premise that a small economy with reliance on wholesale funding should reduce the risk of financial crises by maintaining an ‘unquestionably strong’ level of resiliency. For that reason, they (like many countries) have pushed their total loss absorbency requirements beyond the minimums in the Basel framework. For New Zealand, which is smaller and even more peripheral to the global economy, the arguments for building resilience seem to apply even more than they do for Australia. For example, Standard and Poor’s Banking Industry Country Risk Assessment deem New Zealand’s banking industry risk the highest of the comparator countries in the Oliver Wyman report, and substantially riskier than Australia: they rate Australia as 2/10 and New Zealand as 4/10. This reinforces our view that, while we support the underlying rationale of Australia’s approach to capital adequacy, directly replicating APRA’s calibrations would not be appropriate for New Zealand’s risk profile. Levels for Groups 2 and 3 We carefully considered whether there were options that could increase proportionality more than the option we decided on (Option 2), without increasing Group 1 capital levels. This included considering whether any of the other prudential standards could be eased for those groups. While subject to a level of judgement, we consider the prudential standards already represent the minimum level required to meet our statutory purposes. Accordingly, any recalibration of the settings to change the relativities would require implicitly or explicitly increasing requirements for Group 1 deposit takers.

15 Section 259 of the DTA sets out additional purposes for crisis management and resolution. Section 261 of the DTA requires the Reserve Bank to publish our expected resolution strategy for dealing with deposit takers and our intended approach to co-operating with relevant Australian authorities among other content in a Statement of Approach to Resolution.

2025 Capital Review Summary of Submissions and Policy Decisions 25 Group 2 and Group 3 deposit takers already have a smaller buffer, so a shorter runway for recovery, than Group 1. Further, the proposals in the consultation paper and the additional risk weight changes mean Group 2 and Group 3 are expected to see substantial drops in capital requirements relative to the 2019 status quo. Given one of the purposes under the DTA is to promote the safety and soundness of individual deposit takers, and that smaller deposit takers are inherently more likely to fail, there is a minimum amount of capital the smallest deposit takers must maintain to ensure soundness. Our analysis suggests further reductions in capital ratios would be unlikely to change outcomes for Groups 2 and 3 significantly as a whole. This is because lower ratios would allow capital to drop to levels observed before the 2019 Capital Review. In that context, regulatory settings were less binding - instead, factors like desired credit ratings and ICAAPs16 would be the main determinant of capital levels, as they were prior to the 2019 Capital Review. Other decisions also support Group 3 entities, such as allowing a slower transition to higher capital ratios while bringing forward new risk weight changes and allowing perpetual preference shares to count as Tier 2 capital for Group 3 (which gives smaller deposit takers more options for raising Tier 2 capital if desired). 2.1.4 Additional Tier 1 capital Background Additional Tier 1 (AT1) capital is a form of regulatory capital positioned between Common Equity Tier 1 (CET1) and Tier 2 capital. It was introduced in New Zealand as part of post-GFC banking reforms and is designed to absorb losses on a going concern basis - supporting deposit takers during stress periods. Globally, AT1 was intended to achieve this by cancelling distributions to preserve capital and converting into ordinary shares (or being written off) at a trigger point to provide additional CET1. While more effective than Tier 2 in absorbing losses, AT1 is less effective than CET1. Following the 2019 Capital Review, we revised the definition of AT1 capital in New Zealand, diverging from Basel standards by excluding contingent convertibility and write-off clauses and requiring AT1 to be treated as equity, in the form of redeemable perpetual preference shares. The allowable proportion of Tier 1 capital met by AT1 was increased from 1.5% to 2.5% of risk-weighted assets to offer banks greater flexibility. These changes aimed to reduce uncertainty about whether new shares would be issued in the New Zealand subsidiary or the Australian parent when the contingent convertibility feature of AT1 instruments was triggered. They were also intended to clarify the practical loss-absorbing value of AT1 instruments, simplify the capital framework, support mutual ownership structures, and improve transparency. Banks have been transitioning to these new requirements since 2021. Proposal We consulted on removing AT1 as a form of regulatory capital. In our consultation paper, we discussed challenges faced by deposit takers issuing AT1. In addition, we discussed international experience with AT1 where it has not provided the loss-absorption nor had the stabilising effect it

16 Internal Capital Adequacy Assessment Process – Under ICAAP each deposit taker is required to assess its own capital adequacy relative to the risks it faces, beyond just meeting regulatory minimums.

2025 Capital Review Summary of Submissions and Policy Decisions 26 was designed to achieve. We also noted that APRA is phasing out the use of AT1 as eligible bank capital in Australia. Our conclusion was that removing AT1 as a form of regulatory capital meets the assessment criteria we developed for the purpose of this review. We posed four questions relating to AT1 asking for views on the proposal to remove AT1 as a form of regulatory capital, on what transitional arrangements should look like and whether there were particular factors that should be taken into account for Group 3 deposit takers. Consultation feedback We received 19 submissions in relation to the proposal. Responses were received from deposit takers in all groups as well as industry bodies and individuals. Respondents overwhelmingly supported removing AT1 from the capital stack. Conversely, one market respondent observed that this would remove the ability for domestic investors to gain a quasi-equity exposure to the banking industry. Respondents agreed with our analysis of the challenges presented by AT1, both from the perspective of issuers and from a prudential perspective. Respondents liked the alignment with APRA. Furthermore, the general simplification of the framework was seen as a positive, as well as providing benefits from a crisis resolution angle. Feedback broadly focussed on two aspects: what would replace the AT1 slice in the capital stack, and the nature of the transitional arrangements. This section focuses on feedback relating to the proposal to remove AT1 – we discuss what the capital stack will comprise in Section 2.1.3. Care needs to be taken with transitional arrangements Respondents were keen to ensure the transition would be managed carefully, with clear and early communication, to avoid market disruption. Some respondents asked for a clear grandfathering process for legacy instruments. Many respondents felt that grandfathering existing instruments to the first call date was an appropriate mechanism. A large deposit taker made the point that there was a risk of unintended consequences and that, in developing transitional provisions, we should take care not to penalise any deposit taker that has been proactive in trying to overcome the current challenges of AT1. Many respondents raised the prospect of the change in policy constituting a regulatory event that would trigger the ability for deposit takers to redeem instruments early (subject to having the appropriate provisions in their AT1 instruments).17 There were mixed views on whether we should incentivise the early redemption of legacy instruments (for example, to remove any complexity or ambiguity that might arise in the context of a stress situation) or whether that would lead to unnecessary market disruption. From a Group 3 perspective, one respondent highlighted the need to have an adequate timetable to limit the impact of removing AT1 on smaller deposit takers, particularly those with a mutual structure. This respondent suggested a derecognition schedule (as Tier 1) as well as ongoing

17 See D2.5(3) in Reserve Bank of New Zealand. (2023). BPR110: Capital Definitions. https://www.rbnz.govt.nz/- /media/project/sites/rbnz/files/consultations/banks/review-capital-adequacy-framework-for-registered-banks/bprdocuments/bpr110-capital-definitions-oct-23.pdf

2025 Capital Review Summary of Submissions and Policy Decisions 27 recognition of PPS as Tier 2 capital would support a smoother transition. A hybrid approach was also suggested by another respondent. Comment Impact on different groups We are conscious this decision will impact deposit taker groups in different ways depending on their current capital structure. Groups 1 and 2 deposit takers As at the date of this document, AT1 instruments have been issued by three Group 1 deposit takers and one Group 2 deposit taker. Two of these also have transitional AT1 still on issue.18 For the larger deposit takers, AT1 provides a cheaper form of Tier 1 capital than CET1. However, as discussed in the consultation paper, there are challenges that impact a deposit taker’s ability to issue it. One large deposit taker has been able to utilise its corporate structure in a way that enables it to issue AT1 in an efficient manner, but this is not something available to most current Group 2 deposit takers. Group 3 deposit takers There is one future Group 3 deposit taker with a significant amount of PPS in its capital structure. Otherwise, PPS have generally not been issued by NBDTs. We have also heard from respondents that very small deposit takers are often more locally situated and connections to the community can be important. In some cases, they see issuing PPS (typically to existing customers or members) as more in line with this ethos than issuing wholesale debt to external investors (assuming they have the ability and market access to do so). Response In light of the feedback, we are proceeding with the proposal to remove AT1 from the capital stack. Proportional approach to transitional arrangements We have considered the feedback received about transitional arrangements. There are some key elements to transitional planning that we have sought to balance with our prudential objectives. These are to:  Avoid market disruption  Reduce the risk of unintended consequences  Reduce/remove complexity and ambiguity  Mitigate impacts on small deposit takers

18 Transitional AT1 are the instruments issued prior to 2019 when the eligibility criteria for AT1 was significantly revised. These instruments are subject to the phasing out arrangements in Part A of BPR110: Capital definitions.

2025 Capital Review Summary of Submissions and Policy Decisions 28 We have sought to reconcile these in a way that incorporates the different impacts on different groups. Our intended transitional arrangements are summarised in the table below. These arrangements are different for Groups 1 and 2 on the one hand, and Group 3 on the other. For all groups, current AT1 capital instruments or PPS will be able to be treated as Tier 1 capital to some extent and for a limited period of time. However, we will amend the current Banking Prudential Requirements such that no further AT1 eligible capital can be issued. For Groups 1 and 2, the intention is for capital recognition for AT1 instruments to ultimately phaseout completely. For Group 3, a portion of existing issued PPS can be recognised as Tier 1 and a portion as Tier 2 capital during the transitional period. This balance will gradually shift until no PPS will be recognised as Tier 1 capital. However, Group 3 deposit takers will be able to continue to issue PPS (as an alternative option to subordinated debt) and for it to be treated as Tier 2 capital on an ongoing basis. We will seek feedback from Group 2 deposit takers during the Capital Standard Exposure Draft process to see if there is demand for this approach to be extended to Group 2. The table below provides a snapshot of how the transitional arrangements are expected to work. Table 7: Transitional arrangements for AT1 capital From Deposit taker group Change 1 October 2026 Groups 1 and 2 We will update the BPRs so that no further AT1 can be issued. However, any existing AT1 (or transitional AT1) will continue to be recognised as Tier 1 regulatory capital. Transitional AT1 will continue to be phased out according to the schedule in BPR110: Capital Definitions. Group 3 We are not making any changes to the treatment of PPS within the Deposit Takers (Credit Ratings, Capital Ratios, and Related Party Exposures) Regulations 2010 (NBDT Capital Regs). 1 December 2028 Groups 1 and 2 The Capital Standard comes into force. Existing issued AT1 capital instruments will cease to be recognised as Tier 1 capital in accordance with a derecognition schedule commencing 1 December 2029. By 1 December 2031, AT1 will have been completely phased out. Group 3 The Capital Standard comes into force. Half of existing issued PPS will be recognised as Tier 1 capital, with the balance able to be recognised as Tier 2 capital.

2025 Capital Review Summary of Submissions and Policy Decisions 29 From Deposit taker group Change This balance will gradually shift such that by 1 December 2032, no PPS will be treated at Tier 1 capital but 100% of any PPS can be treated as Tier 2 capital. Groups 1 and 2 phase out schedule The table below sets out the maximum proportion of AT1 capital instruments on issue on 1 October 2026, that Groups 1 and 2 deposit takers may include in their total Tier 1 capital on any date on or after 1 October 2026. This is the same approach that was taken to phasing out the recognition of AT1 instruments issued before 1 October 2021. This shows that there is no reduction in the recognition of AT1 capital until the DTA has been in force for a full year. Table 8: Groups 1 and 2 phase out schedule Group 1 and 2 Tier 1 1 October 2026 to 30 November 2028 100% 1 December 2028 to 30 November 2029 100% 1 December 2029 to 30 November 2030 67% 1 December 2030 to 30 November 2031 33% 1 December 2031 to 30 November 2032 0% Group 3 phase out schedule For Group 3 deposit takers, the table below sets out the maximum proportion of Tier 1 capital requirements that may be met with PPS issued in accordance with NBDT Capital Regs with the balance able to be treated as Tier 2 capital. From 1 December 2032 onwards, any PPS issued by Group 3 deposit takers may only be treated as Tier 2 capital.

2025 Capital Review Summary of Submissions and Policy Decisions 30 Table 9: Group 3 phase out schedule Group 3 Tier 1 Tier 2 1 October 2026 to 30 November 2028 Existing requirements apply (i.e. for NBDTs) 1 December 2028 to 30 November 2030 50% Any remaining PPS can be classified as Tier 2, up to the limits on Tier 2 1 December 2030 to 30 November 2032 25% Any remaining PPS can be classified as Tier 2, up to the limits on Tier 2 1 December 2032 onwards 0% All PPS can be classified as Tier 2, up to the limits on Tier 2 This approach addresses transitional concerns Avoiding market disruption We consider that continuing to recognise existing AT1 instruments as Tier 1 capital for a period of time mitigates the risk of market disruption as deposit takers will not be incentivised to immediately replace instruments in issue. Deposit takers will have scope to redeem their AT1 instruments in an orderly manner (subject to meeting the requirements of BPR120: Capital adequacy process requirements). Based on our analysis, half of existing issued AT1 instruments will have reached their first optional redemption date by November 2029 with the remaining existing issued AT1 instruments doing so by November 2030. In addition, under existing policy, deposit takers have the right to call or redeem AT1 instruments if a regulatory event has occurred provided their AT1 instruments include the relevant provisions.19 Following publication of our decisions on 17 December 2025, some deposit takers announced that they have determined a regulatory event has occurred. 20 Nevertheless, deposit takers still need to meet the requirements in BPR120: Capital adequacy process requirements21 prior to being able to redeem their AT1. This means that, where deposit takers are not replacing the instrument with the same or better-quality capital, they need to demonstrate to the RBNZ’s satisfaction that their capital ratios (including their prudential capital buffers) would be sufficiently above minimum requirements following the redemption. Regulated entities can contact their supervisor for more information on this process.

19 See D2.5(3) of BPR110: Capital definitions 20 See ANZ: NZX, New Zealand's Exchange - Announcements, Update on RBNZ’s Capital Review Decisions; Westpac NZ Ltd: NZX, New Zealand's Exchange - Announcements, AT1 Regulatory Event Following RBNZ Decisions; BNZ: NZX, New Zealand's Exchange - Announcements, BNZ Perpetual Preference Shares - Regulatory Event Kiwibank: NZX, New Zealand's Exchange - Announcements, AT1 Regulatory Event Following RBNZ Capital Review Decision 21 See C2.2 of BPR120: Capital adequacy process requirements

2025 Capital Review Summary of Submissions and Policy Decisions 31 Overall, we consider that the transitional arrangements will enable early redemption of existing AT1 instruments but do not provide any incentive to do so. Reducing risk of unintended consequences By not changing the regulatory treatment of AT1 and phasing it out in a well communicated, clearly defined way, we expect that the risk of unintended consequences is low. Addressing complexity and ambiguity There will be a degree of complexity while we phase out current instruments at the same time as we introduce new requirements. However, we expect that clear and early communication about how the transition will operate will mitigate any concerns. The interaction with the new requirements for the capital stack, including new LAC instruments (if relevant) is discussed as part of that section. We will also consult on the form of LAC next year. Mitigating impacts on small deposit takers This approach recognises the limited market access and issuance capacity that smaller institutions typically face, enabling them to maintain regulatory compliance while the transition is underway. By providing a period during which a proportion of PPS contributes to Tier 1, Group 3 deposit takers will have greater flexibility to adapt to the new capital framework. Over time, the capital recognition will shift wholly to Tier 2, ensuring consistency with the broader regulatory objectives while minimising disruption for smaller entities. We think this is a proportional response. Concerns regarding AT1 primarily arise from the intended functioning of these instruments within the capital stack, rather than from their intrinsic characteristics. 2.1.5 Standardised risk weights Introduction Capital requirements express the amount of capital a deposit taker must have as a percentage of its RWAs. Deposit takers are required to have more capital (such as equity) for riskier loans to provide a large buffer to absorb potential losses. Risk weights help determine the amount of capital the deposit taker is required to have by reflecting the risks of underlying lending. Our framework for calculating risk weights is based on the Basel Framework that sets international standards and minimums for bank requirements. Under this framework, there are two approaches used to calculate risk weights:  The standardised approach, where risk weights are set based on the broad characteristics of loans, such as the loan-to-value ratio (LVR) for residential mortgage lending (RML).  The internal ratings-based (IRB) approach, where deposit takers accredited by the Reserve Bank are authorised to calculate credit risk weights using their own internal risk models, enabling them to get a more precise measure of the riskiness of lending. Currently only Group 1 deposit takers have been approved to use this approach, but others can apply.

2025 Capital Review Summary of Submissions and Policy Decisions 32 Summary of risk weight changes Evidence provided in the submissions suggested that our standardised risk weight proposals in the consultation paper were still too conservative in some areas. We have reviewed the evidence presented in the submissions and agree that more can be done to align the standardised risk weights with the actual risk of lending – a key driving factor for our standardised framework. To achieve this, we are proposing the following changes to the consultation paper proposals to align with that evidence (see also Table 10 below):  Lower risk weights for low-LVR RML – matching APRA and the Basel Framework (Basel) at VRs ≤ 70%.  Lower risk weights for lending to Community Housing Providers (CHPs), particularly for borrowers with long-term funding contracts with the Crown.  Removed housing co-operatives from the CHP category. This lending will be able to be treated as RML, provided it meets the requirements for RML lending.  Not proceeding with the suggested new, higher 150% risk weight for unsecured personal lending.  Not proceeding with the suggested new flat 100% risk weight suggested for unrated commercial property lending. There was also some feedback comparing New Zealand risk weights with Australian risk weights. Our assessment is that the changes above bring risk weights closer to reflecting actual risk, after considering the feedback and additional evidence that we received. The points below summarise the revised changes compared with Australia:  RML: New Zealand would be the same as Australia, except at some high LVRs (over 90%) where New Zealand’s standardised risk weights would be around 5% higher than Australia’s.  Corporate lending, including to small and medium-sized enterprises (SMEs): New Zealand would be the same as Australia, except for some additional granularity in Australia relating to commercial property. We intend to seek additional information from deposit takers in 2026 to allow a more in-depth assessment of options relating to commercial property risk weights and potential impacts in New Zealand.  Agriculture: There is no specific risk weight category for agricultural lending in Australia. However, our assessment is that risk weights in New Zealand would likely be lower than Australia’s for agricultural lending with an LVR of 30% or lower.  Personal lending: Australia would have lower standardised risk weights for credit card lending (75%) compared with 100% in New Zealand.

2025 Capital Review Summary of Submissions and Policy Decisions 33 Table 10: Revised key standardised risk weight changes Type of lending Current standardised risk weight (%) Proposed in August 2025 Consultation New standardised risk weight (%) Owner-occupier residential mortgage lending (RML) with loan-to-value ratio (LVR) ≤ % 35 25 20 Owner-occupier RML with LVR 50.01 – 60% 35 30 25 Owner-occupier RML with LVR 60.01 – 70% 35 35 30 Investor RML with LVR ≤ % 40 30 25 Investor RML with LVR 50.01 – 60% 40 35 30 Owner-occupier past due RML (principal and interest) With LMI: Risk weight that corresponds with the LVR and LMI conditions set out in Table C3.1022 Without LMI: 100 With LMI: 80 Without LMI: 100 With LMI: 80 Without LMI: 100 Other standardised residential property RML With LMI: Risk weight that corresponds with the LVR and LMI conditions set out in Table C3.1023 Without LMI: 100 With LMI: 95 Without LMI: 120 With LMI: 95 Without LMI: 120 Small and medium enterprise (SME) retail 100 75 75 SME corporate 100 85 85 Agriculture with LVR ≤ % 100 50 50 Agriculture with LVR 30.01 – 50% 100 75 75 Agriculture with LVR > 50% 100

22 BPR131 - Standardised Credit Risk RWAs 23 BPR131 - Standardised Credit Risk RWAs

2025 Capital Review Summary of Submissions and Policy Decisions 34 Type of lending Current standardised risk weight (%) Proposed in August 2025 Consultation New standardised risk weight (%) Community housing providers Treated as either RML, corporate, or incomeproducing real estate (IPRE), depending on the deposit taker. New standardised category with risk weights aligned with property investment RML. Both standardised and IRB deposit takers required to use the new standardised category. New standardised category with risk weights aligned with owner-occupier RML, plus a maximum risk weight cap of 30% where there is longterm government contract. Both standardised and IRB deposit takers required to use the new standardised category. Housing co-operatives Treated as either RML, corporate, or IPRE, depending on the deposit taker. Included in same category as CHPs above To be treated as owner-occupier RML by deposit takers, under either the standardised or IRB approach. Note: This table only lists the key standardised risk weights that are being changed as a result of this Review. If a standardised risk weight category is not discussed in this table or the following section, then no change has been made. Residential mortgages We suggested in the consultation paper that there was potential scope to add more granular, lower risk weights for lower risk low-LVR RML. To this end, we suggested some specific changes to the risk weights for owner-occupier and property investor RM with VR ≤ 60%. Consultation feedback Almost all respondents supported introducing more granular risk weights, however most wanted further reductions than what we had proposed – to match the risk weights used by APRA and Basel. Respondents that acknowledged higher risk in New Zealand compared with other jurisdictions suggested that higher risk was better reflected in overall capital ratio requirements rather than individual risk weights. A number of respondents also noted that the RBNZ Bank Financial Strength Dashboard shows that IRB deposit takers have higher RML non-performing loans (NPLs) and expected losses than standardised deposit takers, despite deposit takers having relatively homogenous RML books. Respondents also highlighted the following:  The consultation paper proposals would give IRB deposit takers a significant competitive advantage for low-LVR RML. For example, respondents suggested that a 50-60% LVR loan at

2025 Capital Review Summary of Submissions and Policy Decisions 35 an IRB deposit taker would attract a ~15% risk weight. Under the consultation paper proposals, the same loan at a standardised deposit taker would attract a risk weight twice as high at 30%.  Standardised deposit takers have reported peak NPLs of just 0.3% for their mortgage portfolios over the last seven years. This is ten times lower than the implied probability of default (PD) for a 25% risk weight, suggesting a 25% risk weight may be overly conservative.  Australian banks have a higher percentage of NPLs than banks in New Zealand. Respondents suggested that this means that New Zealand’s risk weights should not be more conservative than Australia’s.  Setting higher risk weights than justified by actual risk leads to inefficient outcomes and higherthan-necessary costs. Comment The evidence presented in the submissions makes a strong case for reducing low-LVR RML risk weights to levels similar to those used by APRA. Data available in New Zealand does not incorporate a significant loss experience. This suggests that estimating risk weights directly from the historical loss analysis reported by some of the respondents would be inadequate. However, as Figure 2 below shows, IRB risk weights are consistently lower than standardised risk weights at lower LVRs, but higher at higher LVRs. Figure 2: Average risk weights for residential mortgages, by LVR (%) and borrower category Source: How risk weights affect bank lending We have also assessed stress test results from a range of previous RBNZ bank industry solvency stress tests. The figure below compares stress test results for RML across the different LVR buckets. In Figure 3, loss given default (LGD) accelerates quickly as LVR increases, and is nearly three times higher at 60-70% LVR compared with 0-40% LVR. The implication of this, is that when the LVR is high, the loss to the bank will be substantially higher, since the bank may not have as much value available in the security to cover losses. 0 20 0 60 0 100 120 0 20 0 60 0 100 120 % % Borrower oan to Value Ratio IRB Approach tandardised approach wner occupier Investor

2025 Capital Review Summary of Submissions and Policy Decisions 36 Figure 3: Mortgage stress LGD by LVR (%), 2022 stress test The variability in stress test losses was reflected in the consultation paper proposals. However, after reviewing evidence from the submissions, we have concluded that there is more scope for granularity. For example, respondents argued for different risk weights for owner-occupier RML with LVR 60-70% and LVR 70-80%, where our consultation paper proposal had a risk weight of 35% for both LVR buckets. The additional analysis submitted by respondents encouraged us to revisit these risk weights and our stress test results, and we concluded that assigning different risk weights (30% and 35% respectively) to these two LVR buckets would better reflect actual risk (see Table 11 below). Response In response to the evidence presented, we amended the consultation paper RML risk weight proposals to match APRA risk weights at lower VRs (≤70%). However, at higher LVRs, the consultation paper RML risk weight proposals still appear better aligned with IRB model outputs and stress test results. Therefore, there is significantly less justification for reducing risk weights at higher LVRs so we are not making any changes to the proposed RML risk weights at LVRs > 70%. Our final RML risk weights are shown in Table 11 below. A number of submissions provided solid evidence that the low-LVR RML risk weights we proposed in the consultation paper were too high. This conclusion is based on assessment of standardised deposit takers’ historical losses, including comparisons with IRB loan performance and stress test results. These metrics support lower RML risk weights at low LVRs, which IRB deposit takers are currently able to act upon, but standardised deposit takers are not. Therefore, we concluded that there is limited evidence to support maintaining our current higher risk weights for low-LVR RML. Instead, our assessment is that any New Zealand-specific risk factors are best reflected in overall capital ratio requirements rather than in individual risk weights for exposure categories. 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 0-40 40.01-50 50.01-60 60.01-70 70.01-80 > 80 LGD LVR

2025 Capital Review Summary of Submissions and Policy Decisions 37 At higher LVRs, the evidence is more supportive of maintaining our higher risk weights. In particular, IRB models produce higher risk weights than our standardised risk weight proposals and stress tests show significantly higher losses for RML with higher LVRs. Therefore, there is little evidence to support lowering standardised risk weights for high-LVR RML. Lower risk weights for low-LVR RML will reduce capital requirements, all else equal. Lowering these risk weights may also encourage more lending to flow into housing. However, neither of these are strong reasons to retain higher risk weights for low-LVR RML, given our focus on aligning risk weights with actual risk. Based on the additional evidence submitted, keeping low-LVR RML risk weights artificially high in order to support higher capital requirements would be challenging to reconcile with our approach to prudential regulation. We estimate that the changes below could result in an approximate 7% reduction in RWAs, and overall capital. For Group 1 deposit takers, it is an estimated 5.8% reduction, and for Group 2 it is an estimated 15.3% reduction. However, these estimates expect that the output floor will generally stop binding for three of the four Group 1 deposit takers, due to the combined effect of our lower standardised risk weights across the different lending portfolios. This occurs as the IRB risk weights will be higher than 85% of the standardised risk weights, due to the standardised risk weights falling. This means that those deposit takers would switch to using their IRB risk weights that are unbound by the 85% output floor, so the full effect of the changes to standardised risks weights may not affect most Group 1 deposit takers. If the output floor were to continue to bind, then roup 1’s risk weights would fall by 85% of the standardised risk weight reduction, and overall RWA and capital would fall further than the impacts provided above. Table 11: Overview of original proposal and revised changes for standardised risk weights for RML Loan type RBNZ (%) APRA (%) New (%) Lenders mortgage insurance (LMI)/No LMI, nonproperty investment loan, LVR <=50% 35 20 20 (was 25% in consultation) LMI/No LMI, non-property investment loan, LVR 50.01 – 60% 35 25 25 (was 30% in consultation) LMI/No LMI, non-property investment loan, LVR 60.01 – 70% 35 30 30 (was 35% in consultation) LMI/No LMI, non-property investment loan, LVR 70.01 – 80% 35 No LMI, non-property investment loan, LVR 80.01 – 90% 50 No LMI, non-property investment loan, LVR 90.01 – 100% 75 70 75 No LMI, non-property investment loan, LVR > 100% 100 85 100

2025 Capital Review Summary of Submissions and Policy Decisions 38 Loan type RBNZ (%) APRA (%) New (%) LMI, non-property investment loan, LVR 80.01 – 90% 35 40 35 LMI, non-property investment loan, LVR 90.01 – 100% 50 55 50 LMI, non-property investment loan, LVR > 100% 100 70 100 LMI/No LMI, non-property investment interest-only loan, LVR > 80%, loan term > 5 yrs / unknown N/A 100 N/A LMI/No LMI, property investment loan, LVR <=50% 40 25 25 (was 30% in consultation) LMI/No LMI, property investment loan, LVR 50.01 – 60% 40 30 30 (was 35% in consultation) LMI/No LMI, property investment loan, LVR 60.01 – 70% 40 LMI/No LMI, property investment loan, LVR 70.01 – 80% 40 45 40 No LMI, property investment loan, LVR 80.01 – 90% 70 65 70 No LMI, property investment loan, LVR 90.01 – 100% 90 85 90 No LMI, property investment loan, LVR > 100% 100 105 100 LMI, property investment loan, LVR 80.01 – 90% 50 LMI, property investment loan, LVR 90.01 – 100% 75 70 75 LMI, property investment loan, LVR > 100% 100 85 100 LMI/No LMI, property investment interest-only loan, LVR > 80%, loan term > 5 yrs / unknown N/A 100 N/A

2025 Capital Review Summary of Submissions and Policy Decisions 39 Past due Summary of consultation proposal In our consultation paper, we proposed aligning our past due RM risk weights with APRA’s to better reflect the underlying risk of this type of lending and more closely align with international benchmarks. Our current approach to past due RML risk weights is:  The risk weight for RML without qualifying LMI that is a 90-day past due asset is 100%.  The risk weight for RML with qualifying LMI is the risk weight that corresponds to the LVR and LMI conditions as set out in Table C3.10 (see Table 11 above).24 Table 12 shows APRA’s approach to past due RML risk weights. Table 12: APRA’s risk weights for defaulted residential property exposures Risk weight (%) Owner-occupied principal-and-interest with LMI 80 Owner-occupied principal-and-interest without LMI 100 Other standardised residential property with LMI 95 Other standardised residential property without LMI 120 Consultation feedback Most respondents did not comment on the past due proposal. Of those that did, two were opposed, while others sought clarification of what the new risk weights would be but did not offer any strong views one way or the other. The respondents that were opposed to the past due risk weight changes cautioned that applying a 100% risk weight to past due RML regardless of LMI or LVR would not accurately reflect the risk of the lending and could disincentivise deposit takers from supporting customers through hardship. They proposed an alternate approach, which would take into account both LMI and LVR for owner-occupied and investor RML. Comment Our current approach also applies a flat 100% risk weight to past due RML regardless of LVR if the lending does not have qualifying LMI. Most lending in New Zealand does not have LMI, so changing to the new risk weight approach for past due RML is unlikely to affect deposit takers’ approaches to supporting customers through hardship.

24 Reserve Bank of New Zealand. (2024). BPR133: IRB Credit Risk RWAs. https://www.rbnz.govt.nz/- /media/project/sites/rbnz/files/regulation-and-supervision/banks/banking-supervision-handbook/bpr133-irb-credit-risk-rwa-1-july2024pdf.pdf

2025 Capital Review Summary of Submissions and Policy Decisions 40 There is also currently no international justification for taking an alternative approach like that suggested by the respondents. The Basel approach is generally to apply a flat 100% risk weight to defaulted mortgage exposures. As we stated in the consultation paper, we have used data provided to us by deposit takers through our information request in October 2024 to calculate the impact of these changes to standardised risk weights for past due RML on overall RWA. The overall share of loans in these categories is very small, making up approximately 0.7% of RML for Groups 1 and 2 deposit takers.25 As a result, making these changes to past due RML standardised risk weights would make very little difference to the overall RWA, and therefore capital levels, of deposit takers (approximately 0.2 percentage points). However, it would mean that these higher risk lending categories have a generally higher risk weight assigned to them which is more aligned with the actual risk of the lending. Response The majority of respondents did not offer a view on the past due RML proposal in the consultation paper and those that did, did not offer any strong evidence to justify revising it. We are therefore not making any amendments to the consultation paper proposal for past due RML and are proceeding with the changes to past due RML risk weights outlined above. Reverse mortgages Consultation feedback We did not consult on changes to reverse mortgage risk weights, but a few respondents highlighted this as an area where they would like to see some changes. These respondents suggested that the recent changes to reverse mortgage risk weights were still overly conservative and should be more aligned with the newly proposed risk weights for property investor RML (while accounting for tail risks in higher-LVR lending) to reflect the added granularity and reduced risk weights. The most detailed proposal in this area suggested the following:  Align reverse mortgage risk weights with those for investor residential mortgages, while ensuring that tail risks at higher LVRs are appropriately captured.  Removing the 20% valuation discount and instead calibrating LVR buckets to reflect severe but plausible house price declines.  A more granular LVR-based risk weighting schedule. Comment To support these suggestions, respondents provided a range of evidence regarding low loss rates on this type of lending. They also noted that this performance had been achieved across a range of economic cycles, including periods of declining house prices and rising interest rates. One respondent noted that under a range of assumptions, house price falls would need to be in excess of 60% to result in losses on reverse mortgages.

25 As at June 2024.

2025 Capital Review Summary of Submissions and Policy Decisions 41 The evidence presented in the submission makes a solid argument for reducing risk weights. Many of these points are worth considering further. We also acknowledge that with our lowering of standard RML risk weights as part of this Review, including the further changes outlined in this paper, reverse mortgage risk weights are now relatively high compared with risk weights on other RML lending. Response As noted above, some of the points raised in submissions are worth considering further. However, we are not making any changes to reverse mortgage risk weights at this point in time. Risk weights for reverse mortgages were adjusted in 2023, following a review period that exceeded one year. As reverse mortgages were not within scope for this Review and were not part of our original package of proposals, we have not had sufficient time to recreate the financial modelling from our 2023 changes. Without this opportunity we are not comfortable making changes to the approach. Nevertheless, we acknowledge that reverse mortgage risk weights are now relatively high compared with our revised RML risk weights. We will consider this further in 2026, to incorporate in the Capital Standard. Corporate Consultation feedback SME lending and thresholds Currently our standardised approach sets a 100% risk weight for any corporate lending that is not covered by a credit rating. In the consultation paper, we proposed creating two new subcategories aligned with the APRA and Basel approaches:  SME retail (75% risk weight)  SME corporate (85% risk weight) The proposed change was widely supported, and no amendments are being made. Most feedback focused on the definitions and where the threshold would be set for SME loan eligibility. The question of where to set the threshold for qualification as an SME exposure also interacts with the ‘retail’ treatment of ME lending in the IRB approach. The IRB approach states that a loan that is extended to a small business and managed as a retail exposure, and that does not qualify as an RM , is eligible for retail treatment where the banking group’s total business-related exposure to the borrower is less than NZD 1 million. Stakeholders asked us to consider increasing the IRB threshold to a higher number. The current $1 million setting was put in place in 2008. APRA updated their threshold to $1.5 million (AUD) in 2024. We will consider increasing the $1 million IRB threshold during 2026. We intend to set the eligibility threshold for SME loans in the standardised approach at the same level as wherever we end up setting the IRB retail threshold described above – i.e., once the reassessment of the $1 million threshold is complete.

2025 Capital Review Summary of Submissions and Policy Decisions 42 Other corporate There is no granularity in the current approach to corporate exposures without a credit rating, which are risk weighted at 100%. This will change once the new SME and agriculture categories are introduced. Nevertheless, outside of those SME and agriculture exposures, all unrated corporates would be treated the same. However, the risks are not the same across this group. For example, low-LVR commercial property lending will generally be lower risk than high-LVR commercial property lending. In this context, one respondent raised concerns about the impact of the output floor. They suggested that, with RML risk weights falling substantially, and no change in unrated corporate risk weights, outside of SMEs and agriculture, unrated corporate lending would be significantly less attractive to lend to comparative to RML. This is because, in their view, there will be additional incentives to direct lending into RML, if possible, and away from other unrated corporate lending. The output floor aspects of this are considered elsewhere in this document. Comment SME lending and thresholds The proposed change was widely supported and is being implemented with no amendments made relative to what we proposed in the consultation. We agree that the thresholds and definitions will be critical to implementation. These will be worked through in 2026. Other corporate We have considered, and rejected the following approaches to add granularity to the framework:  A new category for commercial property risk weights, based on LVRs. This is discussed in more depth in the commercial property section below, where we note that we will consider this in 2026 – subject to gathering and reviewing additional data from deposit takers and carefully reviewing available evidence.  Introducing within ‘all other unrated general corporate exposures’ an APRA-like split to an ‘investment grade’ sub-category with a risk weight of 5%, or 110% for a ‘non-investment grade’ exposure. We are not proceeding with this approach for the reasons outlined below. While the ‘investment grade’ sub-category has some appeal, the average impact (given the current risk weight of 100% for unrated corporates) is likely to be minimal with some risk weights ending up less than 100% and some more than 100%. In addition, any investment grade rating would need to meet a high bar and deposit takers would need to be able to assess the corporate entities against a complex investment grade definition. For example, in APRA’s rules the corporate must be assessed as having “adequate capacity to meet its financial commitments in a timely manner” and be “robust against adverse changes in the economic cycle and business conditions”. Our assessment is that this approach would add significant complexity and introduce further risk while having limited actual impact on aggregate risk weights.

2025 Capital Review Summary of Submissions and Policy Decisions 43 Response SME lending and thresholds We are proceeding with the consultation paper proposal to reduce the risk weights for SME retail and corporate lending to 75% and 85% respectively. In addition to this, we will review the threshold for standardised and IRB SME lending categories in 2026. The level that the SME threshold will be set at will be confirmed during the exposure draft process. This will provide an opportunity for more detailed analysis and feedback. The impact of changing the SME threshold has not been incorporated into the estimated impacts from changes to risk weights in this paper. Other corporate We are making no further changes to other corporate lending at this time, outside of the changes to SME and agriculture lending. However, in response to feedback from multiple respondents, other potential changes that can be made to corporate exposures regarding securitisation and infrastructure, and further granularity in commercial property risk weights, can be considered in separate work programmes, as discussed below. Agriculture Consultation feedback In our consultation paper, we noted that under the standardised approach, agricultural lending is likely captured under unrated corporate lending, as, like SMEs, it is unlikely they have pursued a credit rating due to the cost and barriers involved. This means that all agricultural lending likely receives the same risk weight under the standardised approach (100%), regardless of how risky the underlying lending is. As a result, we proposed introducing three new exposure categories for agricultural lending to enable lenders using the standardised approach to assign risk weights based on the LVR, which would more accurately reflect the actual riskiness of the lending. The risk weights that we proposed are shown in Table 13 below. Table 13: Proposed standardised risk weights for agricultural lending LVR (%) Proposed risk weight (%) LVR <=30 50 LVR > 30 to 50 75 LVR > 50 100 Most respondents supported the additional granularity in agricultural lending, and the risk weights that were proposed. Of the few that did not agree:  One submission did not agree with introducing a separate category for agricultural risk weights at all, due to this approach being inconsistent with international approaches such as APRA and Basel.

2025 Capital Review Summary of Submissions and Policy Decisions 44  One submission suggested the risk weights for agricultural lending should be higher and aligned with our proposed SME corporate treatment as they believe these types of lending have the same risk characteristics.  Two submissions suggested that agricultural risk weights should be lower, with each submission proposing a different alternate approach. One said this was to bring standardised risk weights closer to the risk weights applied by IRB deposit takers, while the other said this was to reduce the conservatism compared to the APRA and Basel approaches for ‘very wellsecured’ lending. Comment None of the submissions provided robust evidence to support changing our consultation paper proposal. The key assertions for those suggesting even lower risk weights were around levelling the playing field between IRB and standardised deposit takers, and further alignment with APRA. However, under the APRA framework, agricultural lending does not have its own risk weight category. Our assessment is that agricultural lending would most likely receive the following treatment under the APRA framework, with the actual treatment depending on the specifics of each particular loan:  If secured by property, it will likely be captured as commercial property ‘not dependent on property cash flows’ (see Table 14 below). This gives a risk weight of 60% for LVR<60%.  If LVR>60%, it will likely be treated as SME, unrated corporate or commercial property, resulting in a risk weight ranging between 75-110%. Therefore, our new agricultural risk weights would likely be lower than the risk weights under APRA’s framework at the low-LVR end and not too dissimilar to APRA’s risk weights at the highLVR end. Table 14 below shows a comparison of our proposed agricultural risk weights and the treatment that agricultural lending likely gets under the APRA framework, based on Prudential Standard APS 112 Capital Adequacy (APS 112). 26 Table 14: Proposed RBNZ agricultural risk weights vs. APRA agricultural risk weights LVR (%) RBNZ proposed risk weight (%) APRA risk weight (%) LVR <=30 50 60 LVR 30-50 75 60 LVR>50 100 - LVR<=60 - 60 LVR>60 100 75-110

26 Australian Prudential Regulation Authority. (2025). Prudential Standard APS 112 Capital Adequacy: Standardised Approach to Credit Risk. https://www.apra.gov.au/sites/default/files/2024- 07/Prudential%20Standard%20APS%20112%20Capital%20Adequacy%20Standardised%20Approach%20to%20Credit%20Risk%20- %20Clean.pdf

2025 Capital Review Summary of Submissions and Policy Decisions 45 Both submissions advocating for lower agricultural risk weights also stated that our new agricultural risk weights were still too conservative, referring to the lower IRB deposit takers’ risk weights and the fact that agricultural lending is ‘very well secured’ with strong land collateral. ne of these submissions argued that this would continue to distort competition and create a structural disadvantage for standardised deposit takers. However, neither of these submissions, nor the other two that disagreed with our consultation paper proposal (mentioned above) offered additional evidence to support their suggested changes to our proposal – which was based on analysis of previous stress test results, the relative risk profile of agricultural lending comparative to other types of lending, and the wide range of risk weights that IRB deposit takers apply to loans in each of the LVR buckets. Response Aside from the few dissenting submissions described above, our agricultural risk weights proposal was widely supported. We did not receive, and have not found, any evidence to justify changing our proposed approach to agricultural risk weights. Therefore, we are proceeding with introducing the new standardised risk weights for agricultural lending outlined in Table 14. Commercial property Consultation feedback In the consultation paper, we noted that stress tests showed significantly higher losses for aspects of commercial property lending compared to other types of lending. As a result, we suggested that there was potentially scope to add more granular, higher risk weights for these types of lending to more accurately reflect their comparative inherent riskiness. The consultation paper floated the idea of creating a new category for unrated commercial property exposures to allow lenders to more accurately risk weight the lending according to the comparatively high risk it presents. This idea was also floated to address the likelihood of some unrated commercial property lending benefitting from the lower SME risk weights (discussed above), despite commercial property having a different risk profile from SMEs. We suggested the risk weight could be set at a flat rate of 100% for all unrated commercial property lending in the first instance, due to a lack of available data that would allow us to assess a more granular option. However, we also asked deposit takers for more detailed data to be able to assess the appropriateness and potential impact of such a change. The idea of the 100% flat risk weight on all unrated commercial property lending was strongly opposed by most respondents. A key driver of this feedback was that the respondents considered such a risk weight to be too blunt and not sufficiently variable to account for the different levels of risk across different types of unrated commercial property. In addition, some respondents noted that the default risk weight for such lending would already be 100% in the ‘corporate’ category – with the exception of SMEs, where the new 75% or 85% risk weights would apply. These respondents suggested that adding a specific 100% risk weight would therefore add very little to the framework or deposit taker outcomes.

2025 Capital Review Summary of Submissions and Policy Decisions 46 Most respondents supported creating a separate category for unrated commercial property but suggested a range of alternative, more granular, calibrations to the flat 100% risk weight floated in the consultation paper. Most favoured something similar to that used by APRA, shown below. In APS 112, APRA provide the following treatments of commercial property, which is defined as a property exposure that is not a residential property exposure and is not land acquisition, development and construction.27 Table 15: APS 112: Risk weights for commercial property exposures – dependent on property cash flows LVR (%) Risk weight (%) ≤ 60 60.01 - 80 > 80 Standard 70 90 110 Non-standard 150 Table 16: APS 112: Risk weights for commercial property exposures – not dependent on property cash flows Counterparty Risk weight (%) LVR ≤ 60% LVR > 60% or non-standard Rated corporate 60, or the risk weights set by credit rating in APS 11228 Risk weights set by credit rating in APS 11229 All other counterparties 60 According to applicable risk weight in APS 112 Attachment B30

27 Australian Prudential Regulation Authority. (2025). Prudential Standard APS 112 Capital Adequacy: Standardised Approach to Credit Risk. https://www.apra.gov.au/sites/default/files/2024- 07/Prudential%20Standard%20APS%20112%20Capital%20Adequacy%20Standardised%20Approach%20to%20Credit%20Risk%20- %20Clean.pdf 28 Australian Prudential Regulation Authority. (2025). Prudential Standard APS 112 Capital Adequacy: Standardised Approach to Credit Risk. https://www.apra.gov.au/sites/default/files/2024- 07/Prudential%20Standard%20APS%20112%20Capital%20Adequacy%20Standardised%20Approach%20to%20Credit%20Risk%20- %20Clean.pdf 29 Australian Prudential Regulation Authority. (2025). Prudential Standard APS 112 Capital Adequacy: Standardised Approach to Credit Risk. https://www.apra.gov.au/sites/default/files/2024- 07/Prudential%20Standard%20APS%20112%20Capital%20Adequacy%20Standardised%20Approach%20to%20Credit%20Risk%20- %20Clean.pdf 30 Australian Prudential Regulation Authority. (2025). Prudential Standard APS 112 Capital Adequacy: Standardised Approach to Credit Risk. https://www.apra.gov.au/sites/default/files/2024- 07/Prudential%20Standard%20APS%20112%20Capital%20Adequacy%20Standardised%20Approach%20to%20Credit%20Risk%20- %20Clean.pdf

2025 Capital Review Summary of Submissions and Policy Decisions 47 To distinguish between the two categories in these tables, APRA provide a range of circumstances that qualify an exposure to be classified as ‘commercial property exposures - not dependent on property cash flows. This includes circumstances where an authorised deposit-taking institution has recourse to a borrower that meets a tightly limited set of criteria, including ones relating to the diversification of the property portfolio, with limits on geographic and sector concentration. Comment Although deposit takers suggested that the 100% risk weight on unrated commercial property would have little to no impact due to the current standardised risk weight approach, some provided additional information which could be useful in assessing their alternative proposals for introducing further granularity in risk weights for unrated commercial property lending. The key insights from the additional information provided were:  Of those that provided additional detailed data, the vast majority of their commercial property lending had a risk weight under 80%.  Some deposit takers were unable to supply commercial property exposure information by LVR as they did not have this level of detail in their data. Currently, there is no granularity in unrated corporate exposures, which are risk weighted at 100%. As mentioned above, this will change once the new SME and agriculture categories are introduced. Nevertheless, outside of those SME and agricultural exposures, all unrated corporates would be treated the same. However, as highlighted earlier, the risks are not the same across the unrated corporate exposures group. Providing a separate commercial property exposure group would be a way to add further granularity to the standardised approach. It could provide for lower risk weights in circumstances where risk is lower, based on characteristics such as LVR or dependence on cash flows. For example, a simplified approach could be to have a category of commercial property risk weights based on APRA's 'dependent on property cash flows' category shown in Table 15 above. Providing for more granularity is appealing, as losses should be lower for low-LVR lending. However, the main drawback is that we have limited information from which to be able to analyse the impact of such a change. In addition, unlike RMLs, we do not have stress test results at different LVRs for commercial property lending. This means that we cannot verify the assumptions that losses are greater when LVRs are higher, nor do we have a sense of the sensitivity of losses to LVRs. The absence of robust, reliable information increases the risks of making such changes to the standardised framework, especially given the potential magnitude of the changes discussed above. Response We are not proceeding with the proposal to create a 100% risk weight category for unrated commercial property lending. It would likely only add unnecessary complexity to the risk weights framework, while bringing little change in actual outcomes. We are open to considering other changes to standardised risk weights for commercial property lending during 2026, based on:

2025 Capital Review Summary of Submissions and Policy Decisions 48  Deposit takers providing detailed data about commercial property lending, by LVR where possible. We will design a template to collect this information.  An assessment of IRB risk weight outcomes to help us consider if there is scope for standardised changes based on those results. We may also consider carrying out a stress test to assess the potential impacts of any changes. Commercial property lending can be high risk. We will therefore assess all the evidence carefully before committing to any changes in this area. Personal lending Personal lending is captured under ‘other’ exposures in the standardised approach to credit risk weights and has a 100% standardised risk weight. In the consultation paper, we suggested that there could be scope for risk weights to be more granular and moved higher to align with the current approach for Group 3 deposit takers. This would mean creating new categories of secured and unsecured personal lending with 100% and 150% risk weights respectively. Our suggested change was based on an assessment that higher standardised risk weights may better reflect the actual risk of personal lending, which has consistently demonstrated higher loss rates than other lending types in our stress tests. It would also mean that personal lending requirements for Group 3 deposit takers would not change when the DTA comes into force and Group 3 deposit takers transition to the standardised approach. We sought feedback on this suggestion and asked deposit takers for more information about their personal lending exposures to support our analysis. Consultation feedback There was general support for creating new secured and unsecured categories of personal lending. Some deposit takers suggested that there should be additional granularity by also splitting out credit card exposures from other personal lending. However, the suggestion that the risk weight for unsecured personal lending could be increased to 150% was unanimously opposed. In addition to opposing the 150% risk weight that we proposed, there was also some opposition to retaining the 100% risk weight for secured personal lending, with some respondents suggesting this risk weight should instead be reduced. Respondents cited the following reasons:  Standardised risk weights for personal lending would be significantly higher than IRB risk weights, which could have detrimental impacts on competition.  The proposed risk weights would be higher than personal lending risk weights in other jurisdictions.  In general, personal lending risk weights should be lower because there is a weaker historical correlation between personal lending defaults and economic conditions compared to other lending types.  Higher risk weights for personal lending could disincentivise deposit takers from offering this product and/or increase lending rates, forcing New Zealanders outside of the regulated banking system.

2025 Capital Review Summary of Submissions and Policy Decisions 49  The rest of the industry should not be aligned with Group 3 deposit takers because of the different nature of their activities and risk profiles. Several respondents recommended adopting the APRA approach, which has a 75% risk weight for credit card exposures and a 100% risk weight for other retail exposures. A few deposit takers also recommended decreasing the risk weight for secured personal lending to 75% to align with Basel requirements. Comment Given the limited data available, our analysis of personal lending risk weights is mainly based on results from our previous bank industry solvency stress tests. Secured and unsecured personal lending (excluding credit cards) are included together in the ‘other personal lending’ category in stress tests. ther personal lending tends to have significantly higher loss rates than any other lending type, but it also has a slightly lower average standardised risk weight than most other categories. It is difficult to assess personal lending due to the lack of granularity in the stress test data. While overall loss rates are high, the data does not show how losses are split across the secured and unsecured personal lending categories. This means it is not clear whether separate categories for personal lending are necessary or what the relative risk weights should be. However, adopting a 150% risk weight for unsecured personal lending would be significantly out of step with the approach taken in other jurisdictions. Stress test results show that personal lending losses could increase significantly more than losses for other lending types in a severe economic downturn. Deposit takers have not experienced losses on personal lending to this extent in recent years, but it is important to note that New Zealand also has not experienced a downturn as severe as what has been modelled in stress tests in recent years. Although these scenarios are severe, we still consider them to be plausible. The current 100% standardised risk weight for credit card lending (included in the personal lending category) is broadly aligned with other lending types that have similar loss rates in stress tests. However, data provided by some deposit takers showed that credit card exposures are usually treated as unsecured personal lending, so the risk weight would have increased to 150% under our suggestion in the consultation paper. Personal lending tends to be a relatively small part of deposit takers’ balance sheets, so changing risk weights for either Group 2 or Group 3 deposit takers would likely have minor impacts on most entities and on the financial system overall. Material impacts would be concentrated in a small number of deposit takers. Response We have not proceeded with changing our approach to personal lending. Our assessment is that separate secured and unsecured categories cannot be justified based on overall stress test results. More data would also be necessary to justify higher risk weights for unsecured lending or lower risk weights for secured lending. The current risk weight for credit card lending seems appropriate as it is aligned with other lending types of similar risk.

2025 Capital Review Summary of Submissions and Policy Decisions 50 However, since we published our decisions in December 2025, some respondents have asked us to reconsider the settings for secured and unsecured personal lending. In the interests of considering feedback as fully as possible, we intend to collect more data from respondents about this over 2026. We will use this data to assess whether further analysis on personal lending is required. We intend to monitor the impacts on Group 3 deposit takers when they transition to lower risk weights for personal lending under the DTA. If we were concerned about the capital adequacy of individual deposit takers, we could consider overlays for those deposit takers that would be more impacted. We think that this approach is appropriate for addressing potential risks, given that the highest risk is concentrated in a small number of deposit takers. Community Housing Providers and Housing Co-operatives Community Housing Providers (CHPs) are generally not-for-profit groups meeting housing needs through a range of affordable rental and home ownership options. CHPs provide around 16% of all of New Zealand’s social housing places, or around 1 ,000 houses.31 A housing co-operative generally refers to a legal entity with members who gain a permanent right to occupy a dwelling. Commonly, residents are also co-operative members and therefore are able to collectively make decisions about the co-operative and its governance. The evidence we considered in the consultation paper indicated that the current risk weights generally faced by CHPs and co-operatives are high, relative to actual risk. We concluded that this was true for both the standardised approach and the Internal Ratings-Based (IRB) approach to credit risk. A detailed discussion of this assessment is available in the consultation paper. The impact of the existing approach to risk weights is that deposit takers may apply a higher risk weight for this type of lending, relative to actual risk, potentially leading to a higher funding cost for the deposit taker and higher interest rates for borrowers. To manage this, and to produce risk weights closer to the actual risk, we proposed creating a new category of risk weights for CHPs and housing co-operatives. We covered a range of different options, with our preferred option (Option 3) consisting of the following features:  A new standardised category of risk weights to cover lending to CHPs and housing cooperatives.  Both standardised and IRB deposit takers required to use the standardised risk weight.  Separate from RML, but the risk weight would be calibrated to the same level as property investment RML.  Exclude lending by third parties.  Exclude property development lending – only provision of dwellings for accommodation would be covered. Other options that we covered included allowing accredited deposit takers to use their IRB models to calculate risk weights.

31 For more information, see Ministry of Housing and Urban Development. (2025, June 30). The Housing Dashboard. https://www.hud.govt.nz/stats-and-insights/the-government-housing-dashboard/public-homes.

2025 Capital Review Summary of Submissions and Policy Decisions 51 Consultation feedback Most of the feedback supported the general direction proposed in the consultation paper, with the majority supporting our Option 3 approach to create a new separate standardised risk weight category for CHP lending. Respondents that favoured the approach that we proposed generally pointed to some of the following factors:  Current treatment often sees borrowing categorised as Income Producing Real Estate in the prudential framework. This is not accurate for CHPs that are providing rental housing as a social good, or for co-operatives where the owners live in the dwelling.  Risk is not adequately reflected in the current approach as CHP and co-operative lending is treated alongside other, higher-risk lending.  Property development lending should be treated differently from the approach for other lending to CHPs and co-operatives, since the risks during construction and development are different from providing accommodation in finished dwellings.  Respondents considered the changes would help reduce funding barriers. While the points above were supported by most respondents, there were a number of areas where respondents disagreed with our proposal, or suggested changes they considered would ensure the risk of this type of lending was more accurately reflected by the risk weights. These areas of disagreement were:  Some respondents wanted us to allow IRB accredited banks to use their internal models to calculate risk weights for lending to CHPs and co-operatives. These respondents considered that this would allow for better risk differentiation, and therefore better alignment of risk weights with actual risk.  A number of respondents said that the approach for housing co-operatives and CHPs should differ. These respondents highlighted that these two forms of housing are quite different in terms of risk and should not be grouped together for the purposes of risk weighting.  A number of respondents also said that our proposed standardised risk weights were still too high. We proposed that the standardised risk weights for CHPs should match those that apply for the property investment sub-group of RML. Respondents suggested these proposed risk weights would still be too high and overstate the risk of lending to CHPs, especially when the CHP has a long-term funding contract in place with the Crown. They suggested that owneroccupied RML risk weights would be more appropriate, although some suggested that CHP risk weights should be even lower than those.  Some respondents wanted coverage of the new category extended to third parties, where a separate borrower owns the property used for providing housing, as long as that borrower has a contract with a CHP. For example, a third-party borrower might own the property, with a leasing agreement in place with a CHP that allows the CHP to use the property to provide social housing.  A small number of respondents wanted the coverage extended to cover property development for CHP housing.

2025 Capital Review Summary of Submissions and Policy Decisions 52 Comment Community Housing Providers We agree with some of the points raised in submissions, and our response below covers the key changes that we have made to the original proposals after considering the feedback. The most significant change that we have made is that we have concluded that the risk weights in the original proposal were too high. As covered elsewhere in this document, we are now planning to reduce standardised risk weights for low-LVR RML loans by more than we originally proposed. This reflects a general assessment that low-LVR loans are low risk, and we intend to reflect this with further declines in risk weights for RML. Since we have calibrated the CHP risk weights in line with standardised RML risk weights, these reductions will also flow into CHP risk weights. In addition to the change flowing through from lower RML risk weights, we have also made the following changes:  Set CHP risk weights equal to owner-occupied RML risk weights, instead of the higher property investment RML risk weights that we originally proposed.  Set a cap on risk weights at 30% for registered CHPs that have a long-term funding contract in place with the Crown, regardless of the LVR. These two changes are designed to reflect the lower risk associated with CHP lending. After reviewing the evidence provided to us, we agree that the original risk weights that we proposed for CHP lending were too high. These changes will also help provide more granularity, by reflecting the risk reduction inherent in the long-term contracts between registered CHPs and the Crown. The long-term Crown contracts mean that those CHPs would have a stable funding source for a prolonged period of time. We have worked with the Ministry of Housing and Urban Development throughout this work to deepen our knowledge of CHPs and existing approaches to regulation. Lending to CHPs without such a contract would be risk weighted in line with the associated LVR. Borrowing by third parties There are a range of service provision models across the sector: a. CHP owns the property and has a long-term contract with the Crown. b. CHP owns the property and has no long-term contract with the Crown. c. CHP has a joint venture with a government agency where the risk is shared and ownership sits in a separate legal entity, partially owned by the CHP. d. CHP leases dwellings from a third-party private property owner, with a contract in place that ensures the dwelling is available for the CHP for a specified period. Our consultation paper proposals would have applied to A and B but not C and D. Respondents asked us to consider expanding coverage to C and D as these may be growth areas in the future under wider policy direction. These respondents noted that some CHPs lease properties from private property owners, then use those properties to provide social housing.

2025 Capital Review Summary of Submissions and Policy Decisions 53 Our primary concern with expanding the coverage to D relates to the practicalities of including projects connected to CHPs but provided by third-party providers, without inadvertently allowing this favourable treatment to apply to other non-CHP related projects that the third-party provider undertakes. Our assessment is that the approach taken in item C would limit these risks. Some respondents wanted borrowing by all third parties to be included in the new category. We are comfortable with lending to these third parties qualifying for treatment in the new category, but only in a tightly defined set of circumstances. In particular, this lending will only qualify if the third-party borrower includes some form of ownership stake with a registered CHP. To clarify what this means:  There must be a contract in place to ensure that the dwelling is available to the CHP to lease the property for the purpose of providing housing – for example this could be a joint venture or a special purpose vehicle.  In addition to the point above, the borrowing entity must have a legal arrangement that includes an ownership stake held by a registered CHP – the connection with the registered CHP helps ensure that the arrangement is within the system of regulated CHPs and sets up a durable, on-going role for the CHP. Stakeholders will have an opportunity to provide feedback on this approach through the Capital Standard Exposure Draft process during 2026. Housing Co-operatives We have reviewed the feedback about the status of lending to housing co-operatives and have concluded that this lending should not be treated in the same category as lending to CHPs. Respondents pointed to a range of factors that are quite different for co-operatives compared with CHPS, including the ownership stake that residents have in the co-operative, the risk being spread across multiple owners, and the absence of government contracting arrangements. These differences between CHPs and co-operatives could have conflicting impacts on risk. Some factors, such as multiple owners, reduce the risk associated with lending to co-operatives relative to CHPs. However, other differences, such as the absence of long-term government contracts, increase risk relative to CHPs. We have therefore concluded that there are few advantages from treating CHPs and co-operatives in the same way. Our assessment is that lending to co-operatives is generally similar to lending for owner-occupied RML. There are some differences, primarily covering the legal ownership structures and the roles of multiple owners. However, both types of lending are secured by housing, and the lending is for the purposes of the borrower, or in the case of the co-operative, the members of the cooperative, living in the dwelling. This is an important difference compared with property investment RML, where the owner intends to rent the dwelling to a tenant to live in – conversely, the members of a co-operative intend to live in the house that they are borrowing from the deposit taker to buy. We have concluded that the simplest way to manage lending to co-operatives is for lenders to classify the lending as owner-occupied RML, provided that it meets the RML definition. We will be seeking feedback on this during the exposure draft changes to Banking Prudential Requirements (BPR) documents in 2026.

2025 Capital Review Summary of Submissions and Policy Decisions 54 For deposit takers required to use the standardised approach to calculate risk weights for RML, classifying lending to co-operatives within that category should be straightforward. Our assessment is that a co-operative ownership structure is not a barrier to classification as RML, provided the other aspects of the RML definition are met. Nevertheless, if there are aspects of the lending that are outside of the RML classification, then the deposit taker should not use the RML category. For deposit takers using the IRB approach to calculate risk weights, the classification of lending to co-operatives may be more complex. We have no concerns about lending to co-operatives being classified as RML. When the lending meets the RML definition we encourage the deposit taker to use that category. However, we do acknowledge the use of IRB models may add some complexity. We are open to IRB-accredited deposit takers using their approved RML models to calculate risk weights for lending to housing co-operatives. However, in some circumstances it may not be practical, or even possible, to use these models. For example, it is possible the deposit taker will not have access to the required variables for lending to the co-operative, meaning their model will not be able to calculate a risk weight. Deposit takers will need to consider on a case-by-case basis whether any particular loan to a co-operative can be modelled within their approved model, taking into the account the requirements that RML models must cover only homogenous, pooled loans. If a deposit taker determines that any particular loan to a co-operative cannot fit into their IRB model, but is an RML, then they should use the standardised RML risk weight instead. This will mean that lending to co-operatives will be classified as RML, provided it meets the definition for that lending. Response The table below summarises the changes that we have made, reflecting the analysis above. Table 17: Summary of changes for housing co-operatives and CHPs Feedback RBNZ response Housing co-operatives and CHPs should not be included in the same category Lending to co-operatives will not be included in the same category as lending to CHPs. There will be a new specific category for lending to CHPs. Lending to co-operatives will be classified as RML, provided it meets the RML definition. IRB-accredited deposit takers should be able to use IRB models to calculate risk weights Deposit takers will need to determine whether their IRB model approach to RML can also be applied to lending to co-operatives. If not, they should use the RML standardised approach. Deposit takers will not be able to use internal models for lending to CHPs. Our assessment is that standardising the method is the simplest approach and ensures that all deposit takers are treated the same way; granularity from internal models is unlikely to add valuable information but would add complexity.

2025 Capital Review Summary of Submissions and Policy Decisions 55 Feedback RBNZ response Proposed risk weights in line with property investment RML are too high We agree that the risk weights proposed in the consultation paper are too high. After reviewing the evidence in the submissions, we have decided to set the standardised risk weights for the new category for lending to CHPs in line with the standardised risk weights for owner-occupied RML. There will also be a risk weight cap – described in the row below – that will apply in some circumstances. Lending to co-operatives will be treated the same as owner-occupier RML. Proposed risk weights do not have sufficient granularity To increase granularity for lending to CHPs we have also introduced a risk weight cap of 30%, regardless of the LVR of the loan. However, this will only apply to lending to CHPs that are registered with HUD and have a long-term funding contract with the Crown. Borrowing by third parties should be covered Third-party borrowing will qualify, if the third-party borrower includes some form of ownership stake with a registered CHP. Property development borrowing should qualify Property development borrowing will not qualify, as we consider property development to have different risks to finished dwellings. Other risk weights feedback – Whenua Māori We sought feedback about barriers to lending when whenua Māori is used as security, as well as any proposals for changes to prudential regulation to address barriers. Respondents identified a handful of barriers, mainly relating to challenges around collective ownership, and concerns about how to manage that during default, rather the prudential regulatory settings. Some also suggested that lending for whenua Māori should have a separate, low risk weight category of lending. Aside from risk weights, no respondents identified prudential actions that would reduce barriers to lending. In addition, we have not found any compelling evidence to indicate lending with Whenua Māori as security is lower risk than any other sort of lending and do not see a strong rationale to reduce risk weights for this lending. However, lending where whenua Māori is used as security will, provided it meets other definitions in the relevant categories, be eligible for the revised risk weights on the other categories covered in our final decisions, such as RML and the new corporate sub-categories. However, the general approach to lower risk weights for CHP lending will be available to lending secured by whenua Māori, so long as the lending is otherwise compliant with CHP definitions, which we will develop next year. In addition, we recommend that this topic remains open, especially for consideration of any guidance gaps that might be a useful step in the future. We continue to be open to suggestions regarding regulatory guidance or other steps that can be considered to address barriers for lending with whenua Māori as security.

2025 Capital Review Summary of Submissions and Policy Decisions 56 Other risk weights feedback – Securitisation Consultation feedback A number of respondents commented on the role of securitisation. Facilitating securitisation funding – particularly through warehouse facilities – for smaller deposit takers and non-bank lending institutions (NBLIs) is one mechanism by which Group 1 deposit takers support the development of competition in New Zealand’s financial markets. Securitisation, especially through warehouse facilities, represents a significant funding source for NBLIs in New Zealand. Respondents noted that Group 1 deposit takers commonly provide senior lending to NBLI via warehouse facilities, which are structured similarly to revolving credit facilities. In many cases, these protect the warehouse provider in similar ways to wholesale market ‘term’ securitisations (e.g. the collateral exceeds the loan amount). Respondents noted that our current requirements do not provide specific capital treatment for exposures to third-party warehouse facilities, focusing solely on the context of bank-originated securitisations. Currently, as a result, Group 1 deposit takers providing funding to unrated warehouse facilities assign risk weights to these exposures as if they are corporate exposures, generally resulting in a standardised weight of 100%, subject to an output floor of 85%. Respondents stated that this does not accurately reflect the underlying credit risk of such exposures. They stated that while the warehouse facilities are generally not externally rated due to the costs and operational demands of obtaining and maintaining ratings, the underlying quality of the exposures would qualify for substantially lower risk weights, possibly as low as 20%. As noted above, the cost of holding the additional amounts of capital under the current standards to fund these exposures is passed on to the NBLIs though either or both reduced funding and higher pricing. This, in turn, affects efficiency and competition. Respondents suggested that, as in offshore jurisdictions like Australia, New Zealand should require warehouse funding deposit takers to hold capital based on the risk of the underlying exposure and structure, rather than on the credit risk of the originating deposit taker or NBLI borrowing entity. Response It was not possible to address this issue in the time available. However, we intend to review this during the next 12-18 months. Securitisation is potentially a wide-ranging topic. Our proposed review will be tightly constrained to the issues described above regarding requiring warehouse funding deposit takers to hold capital based on the risk of the underlying exposure and structure, rather than on the credit risk of the originating deposit taker or NBLI borrowing entity. This is much more limited in scope than issues canvassed in the Ernst and Young (EY) Report titled “External Perceptions of the New Zealand Banking ystem,” 32 which we have released alongside this Response document. The EY paper suggested considering a wider range of topics, including encouraging risk transfer from deposit takers to investors and/or creating a New Zealand Freddie Mac/Fannie Mae. We have not considered these options for inclusion in the next stages of the work. These topics will be outside of scope of our more limited focus on securitisation, due to the complexity involved.

32 https://www.rbnz.govt.nz/-/media/project/sites/rbnz/files/research/external-perceptions-on-the-new-zealand-banking-market.pdf

2025 Capital Review Summary of Submissions and Policy Decisions 57 Therefore, at this time, we do not intend to carry out a review of existing approaches to securitisation, beyond the limited warehousing-related point described above. For example, we will not be reviewing our approach to securitisation requirements relating to circumstances where a deposit taker is originating or supplying assets to a special purpose entity, or other legal structure, for the purpose of securitisation and risk transfer, with consequent reductions in risk weights. We intend to carry over existing requirements in these areas into the DTA Standards, as currently set out in BPR160: Insurance, Securitisation, and Loan Transfers. 33 Other risk weights feedback – Infrastructure Consultation feedback A number of respondents identified a gap in our framework for the treatment of borrowing for infrastructure projects. This is relevant for both the standardised and IRB approaches to credit risk. Respondents asked us to consider a concessionary LGD for the development of public infrastructure projects, similar to Australia. APRA allows for a 25% LGD for operators of domestic large public infrastructure projects and utilities that have a tripartite agreement with the Australian Government. A lower LGD recognises that the underlying infrastructure assets support recovery values in the event of a default. Respondents stated that borrowers who deliver large-scale public infrastructure, or who operate lifeline utilities (such as water, energy distribution and transmission, gas pipelines and fibre) are backed by assets with strong credit enhancement frameworks such as Public Private Partnerships and the Infrastructure Funding and Financing Act 2020. These are assessed as leading to a reduction of loss in default. Additionally, where borrowers operate under a Regulated Asset Base, regulatory oversight and pricing mechanisms help mitigate default risk, while the essential nature, and defensible value of the assets, further reduce potential losses. Their services are nondiscretionary, demand is stable across economic cycles, and these assets are highly secure. This reduces default risk and supports the case for a lower standardised risk weight. Response While we agree that lending with the infrastructure features described above will have lower risk than lending without these characteristics, we have not had sufficient time to fully assess the risks or to identify how to incorporate this into the existing risk weights framework. We will address this during 2026. If possible, we will complete this during the amendment of the BPRs in 2026. This would include both the consideration of a concessionary LGD in the IRB framework and creating an equivalent standardised category of exposures. It may not be possible to complete this work in time for the 2026 revisions to the BPRs, in which case we may consider it at a later time, to include in the Capital Standard in 2028. Minor and technical issues In this section, we address certain discrete technical topics that were included in the consultation or that have been raised by respondents.

33 BPR160 Insurance, Securitisation and Loan Transfers [Final version]: https://www.rbnz.govt.nz/- /media/project/sites/rbnz/files/regulation-and-supervision/banks/banking-supervision-handbook/bpr160-insurance-securitisationand-loan-transfers-1-july-2024pdf.pdf

2025 Capital Review Summary of Submissions and Policy Decisions 58 Table 18: Minor and technical issues for risk weights Issue RBNZ Response Group 1 Unsecured LGD used in IRB calculations for non-retail lending Some respondents asked us to review the unsecured LGD used in IRB calculations for -retail lending by New Zealand IRB banks non We do not currently intend to review this. We have assessed this to be lower priority than the other topics that we have committed to assessing in 2026. Assess whether the long-standing minimum LGDs for RML and farm exposures are still suitable We do not currently intend to review this. We are not aware of any new information that suggests the minimum LGD needs to be reviewed. Additional topic that we intend to cover in 2026 The sections above cover a range of components of risk weights that we intend to assess further, including those for secured personal lending, infrastructure, commercial property and elements of securitisation. In addition, we also intend to look at a subset of the approach to credit conversion factors. These factors convert off-balance sheet commitments into a credit-risk equivalent amount so they can be included in risk-weighted assets. A small range of these can be set at zero in certain circumstances. We intend to assess the evidence regarding the risks associated with the 0% category and whether this is adequately captured in the existing approach. 2.1.6 Counter-cyclical capital buffer (CCyB) In the 2019 Capital Review, we decided to include a CCyB as part of the PCB for Group 1 and 2 deposit takers and that it would come into effect from 2028. Unlike other components of the capital stack, the CCyB is part of our macroprudential policy toolkit and is designed to be adjusted during the financial cycle to take account of the systemic risks present in the financial system. In 2019, we also decided that the CCyB should be administered using an ‘early-set’ strategy. This means that the CCyB would be set at its long-run level (i.e., the rate we would set the CCyB at most of the time), and it would only be reduced or removed during periods of systemic stress to encourage deposit takers to continue lending. We did not expect to raise the CCyB above its longrun level, instead relying on other tools to mitigate the build-up of systemic risk. We continue to view the CCyB as an important part of our future macroprudential toolkit and proposed that the CCyB form part of the PCB for Group 1 and 2 deposit takers. However, with other components of the capital stack being reviewed, we also reviewed the proposed CCyB settings (due to come into effect in 2028).

2025 Capital Review Summary of Submissions and Policy Decisions 59 Long-run CCyB set at 1% In the consultation, we proposed setting the long-run level of the CCyB at 1% of RWA, rather than 1.5%, to reflect that the PCB will be smaller for Group 1 and 2 deposit takers under the options proposed for the capital stack. Consultation feedback There were mixed views on our proposal to reduce the long-run CCyB from 1.5% to 1% of RWA. Some respondents suggested that the long-run CCyB should be set higher than 1%. They suggested that a CCyB of 1% (reduced in a downturn) would not likely give deposit takers enough incentive to extend more credit during a downturn (and hence support economic recovery) and that a higher long-run CCyB gives additional flexibility in responding to a range of shocks. Further, if the long-run CCyB was higher, it was suggested that the PCB should not also increase (i.e., the capital conservation buffer component of the PCB should be reduced to reflect a higher longrun CCyB). Some respondents agreed with a long-run CCyB of 1% (which aligns with Australia). However, they stated that more guidance is needed under which conditions the settings of the CCyB would be reduced during a period of stress and how the CCyB would then be restored to its long-run level after a period of stress (e.g., clarity that the CCyB was being reduced for an extended period and on what timeframe it would be restored). A key consideration in providing guidance will be how the CCyB interacts with the broader capital framework. Comment We acknowledge the feedback that a long-run CCyB of 1% may not provide deposit takers enough incentive (if reduced to 0%) to extend more credit during a downturn and agree that a higher long-run CCyB gives additional flexibility in responding to a range of shocks. However, if the CCyB makes up a large proportion of the PCB, cutting it to 0% may allow a deposit taker to reduce its capital to a level that undermines its safety and soundness. Alternatively, a deposit taker may not make use of the entire CCyB (if reduced to 0%) as it may not wish to get too close to minimum capital requirements. We also note that given the structure of the capital stack (decided as part of this review), if the long-run CCyB was greater than 1%, then reduced to 0%, Group 2 deposit takers could end up with lower capital requirements than Group 3 deposit takers (in percentage terms), which would be inconsistent with proportionality. Response We will proceed with our proposal to set the long-run CCyB at 1% (which is intended to come into effect from December 2028) and will consider how the CCyB interacts with the broader capital framework at different stages of the financial cycle. We may consider this as part of guidance for the Capital Standard or other associated guidance. We agree that there is merit in providing guidance on the conditions under which the CCyB would be reduced, then restored to its long-run level. We will consider this as part of the updated macroprudential policy framework document that we intend to publish in mid-2026.

2025 Capital Review Summary of Submissions and Policy Decisions 60 Not applying the CCyB to Group 3 deposit takers In the consultation, we proposed to not apply the CCyB to Group 3 deposit takers (i.e., it will not be included in their PCB). This would be consistent with our approach to not apply other macroprudential tools to Group 3 deposit takers (such as restrictions on high loan-to-value ratio and high debt-to-income ratio residential mortgage lending). Consultation feedback There were mixed views on our proposal to not apply the CCyB to Group 3 deposit takers. Some respondents supported the proposal, agreeing that it would not have a large impact on the financial system overall and would be consistent with our approach to using other macroprudential tools. However, a few respondents disagreed, as applying the CCyB to Group 3 deposit takers would provide these deposit takers with some relief to capital requirements during a downturn. It was also suggested that not applying the CCyB to Group 3 deposit takers may lead to a competitive imbalance as Group 1 and 2 deposit takers would have more scope to continue to lend during a downturn (if the CCyB is reduced), whereas Group 3 deposit takers may not. Comment Under the DTA Lending Standard, we do not intend to apply other macroprudential tools such as loan-to-value ratio and debt-to-income restrictions to Group 3 deposit takers. Consistent with that approach, in the consultation we proposed to not apply the CCyB to Group 3 deposit takers. Given that the Group 3 sector is small and not as interconnected as other sectors, the impact on aggregate lending (and therefore overall financial and macroeconomic conditions) would be small if we applied a CCyB to Group 3 deposit takers. We acknowledge the point that applying the CCyB to Group 3 deposit takers may provide them with some relief to capital requirements during a downturn. However, it is noted that many Group 3 deposit takers are less able to increase their capital levels quickly (for example, by issuing capital instruments or retaining profits) compared to most Group 1 and 2 deposit takers. This could be particularly challenging after a downturn when restoring the CCyB to its long-run level. Response We will proceed with our proposal to not apply the CCyB to Group 3 deposit takers. 2.2 Other topics raised We also received feedback on the following topics:  Design of loss absorbing capacity  Alignment with Australia  Output floor and scalar  Leverage ratio  Cost Benefit Analysis (CBA)  DCS Levy

2025 Capital Review Summary of Submissions and Policy Decisions 61 2.2.1 Design of loss absorbing capacity (LAC) In the consultation paper we proposed introducing new requirement for internally issued LAC for Group 1 deposit takers under Option 2. This would be 6% of risk-weighted assets in instruments designed to absorb losses at the point of non‑viability (i.e., written off or converted into CET1). LAC instruments would be issued to the Group 1 deposit takers’ parent – we call this ‘Internal AC’. Tier 2 capital instruments for Group 1 deposit takers would have the same write-down and conversion features as LAC instruments and would also be required to be issued internally. Consultation feedback We received a wide range of views about the role of LAC in the New Zealand prudential capital framework. Some of the feedback supported LAC, some were opposed, and some pointed to a range of risks. Supporters of LAC suggested that the additional complexity of LAC had been overstated while those against LAC suggested LAC was untested and may have limited crisis management benefits in practice. Group 1 deposit takers supported the LAC option (Option 2) over the non-LAC option (Option 1) though many noted that their support was conditional on LAC and revised Tier 2 instruments satisfying APRA rules (e.g. being compliant with Basel/Financial Stability Board (FSB) Guidance) compared with current Tier 2 which is not. Some respondents asked us to consider transitional arrangements for Tier 2 instruments for Group 1 to cover the period between now and when the Capital Standard starts. Some respondents suggested that we should allow for external issuance of Tier 2 and LAC by Group 1 deposit takers rather than internal only, firstly to maintain a greater diversity of funding sources, and secondly to apply the LAC requirement to a future Group 1 deposit taker who is not wholly owned by an Australian parent. One respondent suggested that APRA-eligible Tier 2 issued by Australian subsidiaries of a New Zealand bank should also be accepted as capital for the group. Some Group 2 deposit takers however preferred the option without LAC. This was mostly due to concerns about the use of internal LAC potentially acting as a barrier to domestic banks moving to Group 1. Some respondents, particularly from Groups 2 and 3, were concerned about potential reduction in the depth of the market for Tier 2 instruments for Group 2, if existing Tier 2 instruments currently issued by Group 1 deposit takers were removed from New Zealand debt markets. One respondent commented on potential pricing of internal LAC and the cost of external LAC issued by an Australian parent. One respondent argued that Open Bank Resolution (OBR) had been positioned as an alternative to bail-in capital, and that if LAC is introduced then OBR should be reviewed, and in their view, scrapped. All our independent experts raised a number of concerns about the LAC option and two explicitly advised against this option. Their concerns include the need to confirm the design of LAC instruments, the risks that bail-in may not work as expected, the reduction in the depth and liquidity of markets for deposit takers’ Tier 2 capital instruments and the fact that our crisis management framework is still under development.

2025 Capital Review Summary of Submissions and Policy Decisions 62 Response Readers are encouraged to cross-reference these relevant sections in this document for additional context:  Section 1.2 Transitional arrangements for a consultation on a LAC implementation timeline  Section 2.1.3 Capital stack options for a consultation on a detailed LAC design Crisis management benefits of internally issued LAC and Tier 2 instruments We acknowledge that there is some uncertainty on the efficacy of LAC. However, overall, we expect the higher expected cost of crisis that might result from moving to a framework incorporating LAC to be smaller than the benefit of lower lending rates to New Zealanders in the LAC option. See the full Cost Benefit Analysis34 for details. All existing Group 1 deposit takers in New Zealand have an Australian parent bank. Amongst other things, the LAC option is designed to support our preferred “single point of entry” ( PE) model for recovering or resolving a Group 1 deposit taker. An SPE model is designed to keep the group together rather than recovering or resolving the New Zealand deposit taker on a standalone basis.35 We will therefore require these Group 1 deposit takers to issue internal LAC instruments to their parent. Tier 2 capital instruments for Group 1 deposit takers will have the same write-down and conversion features as LAC instruments and will also be required to be issued internally. Most countries (including Australia) use convertible loss absorbing instruments, especially for their largest banks. There is a potential concern that converting LAC instruments in a crisis could worsen declines in market confidence. However, because New Zealand’s largest banks are all subsidiaries and the LAC will be internal, rather than held by external investors, our assessment (built into the CBA) is that these concerns are more limited in New Zealand. See below for our further response on internal vs external LAC. We currently do not intend to introduce a LAC requirement for Group 2 or Group 3 deposit takers given the nature of their business and, the potential availability of other, more appropriate recovery and resolution options that do not rely as heavily on a recapitalisation. We are not changing any of the current Tier 2 requirements for Group 2 and 3 deposit takers as part of the 2025 Review. However, potential changes are being considered separately as part of the implementation of our crisis management framework under the DTA. Any proposed changes would be subject to consultation. The future role of OBR is being considered as part of our broader work on the implementation of the new crisis management regime under the DTA.

34 CBA: https://www.rbnz.govt.nz/-/media/project/sites/rbnz/files/regulation-and-supervision/banks/capital-review/2025/cost-benefitanalysis.pdf 35 While the SPE model is our preferred approach for dealing with a distressed Group 1 deposit taker, in the coming years, other resolution approaches will need to be developed and prepositioned in order for the Reserve Bank to fulfil its future responsibilities as “resolution authority” under the DTA. The Reserve Bank’s approach to resolution of deposit takers will be consulted on in the coming years as part of the requirement to publish a Statement of Approach to Resolution under the DTA.

2025 Capital Review Summary of Submissions and Policy Decisions 63 Detailed design of internal LAC and Tier 2 instruments We will design internal LAC and Tier 2 instruments to meet two key outcomes:  LAC operates effectively to recapitalise a distressed deposit taker under a range of crisis scenarios, recognising however that there will always be uncertainty in how a crisis will unfold in practice.  LAC qualifies for Tier 2 corresponding deductions under APRA rules. This was a key point made by Group 1 deposit takers in their submissions due to its impact on the flow-on costs facing each group. This means that LAC requirements will follow international guidance. 36 We intend to consult on the detailed design of Tier 2 and LAC requirements for Group 1 deposit takers in June 2026. Through the design process, we will continue working with APRA. Current Tier 2 does not have conversion or write off features (Refer to section D3.4 of BPR110 Capital Definitions). This means that current Tier 2 does not meet APRA requirements for an eligible deduction against Tier 2 capital, such that NZ Tier 2 held by the Australian parent bank is required to be deducted from CET1 capital for the APRA-authorised legal entity on a standalone (“ evel 1”) basis. Being eligible for a deduction against Tier 2 may allow the Australian parent bank to hold proportionately more Tier 2 capital and less CET1 capital which would reduce the overall funding costs of the group. We agree that this is an important feature of the intended design of internal LAC and Tier 2 instruments and have been engaging with APRA to ensure this can be achieved. On transitional arrangements for Tier 2 instruments for Group 1 deposit takers, there are a range of ways to manage the period between now and the new requirements. It includes allowing Tier 2 to be issued with a shorter maturity date. We will address this directly in the consultation for proposed changed to BPRs in March/April. Views of Groups 2 and 3 If domestic banks were to grow and move to Group 1, we could manage the circumstance through variations in each deposit taker’s conditions of licence, incorporating the specific features of that deposit taker. See Section 2.1.3 for details and our response on the capital ratio requirements. Accepting APRA-eligible Tier 2 issued by Australian subsidiaries of a New Zealand bank as capital for the group would largely be relevant for Group 2 banks with Australian operations. We did not directly consider allowing APRA-eligible Tier 2 issued by Australian subsidiaries of a New Zealand bank to be accepted as capital for the group as we did not propose changes to Tier-2 instruments for Group 2 and want to carry over the current rules at this stage. Nevertheless, we are open to considering this alongside our resolution strategies during the exposure draft process for the Capital Standard and will invite further views from stakeholders at that point.

36 The Financial Stability Board has published the Guiding Principles on Internal Total Loss-Absorbing Capacity of G-SIBs ('Internal T AC'). To read the BI ’s summary of the guiding principles, see Bank for International ettlements. (2020). Internal T AC – Executive Summary. https://www.bis.org/fsi/fsisummaries/internal_tlac.htm

2025 Capital Review Summary of Submissions and Policy Decisions 64 Impact on NZ debt issuance and internal LAC pricing As of December 2025, the New Zealand Debt Market (NZDX) 37 lists 145 debt securities with approximately $55.5 billion outstanding. This includes 19 issuances by Group 1 deposit takers totalling $11.6 billion and nine issuances by Group 2 deposit takers totalling $1.7 billion. As a result of the capital review decisions, Group 1 deposit taker issuance outstanding is likely to fall away over time, which will reduce the amount of subordinated debt issued by deposit takers to the New Zealand market. We will monitor the effects of our decisions and publish our findings to ensure that we have calibrated our settings correctly for New Zealand. This will include monitoring of trends in amounts and prices of capital instruments issued, whether costs of capital are tracking in line with our expectations, trends in lending rates (by sector), deposit taker’s profits and return on equity. The changes will be implemented over several years to give deposit takers time to adjust. The new settings will need to be in place for a period of time before we can accurately assess their impacts. Therefore, our first monitoring report would likely be in 2028 to allow deposit takers to transition. Internal vs external LAC Where a Group 1 deposit taker is wholly owned, as they all currently are, internal LAC is our preference over externally held LAC due to the simplified process for recapitalising a distressed deposit taker through triggering internal LAC. There are a number of advantages to internal issuance of LAC within a wholly owned group, Firstly, it does not impose additional losses on that group as the consolidated group-level exposure is unchanged. Secondly, it reduces the number of counterparties affected by LAC being triggered, and thirdly, it minimises ranking complexity among different LAC instruments. If we also allowed external LAC under an internal LAC model, these benefits would be undermined. For the avoidance of any doubt, our decisions do not prevent a deposit taker issuing wholesale debt that is senior to internal LAC. 2.2.2 Alignment with Australia Consultation feedback Major banks suggested an alternative version of Option 2 where the amounts of capital and LAC were reduced to the Australian levels (18.25% in total) would be superior to our proposal. These respondents tended to argue this would be less costly, while still sufficient in financial stability terms, and also suggested alignment with Australia was inherently a good thing. Similar points were made by some other respondents. ome technical aspects of this feedback were around ‘compatibility’ between Australian and New Zealand capital requirements. This included the desirability of not having external instruments (e.g. AT1) that counted as NZ capital but could not count as capital for the wider group.

37 https://www.nzx.com/markets/NZDX

2025 Capital Review Summary of Submissions and Policy Decisions 65 Comment We considered this feedback carefully but continue to hold a view that capital ratios higher than those of Australian deposit takers are appropriate for New Zealand deposit takers. This is partly based on the cost-benefit analysis38. We also noted that:  The largest Australian banks currently have total capital ratios well above the 18.25% level (around 21% on average).  New Zealand also has different risks to Australia, and as a result, the optimal ratio is likely to be higher. For example, the Standard and Poor’s Banking Industry Country Risk Assessment has a higher risk classification for New Zealand versus Australia.  We also face different dynamics as a “host” country (i.e. our banking system is dominated by subsidiaries of foreign banks), which generally results in higher capital levels than those set in “home” countries (like Australia). We agree there are advantages to having capital instruments in New Zealand that are either able to count as group capital (this bolsters the case for removing AT1 instruments), or are able to be issued internally to the parent without the parent facing excessive costs (this appears to be the case for internal LAC instruments). Response As set out above in section 2.1.3 above, we have adopted Option 2 and are comfortable with how that aligns our settings with APRA’s rules for Australian deposit takers. 2.2.3 Output floor and scalar The output floor is a part of the capital framework that sets a lower bound on IRB banks’ credit risk RWAs. Under current rules, total credit risk RWAs calculated using the IRB approach cannot go below 85% of the equivalent standardised approach RWAs for these exposures. The IRB scalar is a multiplier that is applied to RWAs calculated using IRB models. The purpose of the scalar is to adjust overall IRB outcomes, without adversely affecting the risk sensitivity benefits of IRB modelling. The scalar was set at 1.2x as part of the 2019 Review, to reduce the growing divergence in overall outcomes that had been occurring between the two approaches. Consultation feedback There was a divergence in views about the current output floor and scalar calibrations in stakeholder feedback. Group 1 deposit takers generally wanted a reduction of the output floor and/or scalar to align with APRA’s settings (72.5% output floor and 1.1 scalar). Group 2 generally wanted the output floor increased to 100% to support proportionality and competition.

38 CBA: https://www.rbnz.govt.nz/-/media/project/sites/rbnz/files/regulation-and-supervision/banks/capital-review/2025/cost-benefitanalysis.pdf

2025 Capital Review Summary of Submissions and Policy Decisions 66 There were a range of arguments put forward against lowering the output floor and scalar from their current settings:  A lower output floor and scalar would likely overplay the extent of credit risk management benefits of IRB modelling and lead to a larger divergence in capital requirements for similar lending.  Lowering the output floor would create a larger discount for IRB compared to the standardised approach, which in the NZ context means a detrimental impact on proportionality between Group 1 and Group 2 and 3 deposit takers.  The proposed standardised risk weight changes already provide an effective relaxation of the output floor (on its own an approximately 6% reduction in capital required for Group 1). Equally, there are arguments supporting a lower output floor and scalar:  A lower (or unchanged) output floor and scalar would continue to support the incentive to undertake more granular risk assessment that results from IRB modelling and improve New Zealand’s consistency with peer regulators’ settings. Response We decided to retain the current 85% output floor and 1.2 scalar. These settings were examined in detail in the 2019 Review and have been calibrated to a level that we consider is appropriate for New Zealand circumstances. It maintains an incentive for IRB modelling but limits the divergence in capital requirements for similar lending – which we consider appropriate in the New Zealand context, when compared to the use of IRB models by banks in other jurisdictions. In comparison, the four accredited banks here operate relatively simple business models, and there is a lesser level of resourcing at banks and the RBNZ to robustly develop, operate and scrutinise IRB models. We are open to standardised banks applying for IRB accreditation. As part of the new approach under the Deposit Takers Act 2023, we will publish a notice covering the process for those deposit takers applying to be accredited for using IRB models. This process will not be restricted to Group 1, and other deposit takers can apply, as is the case now under current regulations. Nevertheless, IRB modelling requires a depth of data and sophistication in systems that may not be practical for smaller deposit takers. It is a resource intensive process, with current IRB banks in New Zealand being able to benefit from the expertise and support available from their large Australian parents. 2.2.4 Leverage ratio Many countries use an additional set of capital rules that do not require assets to be risk weighted. This means the requirement is that capital is a minimum proportion of exposures (a “leverage ratio”). The Reserve Bank does not use leverage ratio requirements and did not consult on the possibility of adding them. This is because the other capital rules in New Zealand mean that average risk weights are relatively high, so that a leverage ratio would add administrative complexity without being likely to bind on any banks. Consultation feedback Respondents did not advocate for a leverage ratio, but the international experts’ reports mentioned the absence of a leverage ratio. One expert advocated for reconsidering this or at least

2025 Capital Review Summary of Submissions and Policy Decisions 67 checking to make sure that the proposed changes to risk-weighted capital requirements did not reduce capital requirements to the point where a leverage ratio would be binding. Response We will not adopt a leverage ratio requirement as part of this Review. But as suggested by the international experts, we will reconsider the possibility of a leverage ratio and check on the extent to which it might bind on NZ firms in the future. This would likely be in 2027 at the earliest, once the majority of the implementation of this Review’s decisions are complete. 2.2.5 Cost Benefit Analysis Throughout this targeted review we have sought feedback and challenge from multiple sources on the CBA, specifically:  Industry and industry experts through our consultation document; 2025 Review of key capital settings – Policy proposals for feedback;  The international experts through workshops and their final reports; and  The Treasury, through regular engagements. Overall, the feedback was highly constructive, and we thank everyone who engaged with us. There was a range of views provided on the CBA, at times these views were in direct conflict. Nevertheless, we consider the amendments we have made and the outputs from the CBA valuable. The final CBA can be found on our website39 . The use of the 2019 models, with adjustments Consultation feedback At a high level there were a range of views on the use of the 2019 models. Some industry commentators questioned the use of the models, concerned they fail to take into account the full cost of overregulation and the flow on (or dynamic) negative impacts that excess capital has on the New Zealand economy. In submissions, we received feedback that alternative approaches to assessing the costs of capital requirements suggested there were substantially higher economic costs from its prudential regulation than the RBNZ had modelled, in the order of $10-14bn (ongoing, annual cost to the economy). Other respondents supported the CBA highlighting the high degree of transparency and thought the Reserve Bank should place higher importance on the results. The international experts found the 2019 CBA broadly balanced and questioned whether there are grounds to depart from current settings.

39 CBA: https://www.rbnz.govt.nz/-/media/project/sites/rbnz/files/regulation-and-supervision/banks/capital-review/2025/cost-benefitanalysis.pdf

2025 Capital Review Summary of Submissions and Policy Decisions 68 There was industry feedback that the use of the fully implemented 2019 policy decisions as the counterfactual was confusing when trying to analyse what this meant for capital levels and lending rates. Respondents suggested that current settings (2025 actuals) were a better comparator. Respondents requested the CBA should include more distributional or incidence analysis, that is, where the benefits, costs and other impacts would lie if capital requirements changed. For example, to what extent would lending from Group 2 increase by, or would the costs impact savers while the benefits support borrowers. Response We recognise that the models used, and the results provided by the CBA models, are stylised and highly dependent on a handful of parameters. Nevertheless, we continue to see value in using and producing results from the CBA models. The models are based off credible well-recognised studies, for example, the ‘cost of crisis’ model uses an asymptotic single risk factor model which is used widely by internal ratings-based banks globally. The ‘lending rate’ model has been reviewed by international experts in 2019, and internally in 2024.40 With the time available, we reviewed the alternative approaches provided by submitters. Overall, reviewing alternative models helps us to critically review the methods and parameters that we use. However, we remain confident that our estimates of the change in output from the options analysed are reasonable. We discuss alternative CBA approaches further in Appendix 2 of the final CBA. Respondents were at times confused by the use of the 2019 policy decisions as the primary comparator for the CBA. We have therefore sought to include 2025 actuals where possible, however, we continue to use the fully implemented 2019 policy decisions (at times referred to as the “2028 settings”) as the primary status quo. This is consistent with CBA guidance41 and would be the regulatory settings if no policy decisions were made. Finally, we continue to expand the models to take into account the impact non-Domestic Systemically Important Bank requirements have on the cost of crisis, as well as providing additional Group 2 and Group 3 outputs and expectations from the models, including the dollar amount of capital required and RWAs. We continue to rely heavily on data associated with Group 1 deposit takers for our lending model; this is primarily because of the availability of data but also due to their market share. Nevertheless, we do recognise assumptions applying to Group 1 deposit takers may not apply to Group 2 and 3 deposit takers, for example, changes in the cost of equity. Assumptions used - Lending rates The method consulted on used the same framework as in 2019. Specifically, the benefit of lower capital requirements was calculated by estimating the amount of cost savings deposit takers are expected to generate from replacing expensive equity financing with relatively less expensive debt financing (a Weighted Average Funding Cost (WAFC) is calculated). We then assume the cost savings generated are passed on to borrowers through lower lending rates.

40 Tanuvasa, W, Downing R, Martel, J, 2024. Biennial Assessment 2023 - Monitoring Capital Review Implementation, Reserve Bank Bulletin. 41 The Treasury, Guide to Social Cost Benefit Analysis, July 2025. Guide to Social Cost Benefit Analysis - July 2015.

2025 Capital Review Summary of Submissions and Policy Decisions 69 Consultation feedback There were mixed views on the extent to which capital requirements impact lending rates. Consistent with past feedback we have received, respondents from industry suggested we should be more optimistic regarding the impact lowering capital requirements has on lending rates. Medium-sized deposit takers pointed to empirical evidence that shows returns on equity are highly insensitive to risk for medium-sized deposit takers.42 Conversely, in discussions with our international experts they raised concerns that our estimates are too optimistic regarding the impact lower capital requirements have on lending rates. Their arguments are based on two suggestions:  The MM theorem holds; that equity funding is highly sensitive to changes in risk – meaning funding costs should only reduce to the extent there is a tax advantage from debt funding compared to equity funding.  100% pass through of funding costs to lending rates is overstated. That is, due to the market dominance of the large deposit takers in New Zealand, Group 1 deposit takers may have the ability to hold lending rates relatively constant and channel more of the benefit of lower capital requirements to their owners. Response Please see our response in the next subsection. Assumptions used - Crisis costs To calculate the cost of crisis the probability of crisis under each capital option is multiplied by an expected GDP impact. This returns the Net Present Value (NPV) of a crisis event. The expected present value cost if a crisis event did occur is assumed to be 63% of GDP. This central estimate figure is guided by a Basel Committee on Banking Supervision (2010)43 paper that assessed the long-term economic impact of banking crises. As part of the consultation, we added a new bail-in event to our analysis. This was introduced to improve our comparative analysis of introducing LAC. The cost of a minor event was set at 20% of GDP. This level was guided by a Bank of England (2015)44 paper that estimated an orderly crisis resolution reduces the cost of crisis by over 60%. Consultation feedback We received some concerns from respondents that our cost of crisis assumptions were too high, which resulted in an implicit bias towards overregulation. Other respondents suggested we should model alternative stress scenarios. The Treasury suggested that the CBA was subject to inherent uncertainty about key modelling assumptions, including the assumed cost of a crisis. The Treasury

42 Clark, B., Jones, J. & Malmquist, D., 2023. Leverage and the cost of capital for US banks. Journal of Banking & Finance. 43 Bank for International Settlements. (2010). An assessment of the long-term economic impact of stronger capital and liquidity requirements. https://www.bis.org/publ/bcbs173.pdf 44 Brooke, M., Bush, O., Edwards, R., Ellis, J., Francis, B., Harimohan, R., Neiss, K., & Siegert, C. (2015). Measuring the macroeconomic costs and benefits of higher UK bank capital requirements. Financial Stability Paper No. 35. https://www.bankofengland.co.uk/- /media/boe/files/financial-stability-paper/2015/measuring-the-macroeconomic-costs-and-benefits-of.pdf

2025 Capital Review Summary of Submissions and Policy Decisions 70 suggested we undertake additional sensitivity analysis to assess whether results were robust to uncertainty. Response Given the available empirical evidence and the lack of New Zealand specific studies or domestic crisis events, we continue to consider the assumptions used in the consultation as our best central estimates. However, we do agree with respondents that modelling tail end risks is highly sensitive to the underlying input and assumed distribution. The assumptions used in the models generally fall in the middle of estimates in academic studies and were internationally peer-reviewed in 2019. For example, there is evidence that points to both higher and lower estimates of the MM offset used and the costs of crisis (63% and 20% costs are used for a severe long-term crisis, and a shorter-term recoverable crisis, respectively). We have completed a wide range of sensitivity analysis to understand how sensitive our conclusions are, including testing:  lower crisis cost assumptions; and  variations to the lending rate assumptions, including a 50% pass-through, and different MM offsets. Overall, the relative ranking of each option is broadly stable over a wide range of sensitivities. Further detail is available in the appendix of our cost benefit analysis overview. Net wealth transfer impact In 2019, following industry feedback we added a net wealth transfer impact to our full CBA. The wealth transfer impact recognises the lower interest payments paid by New Zealand borrowers. This could be considered the direct benefits of the lending rate changes, while the previously discussed ‘lending rate’ benefit is about the indirect benefit on expected growth (via increased investment). The wealth transfer impact is reduced by a tax adjustment. Consultation feedback Feedback from our international experts raised scepticism regarding the wealth transfer impact that is included in the CBA, noting that it was counterintuitive. They questioned why industry would favour an option that reduced the transfer of wealth to their owners. Response We continue to include a net wealth transfer impact in the CBA. However, we do recognise the uncertainty in modelling an expected lending rate impact. We have undertaken additional sensitivity, adjusting lending rates to only increase by half as much as our central estimate, that is, a 50% pass through to borrowers.

2025 Capital Review Summary of Submissions and Policy Decisions 71 Alternative options should be explored Consultation feedback We consulted on two main options as part of the consultation. Some respondents argued the range of options analysed was too limited and an option similar to the headline rates in Australia should be tested. Response Following feedback and further analysis we have refined the models and explored alternative assumptions:  Included additional capital stack options for analysis, including headline capital rates that are similar to APRA.  Provided lending rate impacts for changes in scalar and output floor.  Measured the expected benefit of risk weight changes on each deposit taker.  Updated the models to reflect the proposed post-consultation changes to RWA.  Adjusted the definition of a ‘crisis’ to one that absorbs all Tier 2 capital, in order to not overstate the benefits of LAC.  Completed additional sensitivity analysis on lending rate changes, including pass-through assumptions, as well as lower cost of crisis assumptions.  Attempted to provide further disaggregated outputs, for example data on Group 2 and 3 deposit takers. 2.2.6 DCS Levy The DCS commenced in July 2025, protecting eligible depositors up to $100,000 if a deposit taker fails. Compensation to depositors is made from the DCS fund, which is funded through annual levies on deposit takers who offer DCS-protected deposits. The amount of levies each deposit taker is required to pay is calculated to reflect the expected payout if the deposit taker fails, and the likelihood of that failure. Larger and/or more risky deposit takers pay a higher levy than smaller or less risky deposit takers. As set out in the Deposit Takers Regulations 2025 the method for calculating a deposit taker’s risk is based on a range of prudential indicators. One of these indicators is the deposit taker’s capital ratio. Currently the parameters used in converting a deposit takers’ capital ratio into a risk score is based on the capital ratio requirements contained in the 2019 Policy Decisions. Consultation feedback A deposit taker raised concerns that the current DCS levy methodology penalises standardised banks (compared to IRB banks), even when the deposit taker is meeting their capital requirements. The deposit taker suggested that the capital ratio used when calculating a deposit taker’s risk score should be their standardised capital ratio for all banks (including IRB banks). The deposit taker

2025 Capital Review Summary of Submissions and Policy Decisions 72 noted this was possible given the recent shift to dual reporting for IRB banks (that is, reporting using both standardised and IRB risk weights). Response The relationship between the DCS levy method and the capital ratio requirements was not in scope of the 2025 Capital Review. However, we appreciate that certainty for deposit takers is important. We are already aware of the concern raised by the deposit taker, and will be reviewing the DCS levy methodology as part of changes to prepare for the commencement of the DCS and other DTA standards in 2028. The public will have an opportunity to provide feedback on the method as part of this process. We expect public consultation on the revised method to occur in 2027.

2025 Capital Review Summary of Submissions and Policy Decisions 73 Appendix A: Glossary of technical terms Additional Tier 1 (AT1) capital The second highest quality of capital behind Common Equity Tier 1. Under current Reserve Bank policies, Additional Tier 1 capital is made up of perpetual preference shares that offer fixed dividends, no redemption date and which limit other rights of the holder. Preference shares rank ahead of ordinary shares in a liquidation. Australian Prudential Regulation Authority (APRA) An independent statutory authority responsible for the prudential supervision of financial institutions in Australia. Bail-in A crisis management strategy that seeks to recapitalise a deposit taker that is likely to fail (or has failed) by writing down, or converting into ordinary shares, selected capital instruments and liabilities. There are different ways bail-in can be effected using different legal mechanisms. Banking Industry Country Risk Assessment (BICRA) A methodology used by S&P Global Ratings to evaluate the strength of banking systems around the world. It scores banking systems on economic risk and industry risk. Banking Industry Country Risk Assessment reports are typically updated monthly. Basel (Basel I, II, III) framework A set of standards developed by the Basel Committee on Banking Supervision, which is the primary global standard setter for the prudential regulation of banks. The most recent of these frameworks is Basel III, introduced in December 2010. Capital buffer (or ‘prudential capital buffer’ or ‘CET buffer’) Absorbs losses during stress and protects deposit takers from breaching their minimum capital requirements. The capital buffer must be made up of entirely Common Equity Tier 1 capital. The capital buffer can be made up of a number of components (see definitions below): a Conservation buffer, Domestic Systemically Important Banks Capital buffer and Counter-Cyclical Capital Buffer. Capital ratio A deposit taker’s capital divided by its risk-weighted assets (see definition below). A capital ratio is a key indicator of the financial strength of a deposit taker, measuring the losses it can withstand relative to the risk of the deposit taker’s business. Capital Requirements (or ‘Prudential Capital Requirements’) The minimum investment in a deposit taker that must be funded through the issuance of capital instruments (e.g. Common Equity Tier 1, Additional Tier 1, Tier 2 and Loss-Absorbing Capacity) and amount of capital that the Reserve Bank requires deposit takers to have. It includes minimum capital requirements, prudential capital buffers and requirements regarding the approach to risk weights.

2025 Capital Review Summary of Submissions and Policy Decisions 74 Capital Review decisions in 2019 (or ‘ 9 Capital Review’) These decisions introduced higher capital requirements for registered banks under the Bank (Prudential Supervision) Act 1989 which are split into two broad categories: minimum capital requirements (see definition below) and capital requirements (see definition above). The combined impact of these is in the process of gradually shifting up to 18% of riskweighted assets (see definition below) for the four largest banks. An outcome of the 2019 Capital Review was to no longer recognise convertible debt securities for capital purposes. Capital stack The full set of capital instruments for a deposit taker. It includes Tier 1 and Tier 2 capital instruments (see definitions below). Capital Standard One of the standards that banks and non-bank deposit takers will be licensed against under the Deposit Takers Act 2023. This will replace existing prudential requirements to form new capital rules for deposit takers. Decisions from the 2025 Review of key capital settings will be incorporated into the Capital Standard. Common Equity Tier 1 (CET1) capital The highest quality of capital as it is permanently available to absorb a deposit taker’s financial losses. Common Equity Tier 1 capital includes shareholders’ investment (ordinary shares) and the deposit taker’s retained earnings. Community Housing Providers (CHPs) Generally not-for-profit groups that meet housing needs through a range of affordable rental and home ownership options. Conservation buffer A type of prudential capital buffer that applies to all deposit takers. The conservation buffer promotes capital resilience by requiring deposit takers to maintain capital levels above their minimum capital requirements. Cost benefit analysis (CBA) This involves estimating, where possible, the monetary value of all the costs and benefits of a decision to determine an expected net impact of the decision. A cost benefit analysis is one tool that the Reserve Bank uses to compare different policy options. Counter-Cyclical Capital Buffer (CCyB) A type of prudential capital buffer that the Reserve Bank may increase or decrease over the financial cycle. Increasing the Counter-Cyclical Capital Buffer aims to build deposit takers’ capital resilience and guard against financial stability risks. Lowering the Counter-Cyclical Capital Buffer enables deposit takers to operate at lower capital levels during periods of financial system stress, to promote their ability to continue lending to support the economy. Credit Contracts and Consumer Finance Act 2003 (CCCFA) This sets the legislative framework for credit contracts, consumer leases and buy-back transactions of land. Its primary purpose is to protect the interests of consumers in connection with these activities.

2025 Capital Review Summary of Submissions and Policy Decisions 75 Crisis event For the purposes of the cost benefit analysis (see definition below), this means an event where all deposit takers’ capital is absorbed by losses. Crisis management This refers to the responses of the Reserve Bank, deposit takers and other relevant stakeholders to manage the impact of financial distress when it arises. This is both when there is the potential for recovery, and when a deposit taker is likely to fail (or has failed). Crisis management framework This includes the powers, regulations, policies, tools, strategies and processes in place that inform and enable the Reserve Bank and deposit takers’ actions in response to financial distress and potential failure. It also includes business-as-usual preparations and the governance and testing of these arrangements, to ensure they operate effectively in practice. Debt-to-income (DTI) This measures the amount of debt a borrower has, relative to their gross income. Debt-to-income restrictions are a macroprudential tool used by the Reserve Bank to limit the portion of banks’ new lending towards home loans that exceed debt-to-income thresholds. Debt-to-income restrictions aim to reduce the probability of default. Deposit taker An entity that meets the definition of deposit taker in clause 2 of Schedule 2 of the Deposit Takers Act 2023. Deposit Takers Act 2023 (DTA) Legislation that provides for the prudential regulation of deposit takers. From 2028 it is intended that the Deposit Takers Act 2023 will replace the Banking (Prudential Supervision) Act 1989 and the Non-bank Deposit Takers Act 2013. Domestic Systemically Important Banks (D-SIBs) Banks whose failure would result in significant disruption to the New Zealand financial system and economy due to their size, interconnectedness, lack of substitutability, and complexity. ANZ, ASB, BNZ, and Westpac are currently identified as Domestic Systemically Important Banks. Domestic Systemically Important Bank Capital buffer (D-SIB buffer) A type of prudential capital buffer that applies to deposit takers that are identified as Domestic Systemically Important Banks. A Domestic Systemically Important Bank Capital buffer promotes higher capital strength of deposit takers and lowers their probability of failure. Finance and Expenditure Committee (FEC) A select committee of the New Zealand parliament. The business that the committee looks at includes economic and fiscal policy, taxation, revenue, and banking and finance. In June 2024, a Select Committee inquiry commenced on banking competition, which also focused on rural banking.

2025 Capital Review Summary of Submissions and Policy Decisions 76 Financial Policy Remit (FPR) Specifies or provides for matters that the Minister of Finance considers are desirable for the Reserve Bank to have regard to in relation to our financial stability objective, the objectives or purposes of our prudential regulation, and acting as a prudential regulator and supervisor. The Financial Policy Remit is issued by the Minister of Finance under the Reserve Bank of New Zealand Act 2021. Global financial crisis (GFC) The period of extreme stress in global financial markets and banking systems between 2007 and 2009. Going concern capital Instruments that absorb losses while the deposit taker remains an economically viable entity. These instruments help maintain ongoing operations and market confidence. Gone concern capital Instruments that absorb losses once the deposit taker is no longer economically viable. This includes a regulator led bail-in using crisis management powers, as without that intervention the deposit taker could not have continued operating or sustained market confidence. Gross Domestic Product (GDP) A way of measuring economic activity and income in a country in a given period of time. Gross Domestic Product is all the consumption, investment, government spending, and net exports (exports less imports) in an economy. Changes in Gross Domestic Product are New Zealand’s official measure of economic growth. Group 1, 2 and 3 deposit takers Categories of deposit takers that are set out in the Proportionality Framework. Group 1 includes deposit takers with total assets NZ$100 billion or more. Group 2 deposit takers have total assets of NZ$2 billion or more, but less than NZ$100 billion. Group 3 have total assets of less than NZ$2 billion. We allocate deposit takers into groups to support the consistent application of requirements to similar deposit takers, and to allow for requirements to be set proportionately for each group. Internal ratings-based (IRB) approach Allows accredited deposit takers to use the internal models-based approach to calculate their risk weights for credit risk; otherwise, they must use the standardised approach. Accredited deposit takers are sometimes called ‘Internal ratings-based banks’ or ‘IRB banks’. Risk weights for other types of exposures, including operational risk and market risk, must be calculated using the standardised approach. Lenders Mortgage Insurance (LMI) Protects a lender from incurring losses in the event that a borrower defaults on a home loan. Loan-to-value ratio (LVR) A measure of how much a bank lends against mortgaged property, compared to the value of that property. Loan-to-value ratios are used in credit risk weights for some exposures, including residential mortgage loans, in the standardised approach to credit risk. Separately, loan-to-

2025 Capital Review Summary of Submissions and Policy Decisions 77 value ratio restrictions are a macroprudential tool used by the Reserve Bank to limit how much new high loan-to-value ratio lending banks can make. We vary these restrictions in response to changing financial system risks. Tighter loan-to-value ratio restrictions help to reduce the number of highly leveraged borrowers and lower loss given default, supporting the stability of the housing market and reducing the risk of a sharp correction in house prices. Loss-Absorbing Capacity (LAC) instruments Debt instruments that make up part of a deposit taker’s funding and are pre-positioned to allow for bail-in. A Loss-Absorbing Capacity requirement would be in addition to the minimum capital requirements and capital buffers. LAC can be issued to other members of a deposit taker’s group (internal oss-Absorbing Capacity) or to other parties (external Loss-Absorbing Capacity). Loss given default (LGD) The proportion of exposure at default that is expected to be lost following default, calibrated to economic downturn conditions. Loss rate The cumulative impairment expense over a given time period as a proportion of the opening exposure. Minimum capital (ratio) requirements The minimum capital ratio must be met in order to be licensed and operate as a deposit taker. If a deposit taker has a capital ratio below the minimum requirement, it is likely to be in financial distress from a prudential perspective. Modigliani-Miller (MM) theorem tates that changes in a firm’s funding structure (i.e. the ratio of equity finance to debt finance) would have no impact on its weighted average cost of capital. Increases in profitability through the greater use of leverage when capital is lower will be offset by a higher unit cost for the remaining equity capital, since that equity becomes relatively riskier. Therefore, in the context of banks, if the Modigliani-Miller theorem holds fully, changes in capital requirements should have no effect on the cost of lending. Non-bank deposit taker (NBDT) An entity that meets the definition of non-bank deposit taker in section 5 of the Non-bank Deposit Takers Act 2013. Non-bank Deposit Takers Act 2013 (NBDT Act) Legislation that provides for the Prudential regulation of non-bank deposit takers. It is intended that the Non-bank Deposit Takers Act 2013 will be replaced by Deposit Takers Act 2023 from 2028. Output floor A limit on the internal ratings-based (see definition above) approach for deposit takers that calculate the credit risk-weighted assets (see definition above) using the internal ratings-based approach. When determining its capital ratio, the risk-weighted assets for credit risk

2025 Capital Review Summary of Submissions and Policy Decisions 78 cannot go below 85% of the risk-weighted assets that the deposit taker would calculate under the standardised approach (see definition below). Pillars 1, 2, and 3 Components of the Basel framework (see definition above). Pillar 1 requirements are minimum capital requirements to cover credit risk, market risk and operational risk. Pillar 2 includes additional capital requirements for other risks identified as part of the supervisory review process. Pillar 3 covers disclosure requirements and is designed to enforce market discipline on banks. Probability of default (PD) The likelihood that a credit exposure will default, averaged over a range of economic conditions. It is expressed as an annual rate. Proportionality Framework Sets out how the Reserve Bank takes into account the proportionality principle when developing standards for deposit takers licensed under the DTA. See Group 1, 2 and 3 deposit takers defined above. Regulatory impact assessment An analysis of the likely impact of proposed regulatory changes. Recapitalise The process of restoring a deposit taker’s capital to an adequate level by generating new capital from external sources or bail-in. Residential mortgage lending (RML) Defined in section C3.2 of BPR131: Standardised credit RWAs. It is a loan secured by a first ranking mortgage over a residential property used primarily for residential purposes by the mortgagor, a related party of the mortgagor, or a tenant of the mortgagor Risk-weighted assets (RWA) An adjusted picture of a deposit taker’s financial position (for example, its loan portfolios and other investments, and its operational and market trading activities) that takes into account the risk profile of that financial position. Scalar A scaling factor that must be applied to risk-weighted assets for credit risk calculated using the internal ratings-based approach. A New Zealand deposit taker must multiply its risk-weighted assets for credit risk calculated using the internal ratings-based approach by 1.2 when determining its capital ratio. Sensitivity analysis Shows the impact of changing one variable in a model at a time, or changing a small set of closely related variables, while holding everything else constant. Sensitivity analysis shows how different model calibrations affect the conclusions that can be drawn when there is some uncertainty around the model parameters. Single point of entry (SPE) A model of recovery or resolution where the group the deposit taker is part of is kept together. For example, under our preferred single point

2025 Capital Review Summary of Submissions and Policy Decisions 79 of entry model, the Australian parent entity would transfer or ‘downstream’ sufficient capital to the New Zealand subsidiary to restore its viability in a crisis. Small and medium-sized enterprise (SME) In the internal ratings-based approach, banks may separately address exposures to small and medium-sized enterprises in a separate retail small and medium-sized enterprise sub-category of exposures. A loan that is extended to a small business and managed as a retail exposure, and that does not qualify as residential mortgage lending, is eligible for retail treatment where the banking group’s total business-related exposure to the borrowing enterprise (on a consolidated basis, where applicable) is less than NZ$1 million. Standardised approach to credit risk One of the two methodologies available to calculate risk-weighted assets for deposit takers’ credit risks. The standardised approach requires deposit takers to use Reserve Bank specified rules to determine the risk weights to apply to different types of loans and other assets. Statement of Approach to Resolution (SoAR) A document that will be issued under the Deposit Takers Act 2023 setting out the Reserve Bank’s expected resolution strategies and intended approach to co-operating with relevant stakeholders when performing or exercising its functions powers or duties under the crisis management provisions of the Deposit Takers Act 2023. The Deposit Takers Act 2023 requires the Reserve Bank to publish this statement and regularly review it. The first Statement of Approach to Resolution is expected to be published by mid-2029. Terms of reference The 2025 Review of key capital settings sets out the purpose, approach, scope and timing of the 2025 Review. See 2025 Review of key capital settings - Reserve Bank of New Zealand - Te Pūtea Matua. Tier 1 capital Is made up of a combination of Common Equity Tier 1 capital and Additional Tier 1 capital. See definitions of these types of capital above. Tier 2 capital Comprises certain types of reserves and subordinated debt instruments that do not qualify as Common Equity Tier 1 capital or Additional Tier 1 capital but are available to absorb losses ahead of more senior creditors of the banking group in a winding up. Total capital ratio Defined in BPR100: Capital Adequacy. Measured as total capital divided by total risk-weighted assets. Total Loss-Absorbing Capacity (TLAC) An international regulatory standard requiring global systemically important banks to have sufficient equity and bail-in debt that can absorb losses and recapitalise the deposit taker during a crisis, minimising the application of government funds.

2025 Capital Review Summary of Submissions and Policy Decisions 80 Weighted Average Funding Cost (WAFC) A deposit taker’s average funding cost across all funding sources. Whenua Māori Māori freehold land.

2025 Capital Review Summary of Submissions and Policy Decisions 81 Appendix B: Consolidated questions Q1 Do you have any comments on the proposed assessment criteria? Q2 Do you have any comments on the appropriate risk appetite for New Zealand’s capital settings? Chapter 2: Context Q3 Do you have any feedback on our assessment of the impacts of legislative and policy changes since 2019? Q4 Do you have any feedback on our assessment of the new evidence since 2019? Q5 Is there other new evidence not discussed in this section that we should be considering? Q6 Do you have any feedback on this analysis of how New Zealand deposit takers’ current and planned capital levels compare to other jurisdictions? Chapter 3: Capital stack options Q7 Do you have any feedback on the two high-level options for Group 1? Q8 Do you have any alternative proposals? Q9 Do you have any feedback on the proposal for Group 2? Q10 Do you have any alternative proposals? Q11 Do you have any feedback on the proposal for Group 3? Q12 Do you have any alternative proposals? Q13 Do you agree with the proposal of a 1% Counter-Cyclical Capital Buffer for Group 1 and 2 deposit takers under the options proposed? Q14 Do you agree with the proposal that the Counter-Cyclical Capital Buffer should not apply to Group 3 deposit takers? Q15 Do you have any feedback on our analysis of the proposed options against the criteria? Q16 Do you think it would be preferable from a crisis management perspective to maintain a higher Prudential Capital Buffer or have a lower Prudential Capital Buffer and LossAbsorbing Capacity for Group 1? Q17 If you consider that one option is preferable, what are the reasons why?

2025 Capital Review Summary of Submissions and Policy Decisions 82 Q18 Do you have any feedback on the degree of proportionality across the proposed options and capital stacks? Q19 Do you have any feedback on the implications for competition from our proposed options? Q20 Do you have any feedback on our analysis of the options against the assessment criteria? Q21 Do you have any feedback on our approach to the cost benefit analysis? Q22 Do you have any feedback about the results of the cost benefit analysis? Q23 Do you have any additional evidence that should be considered in the cost benefit analysis? Q24 Do you have any comments about the way that Loss-Absorbing Capacity has been incorporated into the approach? Chapter 4: Additional Tier 1 Q25 Do you agree with the proposal to remove Additional Tier 1 capital as a form of regulatory capital? Q26 Are there any other factors that you think we should take into account in making this decision? Q27 Do you have any views on the most appropriate transitional arrangements, including how Additional Tier 1 capital instruments should be recognised after any possible removal? Q28 Are there any additional factors that should be taken into account for Group 3 deposit takers? Chapter 5: Standardised risk weights Q29 Do you agree that the Reserve Bank should introduce more granular standardised risk weights for mortgage, corporate and agricultural lending? Q30 Do you have any comments on the proposed changes to standardised risk weights for mortgage, corporate and agricultural lending? Q31 For deposit takers: Can you quantify the overall and sectoral impact that the proposed changes to standardised risk weights for residential mortgage, corporate, and agricultural lending would have on your institution? Q32 Would you expect more granular residential mortgage lending risk weights to lead to more differentiation in loan pricing to borrowers? Q33 For deposit takers: Can you provide a lending breakdown for your institution by the following corporate sectors: rating, small and medium-sized enterprise retail, small and medium-sized enterprise corporate, and other unrated corporate?

2025 Capital Review Summary of Submissions and Policy Decisions 83 Q34 Do you agree with creating a new standardised risk weight category for all unrated corporate commercial property lending? Q35 For deposit takers: Can you quantify the impact that a 100% risk weight under the standardised approach on all unrated commercial property lending would have on your institution? Q36 Do you have any comments on increasing risk weights for personal lending? Q37 For deposit takers: Can you quantify the impact that a 100% risk weight on secured personal lending and a 150% risk weight on unsecured personal lending would have on your institution? Q38 For deposit takers: Can you provide a lending breakdown for your institution for the following sectors: commercial property (investment, development, and a loan-to-value ratio breakdown within these categories), and personal lending (secured, unsecured, credit card and other)? Q39 Do you think the proposed standardised risk weights more closely align with the actual risk of the underlying lending? If not, where do you think the biggest discrepancies are? Q40 For deposit takers: Is there a desired lead-in time to adopt the proposed standardised risk weight categories and updated minimum capital ratio? What are the expected costs (and their magnitude) to systems and processes of the proposed standardised risk weight categories? Q41 Is there anything else you think we should consider when contemplating changes to standardised risk weights or analysing their impacts? Q42 Do you think the proposed approach to standardised risk weights aligns with the main purpose of the Deposit Takers Act 2023 (section 3(1)) and the additional purposes (section 3(2))?45 Q43 Do you agree with the proposed approach for risk weights on lending for Community Housing Providers and housing co-operatives? Will this approach accurately reflect the risk of that lending? Q44 Do you think the proposed approach for risk weights on lending for Community Housing Providers and housing co-operatives aligns with the main purpose of the Deposit Takers Act 2023 (section 3(1)) and the additional purposes (section 3(2))?46 Q45 How has the Māori and Court whenua Māori practice note altered borrowing and lending decisions? Q46 For deposit takers: How do you treat lending where whenua Māori is the security? Does this affect your assessment of risk?

45 Deposit Takers Act 2023, s 3. https://legislation.govt.nz/act/public/2023/0035/latest/LMS469449.html. 46 Deposit Takers Act 2023, s 3. https://legislation.govt.nz/act/public/2023/0035/latest/LMS469449.html

2025 Capital Review Summary of Submissions and Policy Decisions 84 Q47 Does lending secured by whenua Māori have different risk characteristics than other lending, and if so, how should this be incorporated into prudential requirements? Is this relevant for residential mortgage lending, and/or other forms of lending? Q48 Will lending secured by whenua Māori benefit from the other changes proposed in this Review? Q49 Are there other aspects of the prudential framework that could be addressed to more accurately align risk weights with actual risk for lending secured by whenua Māori? Q50 What are the barriers to borrowing/lending when whenua Māori is used as security? Q51 For deposit takers: Do you participate in the whenua Māori enders Mortgage Insurance underwriting programme run by Kāinga Ora? Q52 Do you support excluding lending for property development from the proposed approach to risk weights for lending to Community Housing Providers and housing cooperatives? Q53 Are the risks during the property development and construction phase different from providing accommodation in finished dwellings? Q54 Do you support excluding lending to third-party providers (who intend to lease to Community Housing Providers or housing co-operatives) from the proposed approach to risk weights for lending to Community Housing Providers and housing co-operatives? Q55 Are the risks of lending by third-party borrowers different from lending directly to Community Housing Providers?