Part 2 (August 2019) capital review

MEMORANDUM FOR FSC FROM Financial Policy (Ian Woolford) MEETING DATE 14 August 2019 SUBJECT Capital Review: Decisionmaking approach FOR YOUR Information This paper recommends that the Committee: i. Note the indicative approach and timetable for the key components supporting decisionmaking for the Capital Review over coming months. Purpose of this paper

1. To outline how Financial Policy will facilitate and support making final decisions in the Capital Review and to seek feedback on the approach, timing and so on. Summary
2. This paper provides an overview of the inputs, activities, and outputs supporting the Capital Review decision-making process. The expected outcomes of the culmination of these processes are also laid out for discussion.
3. In doing this, I also lay out the gameplan for managing: a) the multiplicity of technical decisions that need to be taken to reach a set of publicly announced decisions late November/early December; b) the risks of under- or over-calibration or losing sight of the desired outcomes or underlying principles the Capital Review is based on; and, c) the impact on individual banks in a coherent way. Background
4. When the Capital Review was launched we included the six principles that would guide us in our decision making. These haven’t changed in substance over the course of the Capital Review, and remain relevant to our considerations today and when final decisions are taken. This paper sets out the ‘gameplan’ we have for managing the large amount of decisions that need to be taken, and how we propose managing this process and the risks around it. This gameplan is reflected in the sequence and grouping of the key papers timetabled below. We also put the entire process into an Inputs-Activities￾Outputs-Outcomes framework for easy reference.
5. The ‘gameplan’ is simple: i. Focus on the key building blocks first. ii. Eliminate or park peripheral or second order issues. iii. Take incremental in-principle decisions. iv. Support the incremental decisions by illustrating the impact on individual banks at each decision point. v. Support overall decision with detailed cost-benefits analysis.

2 vi. Confirm that all final decisions are feasible; consistent with the Capital Review Principles; and contribute to an overall package that delivers net benefits to New Zealand. vii. Deal with individual bank issues with bespoke arrangements (if necessary). 6. Focus on the key building blocks first. The biggest framework decisions to take that are contentious are a) the role of AT1s/numerator issues, b) the RWA framework for IRB banks, and c) the D-SIB component of the buffer. Settling the approach and response to submissions early on will simplify consequential decisions. 7. Eliminate or park peripheral decisions. Getting clarity early on about what to focus on will be helped by removing distractions on issues that are either unrelated or of only tangential importance to the Capital Review. Examples of this include Deposit Insurance, TLAC, and supervisory intensity, either because further consultation on these issues will come in due course and decisions on these issues can be taken after final Capital Review decisions, or because they are simply separable (i.e. they may be complements but have limited substitutability). A paper seeking FSC’s endorsement that the issues we identify as peripheral to the Review will come to the next FSC. 8. Take incremental in-principle decisions. This is self-explanatory: having blocked up the issues into coherent bits, taking in-principle decisions allows us to build towards the complete package in a way that also allows us to keep track of the capital implications at each step (see next point). If, for whatever reason, there is a need to go back and change in-principle decisions, this is fine, as we are not proposing that any of the in￾principle decisions are announced, they are only for internal purposes. 9. Individual bank impacts. We have developed a bottom-up spreadsheet for each bank that allows us to see the impact of every decision, as we take them. Every good tool needs a name, so we have called it CRISP (Capital Review Impact and Scenario Projections). CRISP not only allows us to estimate the incremental contribution to the overall level of conservatism at each decision point, it also allows us to identify how decisions impact cohorts (IRB vs standardised for example), and, it gives us a way to identify and then manage bespoke challenges. 10.A partial example of the tool is in Appendix One. Reading from left to right for two stylised examples for ANZ and Kiwibank, the intuition is as follows: i. The baseline for an individual bank’s potential capital profile based on retained earnings and recent history is shown against the proposed five year transition to 16% for ANZ in the left panel but no other proposals are included. This gives the baseline capital profile. ii. The next panel shows the incremental impact on ANZ’s baseline against the full suite of proposals, and the final panel shows a slight relaxation of the proposals by way of some adjustment to the numerator proposals (in this case allowing redeemable preference shares). iii. The bottom three panels similarly illustrate the Kiwibank profile. In these cases, note that CET1 is 15% as proposed. In the second panel Kiwibank is unable to meet the full suite of proposals simply from retained earnings (i.e. it would need to raise capital or scale back credit growth). The third panel illustrates that if we ended up implementing a combination of a different D-SIB in combination with redeemable preference shares, Kiwibank could meet the full suite of requirements through retained earnings. 11.At the risk of stating the obvious, the point of the tool is not that it be used to set policy, but simply to be clear about the cumulative and individual bank impacts of the policy. It

3 does help, of course, in measuring one suite of acceptable policies against another equally acceptable set, to be clear about which has the larger ‘sweet spot’. 12.Cost benefit analysis. This is underway. We have commissioned John Yeabsley to act as a sounding board, and will also discuss drafts of the cost benefit with our external experts. The discussions with these people are for internal QA purposes, that is, we are not anticipating these would be made public at this stage. 13.Confirm all final decisions against three key criteria. First, that they are consistent with the principles laid out for the Capital Review (Appendix Two). Second, they, as a package, deliver net positive benefits to New Zealand. And third, they are realistic and feasible for every bank to meet over reasonable timeframes. 14.Bespoke arrangements. Having come to a landing on the Capital Review, if there are residual individual bank issues to deal with, we can always consider bespoke arrangements. For example, we might face a situation where all banks except one could meet the requirements in five years from retained earnings, but one needed say eight years. It is open to us to tailor transitional arrangements to individual circumstances if need be. Next Steps 15.Based on the ‘key building blocks’ approach outlined above, the timetable (firm in the short-term, indicative further out) is: Paper Info/Decision Timing Decision-making Paper Info 20 Aug External Expert Reports Decision 20 Aug Peripheral Issues Decision FSC+1 CRISP tool Info FSC+1 Numerator Issues (AT1s) Info→Decision FSC+1 Denominator Issues (RWA) Info→Decision FSC+2 D-SIB Info→Decision FSC+3 CCyB Decision FSC+4 Leverage Ratio Decision FSC+4 Capital stack/overall calibration Info→Decision FSC+6 16.The papers listed above are the ‘key building blocks’. It is highly likely that additional papers will be needed, following on from the need for further analysis or alternative options and so on. The timetable is tight, reflecting the numerous technical issues to work through. And, we recognise that for key decisions there will be a desire to chew over some of the issues before a decision paper. This is reflected in the table. 17.However, it is fair to say that there are not too many FSC between now and late November, and so some flexibility on timing, or Governors taking some steps outside of Committee, may be needed. Similarly, we will need to balance having a good record of decisions against the 5 page rule for FSC, so we will have to make use of BSG and background notes for the record.

5 Appendix Two – Capital Review Principles The Capital Review is guided by six high-level principles:

1. Capital must readily absorb losses before losses are imposed on creditors and depositors;
2. Capital requirements should be set in relation to the risk of bank exposures;
3. Where there are multiple methods for determining capital requirements, outcomes should not vary unduly between methods;
4. Capital requirements of New Zealand banks should be conservative relative to those of international peers, reflecting the risks inherent in the New Zealand financial system and the Reserve Bank's regulatory approach;
5. The capital framework should be practical to administer, minimise unnecessary complexity and compliance costs, and take into consideration relationships with foreign-owned banks’ home country regulators; and
6. The capital framework should be transparent to enable effective market discipline. Appendix Three – Inputs, Activities, Outputs, Outcomes INPUTS ACTIVITIES OUTPUTS OUTCOMES VISION  Submissions  Focus Groups  Meetings with stakeholders, interest groups, and sectors  External Expert Reports  Economic modelling  Risk appetite framework  Capital Review Principles  Careful consideration of submissions: options, impacts, alternatives  Cost benefit analysis  Mapping incremental policy decisions against individual bank capital projections (CRISP)  Structuring decisionmaking in tractable and logical sequence  Eliminate irrelevant considerations  Provide information papers ahead of decision papers (where relevant)  FSC decision papers  Capital Impact Tool results  Cost benefit analysis  External Expert Reports  Final decisions paper  Detailed Response to Submissions  Sound and efficient system  Consistent with principles  Net benefits established and understood  Strengthened public confidence in financial markets  Intergenerational wellbeing  Capital focused on the long-term Transactional Strategic Endowments: Bank profitability; Time horizon; Sound economy; strong trans-Tasman regulatory relationships; non-G20 (i.e. greater ability to do what is best for NZ).

MEMORANDUM FOR FSC FROM Financial Policy (Richard Downing) MEETING DATE 20 August 2019 SUBJECT Capital Review: Next steps with External Experts Reports FOR YOUR Decision This paper recommends that the Committee: i. Note that Final Reports from the External Experts for the Capital Review are due on 31 August 2019. ii. Note the conclusions from the Final Reports will be summarised for the FSC meeting scheduled for 11 September. iii. Note that briefings will be prepared for the RBNZ Board (19 September) and the Minister of Finance (9 September). iv. Agree that the Final Reports be publicly released on 23 September, subject to approval at the 11 September FSC meeting. Purpose of this paper

1. This paper provides an overview of the draft External Experts Reports for the Capital Review. Final reports are due on 31 August.
2. The paper seeks agreement from FSC for the final reports from the External Experts to be released publicly on 23 September, after the details have been reported to FSC, the Reserve Bank Board and the Minister of Finance. The public release date can be confirmed at the 11 September FSC meeting, where FSC will provided a summary of the Final Reports. Summary
3. Three External Experts are carrying out their assessments of the analysis and advice underpinning the Capital Review. Final Reports are due on 31 August.
4. The External Experts provided their draft reports at the end of July. The drafts show comfort with the quality of the advice underpinning the Capital Review proposals and with the general direction of the Review. Nevertheless each Expert has signalled a few areas where they think more analysis is needed (see Annex for a summary). This work is underway.
5. The rationale for publishing the External Expert reports is to maintain a transparent consultation process. There will also be public interest in these reports, and it is likely that there will be requests for the reports to be published. There are some risks with publishing these reports; namely that some commentators may claim the reports are biased. Background
6. As part of final stages of the Capital Review, three External Experts were commissioned to independently review the Reserve Bank’s analysis and advice underpinning the Capital Review proposals.

2 7. The External Experts have been asked to take into account the objectives of the Capital Review as well as the domestic context, the available literature, the international debate, and policy developments globally relating to the role of bank capital in supporting the soundness and efficiency of the financial system. 8. The External Experts’ Reports will be part of the suite of information considered in the final decision-making process of the Capital Review, currently scheduled for 2 December 2019. The External Experts’ Reports will sit alongside other material, including submissions from industry and the public, additional analysis and advice completed during the course of the Capital Review, and other internal review processes. 9. The three External Experts are:  Dr James Cummings: Senior Lecturer in finance in the Department of Applied Finance at Macquarie University.  Professor Ross Levine: Willis H. Booth Chair in Banking and Finance at the Haas School of Business, University of California, Berkeley.  Professor David Miles: Professor of Financial Economics at Imperial College, London. 10.The External Experts were selected from a shortlist compiled by the Reserve Bank following conversations with contacts at the IMF and at APRA. The Experts were selected due to their international standing, as well as a desire to have geographic spread, and a range of different academic, regulatory and banking experience covered among the Experts. 11.Dr Cummings visited the Reserve Bank from 25 to 28 June. Prof Miles visited from 22 to 26 July. Each expert had a series of discussions with Financial Policy on a range of aspects of the Capital Review. They also each requested separate meetings with:  The New Zealand Bankers’ Association. Dr Graham Scott, and others from Sapere Research Group, who carried out research and analysis for NZBA, were also part of the meetings.  One of the big banks: WestpacNZ (WNZL) was selected by the NZBA to meet with the External Experts to provide an industry perspective on the Capital Review proposals. 12.Prof Levine has been engaging with the Reserve Bank via regular video conferences. Senior executives from NZBA and WNZL travelled to California on 5 August 2019 to meet with Prof Levine. NZBA and WNZL were offered video conferences with Prof Levine but opted to travel to the US to meet face-to-face. 13.Annex One summarises the contents of the draft reports from the External Experts. 14.We have provided feedback to the External Experts on their drafts. The focus of the feedback has been to ask for additional information on a number of the topics that the External Experts have raised. We are assessing the issues raised by the External Experts, to ensure their perspectives are reflected in the final stages of the Capital Review. Rationale for publishing reports 15.There will likely be interest in the reports from the public. Publishing the documents promptly will keep the public and key stakeholders informed of the process. Risks 16.If the final reports are seen as broadly supportive of the Capital Review, some commentators may claim the reports are biased. Releasing our correspondence with the

3 Experts on their draft reports may help provide assurance that the External Experts have remained at arms-length from the Reserve Bank – the correspondence asks for additional information in a few areas raised by each Expert, but does not provide drafting comments, or try to refute any aspects of the Experts’ drafts. Next Steps 17.Key dates for the External Experts Reports are shown below:  31 August: Final Reports due.  9 September: Summary of Final Reports and publication plans to the Minister of Finance.  11 September: Summary of Final Reports to FSC.  19 September: Summary of Final Reports to the RBNZ Board.  23 September: Tentative public release date for Final Reports. 18.There will be a communications plan to support the public release of the Final Reports of the External Experts, which are likely to attract some public and media interest. The key message will be to position the reports as just one piece of the jigsaw ie. “The Reports are just one component of the Review, alongside public submissions, focus group discussions and on-going stakeholder engagement. All these inputs will help us to make robust and well-calibrated policies and decisions that best represent society’s interests.” Specific elements of the communications plan will include the following:  Media Release to accompany the publication of the Final Reports.  Q&A to be published on the RBNZ website with the release covering the following: o How the Experts were chosen, as per paragraph 10 above. o The process the Experts followed: meetings with Reserve Bank and others.  Proactive release of supporting information: o Letters from Reserve Bank providing comments draft External Experts’ Reports. 19.The Financial Policy and Communications teams will continue to work together to finalise the communications plan for the External Experts Reports. This will include considering whether any more active media engagements may be required, for example briefings for the media. Recommendation 20.It is recommended that FSC agree that the Final Reports from the External Experts be publicly released on 23 September, subject to approval at the 11 September FSC meeting.

4 Annex 1: Summary of draft reports from the External Experts Key points – Dr James Cummings Concludes that the proposals are “based on sensible analysis and advice in the New Zealand context”. Recommendations cover three main areas:  Contractual loss-absorption features: supportive of the decision to remove contingent convertible instruments from the definition of approved regulatory capital. However, he suggests the RBNZ continue to monitor the international performance of these instruments, in case they become feasible for NZ in the future.  IRB models: Dr Cummings suggests the RBNZ monitor the extent to which RWAs calculated with internal models provide a more accurate prediction of unexplained losses than standardised models. If not more accurate, he suggests reconsidering the decision to retain IRB models.  Impact of proposals on bank funding costs: Dr Cummings concludes that the RBNZ estimates of the impacts of higher capital requirements on bank funding costs are overstated. This implies that the impact on GDP of higher capital estimated by the RBNZ would be overstated. Key Points – Prof Ross Levine The draft concludes that “with a few exceptions, my conclusion is that the RBNZ conducted a sound analysis of bank regulatory capital requirements in New Zealand, employed appropriate data, methods, and evidence, considered a broad and proper array of factors, addressed thoroughly submissions by banks and others regarding the RBNZ’s analysis and proposals, and focused intently on the New Zealand-specific features of capital regulatory reform.” Prof Levine identifies a few “exceptions” to this conclusion  The proposals do not sufficiently consider incentives on bankers and therefore the stability and efficiency of banks. While he sees this as a gap in the analysis, he does note that he doesn’t expect this to necessarily cause the RBNZ’s recommendations regarding capital ratios to be higher or lower.  The dynamic response of the financial system to increased capital requirements. He recommends the RBNZ spend more time assessing whether or not there are potential barriers to entry to new firms in the financial industry.  Countercyclical buffer (CCyB). He is sceptical about the effectiveness of the CCyB, and is does not believe that banks should be able to distribute dividends (or bonuses) when the CCyB has been removed during times of crisis. Key Points – Prof David Miles The draft concludes that:  “The analysis conducted by the RBNZ seems to have been done with care and in an open minded way.”  “The claim (from submitters to the consultation) that RBNZ analysis consistently errs on the side of caution in modelling risks, and also errs on the side of favouring more equity in its assessment of the impact on the wider economy of banks using more capital, does not seem to me to stand up.” Prof Miles has focused largely on detailed aspects of the modelling approach followed by the RBNZ, identifying a number of areas to consider further, within a broad conclusion that the overall approach is sound. He indicates that the RBNZ may have underestimated the overall costs of financial crises, but also concludes, for example, that the RBNZ may have been too conservative in modelling the probability of default on loans. He also rejects the idea that the RBNZ has not carried out an assessment of the costs and benefits – he points to the analysis of the impacts on expected output, noting that this sits at the centre of much of the analysis. Nor does he agree that estimates have been biased by overly conservative calibration.

NOTE FOR Financial Stability Committee FROM Matthew Brunton (Financial Policy) DATE 27 August 2019 SUBJECT CRISP Spreadsheet Model FOR YOUR Information It is recommended that the Committee: Note that the ‘CRISP Tool’ is an internally consistent spreadsheet model that projects individual bank capital ratios. It can analyse the impacts of different capital requirements, and underlying assumptions on interest rate changes and credit growth, for a given bank. Note that FP intends to use the CRISP Tool as a tool to aide decision-making by assessing the cumulative impacts of potential decisions at FSC. Note that the CRISP Tool is not intended to be the sole piece of analysis for decisions around any particular proposal in the Capital Review. These would also rely on qualitative analysis and underlying principles. Note that a demonstration of the CRISP Tool will be given at FSC. Purpose

1. This note describes the ‘CRISP Tool’ (Capital Review Impact and Scenario Projections), an excel spreadsheet model that projects the impact of different capital requirements on individual banks. The purpose of the model is to provide projections of individual bank capital ratios that are internally consistent with assumptions and different threads of work in the Capital Review. Background
2. So far working groups and FSC have been presented with different sets of analysis on: i. the transitional impacts for banks capital ratios, ii. interest rate impacts and MM offsets, and iii. the impact of different definitions of AT1 capital, as well as other pieces.
3. The CRISP tool brings all these different threads of analysis into one internally consistent model to project bank capital ratios under different scenarios. The model
4. The CRISP tool provides a coherent picture in projecting bank capital ratios based on a range of assumptions and different capital requirements. It has a breakdown of key capital and asset line items, providing a relatively flexible framework for analyzing specific issues or questions. For example, we could ask:  What would the removal of Tier 2 requirements look like?  What if we increased the DSIB buffer?  What if banks tilted their portfolios away from Agri?  And what would it look like if any combination of the above were to occur?

2 5. It also allows us to incorporate bespoke scenarios on individual banks. For example, we could assess what the impact of ANZ reducing credit growth significantly in the first few years would be, or if BNZ were to cut back lending in Agri more than other banks. 6. The tool does this by imitating individual bank balance sheets and forecasts their projected capital ratios based on assumptions. It follows a simple logic: i. Capital requirements and other assumptions (credit growth, MM offset etc.) determine what a bank’s balance sheet will look like in 2025. ii. Banks then determine how much earnings they must retain today in order to reach the requirements by 2025. iii. The new equity/debt ratio from the extra retained earnings then determines the interest rate impact for the next period (based on MM offset). iv. The logic is repeated the next period to incorporate the changes in interest rates. Examples 7. The example overleaf shows the impact of the current proposals for DSIB banks on the left. Please note that this is only illustrative and is not intended to be interpreted as advice 8. Based on the historic credit growth over the past five years (which is around 6%) and assuming banks will have a 1% voluntary buffer, this shows that ANZ and BNZ will struggle to with their current credit growth. 9. However, if redeemable AT1 were allowed, their capital ratios reach over 16% by the end of the transition. However, they do not pay dividends until they reach their target ratio (16% + voluntary buffer). This means there isn’t much buffer for them to grow their capital ratios faster than what is currently projected. This is in contrast to WNZL and ASB, who do not need to retain all earnings at their current growth paths to reach their target ratio. 10. Finally, if the AT1 requirement were to go up by 1.5% (to 3%), and CET1 to fall by the equivalent amount (so CET1 requirement is 13%), ANZ and BNZ would be able reach their target capital ratio and make dividend pay-outs during the transition period. 11. This example shows the cumulative impact of the two AT1 requirements (redeemability and higher AT1 ratio) changing. Next steps 12. Financial Policy will provide analysis using the CRISP Model with decision papers for FSC where sensible. We envisage that this will show the cumulative impacts of FSC decisions made prior to papers as well (however we can also use the tool in other ways if FSC would like). For example, the impact of the DSIB decision would also show the impacts of decisions made up to that point, such as the definition of AT1.

Ref #8670697 v1.4 MEMORANDUM FOR Financial Stability Committee FROM Financial Policy (Primary author: Paula Hontalba) MEETING DATE 27 August 2019 SUBJECT Implications of changes to APRA rules for the Capital Review FOR YOUR Information Background

1. The Reserve Bank and Australian Prudential Regulation Authority (APRA) are both currently reviewing bank capital. In our calibration of capital requirements, we have emphasised the need to withstand a 1-in-200 year shock and modelling capital requirements to reflect risks in New Zealand financial system. Across the Tasman, APRA has focused their attention on achieving ‘unquestionably strong’ capital ratios, which they have defined to mean that the capital position of Australian ADIs (Authorised Deposit-taking Institutions) should be within the top quartile of internationally-active banks.
2. We have proposed a CET1 ratio of 14.5% for large banks and 13.5% for non￾systemically important banks, while APRA has proposed that CET1 ratio of at least 10.5% is needed to meet the ‘unquestionably strong’ benchmark. We have also proposed a total capital ratio of 18% for large banks, while APRA has proposed a target total capital ratio of 17% for the four major banks. An important point here is that We recommend that FSC: • Note that one of the principles of the Capital Review is that the capital framework should take into consideration relationships with foreign-owned banks’ home country regulators; • Note that the purpose of this memo is to inform FSC about how changes to APRA’s related entities framework and potential changes to APRA’s capital framework could affect the ability of New Zealand subsidiaries to meet higher capital requirements, and that we will factor this into our assessment of capital proposals; • Note that the potential changes to APRA’s capital framework is somewhat less aggressive than we expected, and essentially APRA is seeking to incentivise any additional funding in the NZ subsidiary to be funded by equity; and • Note that if the New Zealand subsidiaries meet the proposed RBNZ requirement through retaining earnings (instead of seeking capital injections) they would not be impacted by APRA’s tighter constraints on intra-group exposures.

4 Ref #8670697 v1.4 the 400% risk weight, and this risk weight will become less binding if the parent’s equity investment in the subsidiary is less than 10% of the Level 1 CET1. For example, if ANZ Australia has $100 Level 1 CET1, and less than $10 of shares invested in ANZ NZ, then the Level 1 capital requirement for ANZ NZ decreases. If however ANZ Australia has $20 shares invested in ANZ NZ, then the Level 1 capital requirement becomes more binding. 8.

Figure 3: Estimates of Level 1 capital ratios 9. The chart above also shows our estimates of how Level 1 capital ratios would be affected as at 31 March 2019, if the parent invests more equity into the NZ subsidiary to meet our proposed capital requirements and if the parent does not issue more equity to meet unquestionably strong capital requirements. Note that this chart likely overstates the impact on the standalone parent’s capital position, as the NZ subsidiaries will likely meet higher capital requirements via some combination of retaining more earnings, asset allocation, as well as possibly raising more equity from the parent. Related party exposures 10. APRA has recently finalised its policy on the related entities framework (APS 222), which will take effect on 1 January 2021. The key change was that exposure to an individual related ADI would be limited to 25% of Level 1 Tier 1 capital, which is much more restrictive than the previous limit of 50% of Level 1 Total capital. 11. s(9)(2)(ba)(i) s(9)(2)(ba)(i)

5 Ref #8670697 v1.4 In effect, the Australian parent (Level 1) would need to raise additional equity for every dollar of additional equity invested in New Zealand subsidiaries above the 25% limit, if the Australian parent wanted to maintain the same Level 1 capital ratio. 12. It is important to note that the APS 222 limits only applies to exposures that are risk￾weighted, i.e. this only includes ordinary shares issued to New Zealand subsidiaries and this excludes retained earnings held on the subsidiary. Hence, if a NZ bank builds up retained earnings to meet proposed higher capital requirements, then this would not have an impact on the APS 222 limit. Implications of changes to APRA’s rules on Capital Review 13. All else equal, the changes to APS 222 and proposed changes to capital (APS 111) constrain the ability of NZ subsidiaries to raise more equity from the parent. This has an impact on the ability of the NZ banks to meet our capital requirements. Note that on the face of it, this could constrain the ability of Australian parents to support the NZ banks during a crisis. 14. As mentioned earlier, NZ banks can meet higher capital requirements by retaining more earnings, and this would not have an impact on the Level 1 capital ratio or on the APS 222 exposure. This is something that we will consider in deciding transitional arrangements. 15. We will keep engaging with APRA to ensure that we understand how changes to both jurisdictions’ capital requirements can affect capital allocation across the banking group. s(9)(2)(ba)(i) s(9)(2)(ba)(i)

s 9(2)(ba)(i)

1 To: Financial Stability Committee From: Financial Policy (Walter Shea) Subject: Capital Review: Non-core Issues Date: 27 August 2019 For: Discussion and Approval Background

1. During the course of the Capital Review, the Reserve Bank has been asked to consider a large number of issues that could impact final decisions on the appropriate capital framework for New Zealand’s banks.
2. Financial Policy has determined that many of these issues are peripheral to the core issues of the Capital Review and therefore should not impact the decision-making process to any great extent.
3. This paper lists the ‘non-core’ issues identified by Financial Policy and the (high level) reasons for this classification. The Financial Stability Committee is asked to:  note that Financial Policy, after consulting with the Banking Steering Group, proposes to classify certain issues raised during the Capital Review consultation process as ‘non-core’ to the Capital Review, with the consequence that these issues will not factor into the Capital Review decision-making process to any significant extent;  note that these issues are proposed to be classified as such for different reasons (e.g., lack of substance to argument, issue to be addressed at a later date, issue to be monitored);  note that these issues will be addressed in the Reserve Bank’s Response to Submissions that will be released alongside final decisions on the Capital Review in December 2019; and,  discuss and approve the proposed classification of these non-core issues.

2 4. While we believe that some of these issues lack substance and do not warrant further analysis, we also believe some of them are valid and should be monitored and potentially addressed by the Reserve Bank in due course. 5. Some of these non-core issues intersect with issues we have determined to be ‘core’ to the Capital Review and will therefore be addressed indirectly. 6. Subject to discussion and approval by FSC, these issues will not play a significant role in the Capital Review decision-making process. Summary 7. The following table (pages 3 to 5) summarises the list of issues we propose be classified as ‘non-core’ to the Capital Review decision-making process, our rationale for such classification, and any future action to be taken. 8. A more fulsome explanation of each issue is provided in the Appendix.

4 the setting of capital levels for the Reserve Bank’s soundness goal. c) Higher capital requirements will increase barriers to entry for new banks. The Reserve Bank is not proposing to change the entry requirements for new banks. While new banks will be required to maintain higher levels of capital than previously, they will be competing on a more level playing field with incumbent banks (due to the proposed ‘scalar’ and capital floor that would apply to the four internal-ratings based (IRB) banks). None. d) Higher capital requirements will result in ‘disintermediation’ (credit provision moving from the regulated bank sector to the non/lightly-regulated non-bank sector). For this argument to hold, it must be assumed banks willingly cede market share to the non-bank sector in the face of higher capital requirements. While we acknowledge this as a possible outcome, we believe this argument lacks credibility as it would require banks to sever existing profitable banking relationships and cede market share to competitors. None. However, our analysis on how banks may respond to higher capital requirements will address this issue. e) ‘Dual-registration’ banks (banks that have both a subsidiary and a branch in NZ) will originate or transfer business to branches to circumvent higher capital requirements at the subsidiary level. The Reserve Bank controls the size and nature of branch operations and can make adjustments to these controls as needed. The Reserve Bank acknowledges that the dual-registration framework should be reviewed and updated, and will do so in due course, but separate from the Capital Review.

5 f) The Reserve Bank should adopt the Basel III standardised approach (for credit risk, but particularly for mortgage loans), which is more risk sensitive than the Reserve Bank’s current standardised approach, but also has lower average capital outcomes. Our view is that the existing calibration for residential mortgages in BS2A (standardised approach) is appropriate for New Zealand. The Capital Review is focused on the aggregate calibration for standardised banks (rather than risk￾sensitivity). The Reserve Bank will review the appropriateness of the Basel III standardised approach (and APRA’s implementation of it) for New Zealand’s circumstances in due course.

6 APPENDIX – NON-CORE ISSUES a) Relationship between Capital Requirements and Other Regulatory Tools i. Capital Requirements and Deposit Insurance It has been suggested that the existence of a deposit insurance framework should lower a bank’s capital requirements as the impact of a bank failure is less severe with the existence of deposit insurance. While a deposit insurance framework – which the Government has announced it intends to introduce – would alleviate the impacts of a bank’s failure on insured depositors, it would not soften the impacts of the bank’s failure on New Zealand’s economy and many of the adverse social impacts on New Zealanders, nor would it provide compensation for uninsured depositors and other creditors not covered under the deposit insurance scheme. The failure of a bank in New Zealand, and in particular the failure of a large bank, would likely lead to an overall contraction in credit provision to the economy as banks tighten lending standards, resulting in reduced economic growth. It could also impact confidence in New Zealand’s financial system and create funding challenges for New Zealand’s banks. Minimum bank capital requirements ensure that in the event banks incur large unexpected financial losses, they can maintain the confidence of depositors and other creditors, and continue to provide credit and critical services to the economy. Higher capital requirements, and correspondingly higher levels of capital, increase a bank’s capacity to incur such losses. Once implemented, New Zealand’s deposit insurance framework is expected to provide a failed bank’s depositors with relatively quick compensation (possibly within seven days) for insured deposit amounts. The existence of a deposit insurance framework would not increase a bank’s capacity to absorb large unexpected financial losses, the objective of minimum capital requirements. We do not believe that the calibration of minimum capital requirements should take into account the existence (or non-existence) of a deposit insurance framework. 1

Indeed, the Reserve Bank’s Capital Review proposals did not suggest that higher capital requirements were needed given the absence of a deposit insurance framework; in the same way, we do not believe that the existence of a deposit insurance framework suggests lower capital requirements should be acceptable. The objectives of bank capital requirements and deposit insurance are distinct and should not be conflated. While the existence of a deposit insurance framework may alleviate the impacts of a bank’s failure on that bank’s insured depositors, it does not

1 It should be noted that most jurisdictions around the world increased capital requirements following the global financial crisis even with the presence of a deposit insurance framework. We are not aware of suggestions, even by industry, that the extent of those increases should have been limited due to the existence of a deposit insurance framework.

7 reduce the probability of that bank failing in the first place nor does it alleviate the broader economic and social impacts of its failure. 2

In other words, even though an ambulance (deposit insurance) is waiting at the bottom of the cliff, it doesn’t mean there should be a weaker fence at the top (minimum capital requirements). Were a bank to fall off the cliff, the damage to New Zealand’s landscape would be significant and lasting and could only be partly offset by the existence of a deposit insurance framework. The existence of a deposit insurance framework may, however, suggest that a bank’s funding outflows may be lower during a liquidity (run) event as depositors have less incentive to withdraw money from the bank given that they will be compensated in the event of the bank’s failure. This issue will be examined in due course as part of the Reserve Bank’s review of its Liquidity Policy.

ii. Capital Requirements and Resolution Regime In 2015, the Financial Stability Board released its ‘Total Loss-Absorbing Capacity for Global Systemically Important Banks’ (TLAC framework), which set out an expectation that a minimum portion of a bank’s total loss-absorbing capacity consist of debt instruments subject to ‘bail in’ (i.e., write down or conversion to equity) as a means to recapitalise a bank in distress. The Reserve Bank has monitored global TLAC developments since the idea was conceived after the 2007/08 global financial crisis and determined that it would not adopt a TLAC framework in late 2018 (but would continue to monitor TLAC developments). The Capital Review proposals are focused on ensuring that New Zealand banks have adequate ‘going concern’ (Tier 1) capital and were calibrated on the basis that New Zealand would not be adopting a requirement for banks to issue ‘bail-inable’ debt instruments at this time. There have been suggestions that the Capital Review intersects with crisis management policy, because both aim to reduce the probability and severity of a crisis.3 The contention is that lower capital requirements could be justified if a TLAC regime was to be introduced in New Zealand, to which there is a divergence of views globally (particularly in the UK).4 Proponents of a TLAC regime also argue that

2 Deposit insurance may, however, reduce the social impact of a bank’s failure on the failed bank’s insured depositors by providing them with prompt reimbursement of their insured deposits. 3 Crisis management is currently one of the topics being consulted on as part of Phase 2 of the Review of the RBNZ Act. See chapter 5 of Consultation Document 2B of Phase 2 of the RBNZ Act Review entitled ‘What features should New Zealand’s bank crisis management regime have?’ 4 Sir John Vickers states that “[a] gulf this wide […] on a truly fundamental economy policy question is extraordinary”. See Vickers, J, ‘Safer but not safe enough’, keynote address at the 20th International Conference of Banking Supervisors, November 2018. https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=2ahUKEwj40uiSl9njAhU XXn0KHbqqCpQQFjAAegQIAhAC&url=https%3A%2F%2Fwww.bis.org%2Fbcbs%2Fevents%2Ficbs2 0%2Fvickers.pdf&usg=AOvVaw01teXScOIzKf-unHJMmWmQ

8 because it requires banks to issue eligible debt instruments (subject to write down or conversion to equity), it enhances market discipline While a fully developed bail-in regime is conceptually attractive, in terms of reducing the likelihood of both a crisis and a taxpayer bail-out, the key unknown is whether bail-inable debt will operate as intended in a systemic crisis. There is little global experience to date of a resolution authority successfully ‘bailing in’ debt instruments to recapitalise a distressed bank. Further, because New Zealand is the host of domestically significant banks with parents located in Australia, a decision to progress the development of a TLAC framework would require consultation with Trans-Tasman stakeholders. One important issue, for example, would be whether this bail-inable debt issued by the New Zealand subsidiary would be held by the parent bank (internal TLAC), issued to external investors, or both. For these reasons it appears prudent not to offset strong going concern (Tier 1) capital requirements with potential – and as yet undeveloped – proposals for banks to maintain additional (lower quality) gone concern loss-absorbing instruments. The Reserve Bank will continue to monitor global TLAC developments and will consider adopting a TLAC framework after more experience with it is gained globally. iii. Capital Requirements and More Intensive Supervision There have been suggestions that since the Reserve Bank is moving to a more intensive supervisory approach, capital requirements need not be as stringent. A more intensive supervisory approach will provide the Reserve Bank with greater assurance that banks are complying with their legislative and regulatory requirements, however, it will not improve banks’ capacity to absorb financial losses. We do not agree that a more intensive supervisory approach suggests that capital requirements should be less stringent. b) Voluntary (Management) Capital Buffers Banks in New Zealand, like banks around the world, generally maintain levels of capital greater than regulatory minimums and buffer requirements so as to reduce the likelihood of breaching these requirements. It has been suggested that the Reserve Bank’s proposed minimum and buffer requirements should incorporate the level of capital bank’s voluntarily choose to maintain above regulatory requirements. While we acknowledge the common practice of banks to voluntarily maintain capital levels above regulatory minimum and buffer requirements, we have not reduced the proposed required level of capital to make an allowance for the likelihood that banks will maintain a voluntary buffer.

9 As the focus of the Reserve Bank’s capital requirements is to ensure that sufficient capital is available to absorb large unexpected losses (e.g., losses of the size that would be expected to occur only once every two hundred years), and there is no certainty that these voluntary buffers would exist across all phases of the economic cycle, these voluntary buffers do not provide the certainty needed to meet the Reserve Bank’s objectives for minimum and buffer capital requirements. In addition, as the size of these buffers is discretionary and varies across banks, we cannot easily factor them into the setting of minimum capital requirements. c) Barriers to Entry Some submissions expressed concern that higher capital requirements could increase barriers to entry (i.e., make it more difficult for new banks to enter the New Zealand market). The Reserve Bank recognises the importance of facilitating competition in the banking sector, which has also been highlighted by one of the External Experts. Currently, the requirements for registering as a bank in New Zealand are fairly low. For example, an applicant is required to have only a minimum of $30 million in capital. This minimum capital requirement is intended to ensure that an applicant has sufficient substance to carry on business as a registered bank and to demonstrate that the owners have made a reasonable commitment to the business.5

More stringent regulatory requirements, such as capital and liquidity requirements, are applied to banks incorporated in New Zealand. In general, an applicant is required to incorporate in New Zealand if6 :  it is considered to be systemic (i.e., it has net New Zealand liabilities greater than $15 billion);  the home jurisdiction of the applicant has explicit depositor preference;  the applicant plans to accept retail deposits; or,  the home jurisdiction of the applicant has an inadequate supervisory and disclosure approach. While new banks incorporated in New Zealand would be subject to higher capital requirements following the Capital Review, we do not believe that such requirements will increase barriers to entry, particularly as the minimum $30 million capital requirement for entry would remain unchanged. Following entry, the owners of new banks would be required to maintain a more meaningful stake in the bank through higher levels of capital, but they will be competing on a more level playing field with incumbent banks (due to the proposed

5 RBNZ Statement of Principles, BS1, May 2019 (paragraph 43). 6 RBNZ Statement of Principles, BS1, May 2019 (paragraph 25).

10 ‘scalar’ and capital floor that would apply to the four internal-ratings based (IRB) banks). d) Disintermediation Impacts Many submissions claimed that higher capital requirements for banks would result in disintermediation in the financial sector (a shift in credit provision from the regulated banking sector to the non/lightly-regulated sector, such as finance companies, building societies, and credit unions). For this to occur, it must be accepted that higher capital requirements for banks would result in a decrease in credit provided by banks; however, we do not believe that banks will cede market share to competitors to any great extent, even in the face of higher capital requirements. Our analysis on how banks may respond to higher capital requirements will address this issue indirectly. It should also be noted that New Zealand’s financial sector is currently heavily dominated by banks, so a potential shift in credit provision from the banking sector should not necessarily be viewed as a negative development given that it would reduce the concentration of credit provision in the financial sector and may increase competition for the benefit of consumers. In any event, the Reserve Bank will monitor developments in the financial sector (both regulated and non/lightly-regulated sectors) over the Capital Review transition period and will take actions, or suggest actions be taken, to maintain a sound and efficient financial system. e) Use of Branches It has been suggested that, in the face of higher capital requirements, some of the large four banks may attempt to reduce the impact of higher capital requirements by originating more business from their branch operations or by transferring business from their subsidiaries to their branches (ANZ, ASB, and Westpac have branches in New Zealand; BNZ and other banks could establish one). However, as the Reserve Bank controls the nature and size of branch operations, a bank’s ability to circumvent higher capital requirements imposed at the subsidiary level through branch operations would be limited. Additionally, to the extent that banks attempt to inappropriately reduce their capital requirements through the use of branch operations, the Reserve Bank could take the necessary actions to cease this activity. The Reserve Bank recognizes that its ‘dual-registration’ (subsidiary and branch) framework should be updated, which will be done in due course.

11 f) Basel III Standardised Approach (Credit Risk) The Reserve Bank’s current Capital Adequacy Framework (Standardised Approach), also known as ‘BS2A’, is a modified version of the Basel II capital framework (2006). Under the standardised approach, banks are required to use prescribed risk weights for calculating risk-weighted assets (RWAs). The risk weights for credit risk in BS2A are similar to those contained in the Basel II capital framework, with the main exception being more granular risk weightings in BS2A for residential mortgages. Following the Global Financial Crisis of 2007/08, the Basel Committee on Banking Supervision (BCBS) launched a large number of reforms to its capital framework (known as ‘Basel III’), which included proposed revisions to its standardised approach for credit risk. The Basel III standardised approach for credit risk, which includes more risk-sensitive risk weights than those contained in Basel II, was finalised in December 2017 (with implementation to occur by 1 January 2022). Some submissions suggest that the Reserve Bank adopt the Basel III standardised approach for credit risk, particularly for mortgage loans. It is important to separate risk sensitivity in the Basel III standardised approach (the increased granularity of risk weights) from the overall calibration (quantum of capital) of the framework. In the Response to Submissions to the ‘Denominator Paper’, the Reserve Bank indicated it would not consider changes to the standardised risk￾weight framework (BS2A) for the time being, as BS2A already incorporates a higher degree of risk sensitivity than the Basel II standardised approach. While the Reserve Bank intends to review the Basel III standardised approach for credit risk (and APRA’s implementation of it) in due course, at this time our view is that the aggregate calibration for mortgage risk-weights is appropriate for New Zealand.