Reserve Bank of New Zealand Capital Review Submissions

Adrian Tocker I agree with the RBNZ's proposal to increase the capital banks are required to carry to improve the robustness of our banking system and protect against significant financial and monetary shocks. OIA s9(2)(a)

Alan Jones The less credit that Banks' have at their disposal the more productive their loans have to be i.e. the banks take a greater share of risk to retain their level of profit. The taxpayer (the Reserve Bank) has given the banks the facility to make profit from capital not earned or owned by the bank and yet continues to virtually guarantee those same banks performance i.e. if they screw it up the taxpayer steps in to bail out the bank. We have seen what happens when pure profit motive drives a bank's decision making (derivatives anyone?) but with no effective safeguards in place there is no barrier to continued bad decisions. When you look at the source of a bank's capital (money directly invested i.e. cash plus the amount of credit that it is allowed to play with by the Reserve Bank) why is the credit component not charged for? OIA s9(2)(a)

Alex Tai I am in full support of the new capital requirements. OIA s9(2)(a)

Alistair Morrison My concern is that bank margins will rise and how does a client know whether or not that some clients are being unfairly burdened with higher margins due to confidentiality of individuals Also how do we know if one sector of the banks clients are subsidising another sector OIA s9(2)(a)

Andrew Bartlett I support the RBNZ proposal. There is an implicit guarantee (many New Zealand depositors think it is explicit) from the sovereign to the banks that is simply not being paid for by depositors/bondholders or bank shareholders. Putting NZ bank capital requirements clearly in the top percentile, commensurate with our narrow economic base, will ensure that even remote bank risks and their costs are appropriately allocated to shareholders. I believe that the additional costs to borrowers/depositors (if material) will be small relative to the implicit cost to the taxpayer and wider economy from a banking crisis. OIA s9(2)(a)

From: Andrew Body To: Capital Review Subject: Bank capital review Date: Friday, 17 May 2019 3:14:33 PM Attachments: Andrew Body Limited submission on RBNZ"s capital review (FINAL).pdf Hello Please find attached our submission on the RBNZ’s capital review. Please confirm receipt of this submission. Regards Andrew Body CAUTION: This e-mail and any attachment(s) contains information that is both confidential and possibly legally privileged.  No reader may make any use of its content unless that use is approved by Andrew Body Limited separately in writing.  Any opinion, advice or information contained in this e-mail and any attachment(s) is to be treated as interim and provisional only and for the strictly limited purpose of the recipient as communicated to us.  Neither the recipient nor any other person should act upon it without our separate written authorisation of reliance. If you have received this message in error: ·         do not copy, disclose or use the contents in any way, and ·         please notify us immediately and destroy this message.  Thank you. Andrew Body Andrew Body Limited PO Box 91342 A.M.S.C. Auckland 1142, New Zealand                      OIA s9(2)(a)

1 MEMORANDUM TO: Mr Ian Woolford Financial System Policy and Analysis Department Reserve Bank of New Zealand FROM: Andrew Body Principal Andrew Body Limited DATE: 17 May 2019 SUBJECT: Bank Capital Review Dear Sir Objectives and macro versus meso level assessment

1. The Reserve Bank of New Zealand’s Capital Review proposals (Proposals) pose the question as to whether the prudential regulation of New Zealand’s banking system optimises the trade-off between system soundness and efficiency to contribute to a productive and sustainable economy.
2. The Reserve Bank of New Zealand’s statutory objectives of soundness and efficiency are a couple. One does not dominate the other and they are inseparable. Optimisation requires a clear understanding of the trade-offs between the objectives.
3. However recent public discourse on the Proposals has resolved itself into a contest between widely divergent views and powerful institutional interests.
4. The Reserve Bank of New Zealand (RBNZ), by being able to determine prudential regulation, has the power to maintain any contest indefinitely. The independent ratings based (IRB) banks, through their market dominance, have the capacity to withstand the challenges of the RBNZ and stay viable.
5. This contest calls to mind a struggle between powerful contestants over issues that are almost completely unfamiliar to those who bear the consequences of the outcome - borrowers, depositors and the general public (the “little guys”). The little guys are currently only on-lookers despite apparently being the only ones to bear the consequences of the outcome.
6. There is a sense about the Proposals that the RBNZ does not want to engage in an efficiency discussion because it takes the organisation out of a familiar institutional and “macro” analytical environment into an unfamiliar and more complex “meso” analytical environment.

2 Possibly the RBNZ’s nervousness about its meso capabilities is even the reason for the Proposals in the first place? 7. We submit that the outcome of any regulatory developments in the New Zealand banking system should optimise the trade-offs between system soundness and efficiency at both the macro economic level and the meso economic level where the impacts of the Proposals on the little guys can be much much more carefully evaluated. Anything less than this is inconsistent with the Reserve Bank Act, costly and politically unsustainable. 8. The current institutional and macro analytical approach taken by both the RBNZ and the banks does not look like either an economically or politically viable way to optimise the system’s soundness and efficiency. Afterall neither the RBNZ nor the banks vote. 9. By way of example of the meso approach we detail one area of the lending market that has been adversely affected, with no discernible benefits in respect of system soundness, by the RBNZ’s capital adequacy (credit risk) rules and which will have that adverse impact exacerbated by the Proposals. Who pays really does matter 10. The RBNZ hasn’t offered a proper cost-benefit analysis to frame the evaluation of the trade￾off between soundness and efficiency under sections 1A and 68 of the Reserve Bank Act. What they have provided does not pass muster of the Government’s guidance for cost￾benefit analysis1 . 11. Without such a cost-benefit analysis the banking system’s stakeholders are not equipped to judge the impact of the Proposals on the system. We also note that so far the RBNZ’s review has not: a. evaluated the merits of the open bank resolution scheme, a possible deposit guarantee scheme or other possible schemes b. considered how different governance and resourcing of the RBNZ’s prudential management will affect system soundness c. considered how Australian banking prudential regulation impacts on soundness of the New Zealand banking system d. considered how new technologies and increased competition could affect system soundness e. provided any substantive distributional impact analysis. 12. In addition to the Proposals these structural opportunities and impacts should be fully explored by RBNZ prior to decisions being made. 13. Our practical experience as an experienced personal and business customer in the New Zealand banking system provides us enables us to judge the impact of these proposals on efficiency.

1 https://treasury.govt.nz/information-and-services/regulation/information-releases/regulatory-review￾programme/cost-benefit

3 14. The cost of the Proposals are likely to be fully recovered by the banks from their customers. This can happen more or less straight away because standard banking and mortgage contracts contain a discretion for banks to increase costs, including because of regulatory changes affecting their cost of capital. 15. Key characteristics of the regulated banking system that support the view that the banks will recover the full cost of the Proposals include: a. by international standards the extremely high concentration of the regulated banking sector in New Zealand (the IRB four banks have approximately ninety percent market share) 2 b. the super-normal returns achieved by the sector in Australasia3 c. by international standards the low cost to income ratio of the regulated banking sector in NZ4 d. the small size of the financial and professional services sector employment market that acts as a transmission mechanism to share strategy and pricing that in turn supports the oligopolistic nature of the sector e. lending ‘hurdle rates’ and pricing that are ultimately determined by the shareholders in Australia, who RBNZ has no or limited ability to influence. 16. These characteristics also support the recent public statements by banking sector leaders in New Zealand to the effect that the costs will be fully recovered by the banks from their customers.5 Shareholders will not pay. 17. However the cost of the Proposals will fall unevenly on customers. Customers whose banking arrangements are less contestable will face disproportionate costs that do not reflect their risk to the regulated banking system. 18. These less contestable customers will have at least some of the following characteristics: a. older and less capable individuals b. limited management capability c. credit experience that is difficult to assess and communicate d. high leverage e. complex and novel financing arrangements f. substantial costs of rearranging and moving their financing arrangements g. out of favour with the regulated banking sector h. cannot access unregulated or disintermediated financing. 19. Costs of the Proposals will fall on customers not necessarily because they disproportionately contribute to the risk of the banking system but simply because they can be charged more or offered less by the banks.

2 Bullock M (2017), Big Banks and Financial Stability, Economic and Social Outlook Conference, Melbourne 21 July 2017 3 Grattan Institute (2017), Competition in Australia. Too little of a good thing? 4 Bullock, Big Banks and Financial Stability 5 RNZ statement by David Hisco, CEO of ANZ, 2 May 2019

4 20. These less contestable customers are where a lot of New Zealand’s productivity and employment growth sits. 21. We expect the following types of business customers would be very significantly negatively impacted by the Proposals: a. Agri-sector b. SME’s and fast-growing enterprises c. Construction and property development d. Infrastructure 22. Indeed the perverse effect of the Proposals is that the regulated banking sector will favour the very sector that generates substantial systemic risk – the residential housing sector. 23. Given the market power of the banks and their likely cost recovery strategies it will only be by chance that those that most impact on system soundness pay the price of their impact. The Proposals run the risk of being economically unfair and therefore inefficient. 24. The misallocation of the cost of the impact of borrowing activity across sectors and individual borrowers could have a substantial negative affect on the New Zealand economy. 25. Related to this cost misallocation point any cost of inefficiencies resulting from problems with the RBNZ’s capital adequacy rules will be substantially increased by the RBNZ’s proposals because the banks will have to recover more capital cost. 26. The RBNZ’s capital adequacy rules will matter even more under the Proposals. 27. For example, we have identified a significant problem with the treatment of bank lending to private housing co-operatives under RBNZ’s capital adequacy (credit risk) rules, the cost of which will be substantially increased as a result of the Proposals. Capital adequacy rules and private housing cooperatives 28. Private housing cooperatives: a. have existed in NZ for many decades b. were the preferred form of tenure for multi-unit residential property in New Zealand until the Unit Titles Act came into being in 1972 c. have been successfully regulated in New Zealand d. are no more risky than residential housing and probably have superior credit characteristics compared to residential freehold tenure e. are very common in other reputable banking jurisdictions (for example Denmark, Germany, Norway and Sweden) where lending to them is supported by their favourable credit experience over many housing cycles f. lending to them is recognised by both the national and supra-national regulators as sharing similar risk characteristics to freehold tenure residential housing lending and are able to be managed by banks accordingly

5 g. are approved for being treated as residential property for bank lending by both the Basel 2 and Basel 3 rules. 29. However, RBNZ’s capital adequacy (credit risk) rules and, with the knowledge of the RBNZ, the banks’ interpretation of those rules, cause lending to private housing cooperatives in New Zealand to be treated inappropriately as commercial lending by banks. 30. There does not appear to be any substantive justification for the RBNZ’s treatment of lending to private housing cooperatives under its capital adequacy rules, reflecting the RBNZ’s macro rather than meso analytical perspective 31. All of this at a time when many or possibly most New Zealanders consider that the market for the provision of residential accommodation is not working for them and home ownership is steadily declining. New Zealanders are probably also beginning to understand the real connection between their experiences in the market for residential accommodation and the performance for them of New Zealand’s banking system. 32. Private housing cooperatives offer opportunities to provide more affordable accommodation in New Zealand with greater amenity than other forms of tenure in some situations, aligning with the new “well-being” purpose in s1A of the Reserve Bank Act. 33. A clear example of the benefits of private housing cooperatives is in the provision of accommodation to retirees. We note that: a. there are over 700,000 New Zealanders over the age of 65 years and this number is increasing rapidly b. less than ten percent of the cohort can afford accommodation in retirement villages c. the cohort’s investment balances are low d. the number of New Zealanders reaching retirement with debt is rapidly increasing e. average incomes in retirement are low and mostly the result of government subsidy f. accommodation can be unsatisfactory for financial, amenity and social reasons g. their choices for alternative affordable accommodation are very limited h. the cohort’s occupation of inappropriate accommodation typologies creates direct and opportunity costs for themselves and other cohorts of the New Zealand population. 34. Private housing cooperatives for retirees offer demonstrable financial, amenity, health and social advantages compared to freehold tenure. 35. However the RBNZ’s approach contributes to the difficulty of accessing these private housing cooperative advantages for New Zealanders by affecting: a. availability of finance b. cost of finance c. structure and term of finance.

6 36. The RBNZ’s approach stifles innovation in the provision of residential accommodation and contributes to the distortion of tenure choices and therefore reduces competitive neutrality and efficiency amongst those choices. 37. Indeed the RBNZ’s implicit and explicit positions in regard to lending to private housing cooperatives not being a retail exposure probably prevents the New Zealand independent ratings based banks making normal, sensible lending decisions that other banks are making perfectly well in other jurisdictions. 38. We note that parallel problems may exist for the many co-development housing projects that are proposed in New Zealand. This problem may also exist for infrastructure credit exposures (for example, we understand that it caused problems for the Milldale housing development north of Auckland). 39. The cost of these unnecessary difficulties will be increased by the Proposals. 40. We submit that soundness and efficiency require the RBNZ to: a. treat review and improvement of its capital adequacy rules as “business as usual” b. have an efficient, timely and well-resourced process for the on-going review of the impact and appropriateness of the RBNZ’s capital adequacy rules c. assess impacts of the Proposals at a meso as well as at a macro level d. immediately undertake a considered and analytical review of the treatment of the IRB banks’ credit exposure of housing cooperatives e. be open to proper discussion about these matters with the little guys and not just the banks. 41. Meeting its statutory soundness, efficiency and well-being objectives under the Proposals makes RBNZ taking action in regard to the capital adequacy (credit risk) rules a natural part of bank capital review. RBNZ’s social licence to regulate 42. While both the RBNZ and the banks appear to be occupied by an institutional contest over the macro-economic merits of the Proposals there is little doubt that the impact of change will be borne, not by them, but by the little guys - borrowers and lenders and ultimately all New Zealanders. 43. The costless gains, like more efficient lending as a result of changes to the RBNZ’s capital adequacy (credit risk) rules should be grabbed by both the RBNZ and the Banks. 44. RBNZ’s endurance as a prudential regulator at least, is ultimately not by virtue of its legal existence, but rather by a social licence to regulate. Maintaining RBNZ’s licence is about maintaining the trust and confidence of the New Zealand public in its ability to achieve the right balance between system soundness and efficiency. 45. Meaningful and sensible demonstrations of RBNZ’s commitment to the little guys is an important element in building and maintaining the public’s trust and confidence.

Good morning Thank you for the opportunity to submit on this discussion paper. I absolutely support this initiative in both my personal and professional capacity. I own and operate multiple businesses so cannot be accused of bias towards my own vested interests. Banks in New Zealand have been under capitalised for far too long now and have taken advantage of the limited financial scrutiny imposed upon them by our regulators. They have been able to make phenomenal returns on other peoples money, which is all good and well in a capitalist society. What I have issue with is when the music stops and due to this under capitalisation banks become financially stretched to the point where they seek help from external parties. They may start asking for help from depositors in the form of a savings haircut, funds not owned by the bank but used to help stabilise their balance sheet. In effect this is stealing depositors funds. If this is not enough and a bank is deemed too big to fail, then they will ask for help from the Government, a form of nationalisation of a private institution could result. So the profits that were generated in the good times quickly get shifted into shareholders hands but when the time comes for the shareholders to put some money back in and support the ailing institution, invariably they wont be able to do so and therefore 3rd parties are relied upon to come to the rescue. This is morally and ethically wrong. If this all seems so off beat and farcical, we do not need to look too far back in history to see examples of this. The Global Financial Crisis in 2007/2008 saw multiple British and Irish banks being nationalised by their governments to prevent them from falling over. This also occurred in the US, Iceland and multiple European countries. This could have been prevented if these banks were required to hold more of their own capital on their balance sheets. Why should they be allowed to distribute nearly 100% of their profits year after year as dividends so as to appease the markets? Such profits must be retained so that banks are actually lending their own money, not that of depositors. We have a culture of borrow, borrow and more borrowing. This extends to our banking sector with only 10.5% of a banks balance sheet its own equity. To highlight just how bizarre and risky this is, how many business customers (residential customers excluded) of our main trading banks do you think have balance sheets showing a Debt to Total Asset ratio of 90%? Very few I would imagine as banks try to minimise the risk of loan default by ensuring its business customers are well capitalised. That to me seems ironic and somewhat hypocritical given that the majority of them operate with such a weak capital structure, relying primarily upon borrowed funds and depositors goodwill. They are all at risk of default given such an overreliance on external funders including depositors. New Zealand’s banking sector was saved form the ravages of the GFC from a default perspective but did suffer as a result of the worldwide credit crunch that quickly followed. This undoubtedly resulted in business failures, bankruptcies and unemployment but would have been far worse if one of our big four banks did fall over at the time. Furthermore many finance companies fell over during this period due mainly to the fact they were undercapitalised, as well of course as being under regulated. That is not the scope of this submission but serves as a timely reminder that we as a country are not immune to such events from occurring and we must collectively look to strengthen our financial institutions whichever way we can. OIA s9(2)(a)

I think the Reserve Bank should be applauded for this overdue attempt to strengthen our banking sector. In fact they should do more ie: increase the proposed buffers to total 15% so as to ensure a repeat of the Global Financial Crisis does not occur here in New Zealand. There is nothing wrong with our financial institutions having “fortress” balance sheets. Thank you. Andrew Hocken OIA s9(2)(a)

Andrew Labrooy Does the Reserve Bank really need the Capital Adequacy Ratio to be 16%? Why 16%? Why not more? Why not less? Why not leave it alone? It makes no difference whatsoever as to whether it is 8% or !0% or 16%. If a bank is going to fail, it is going to fail no matter how many % the CAR is put up to. It looks like the RBNZ is doing this just to comfort themselves that we tried to tell you and stop it. Even if it is 100%, it would make no difference because it would in effect wipe the bank off the the face of the earth just like Barings Bank in 1995. I think the best bet would be for all full licensed banks to be be listed on the NZX and be a separate entity from its parent bank, thereby ensuring that the authorities in New Zealand will have full control over their activities and the failure of the parent bank should and would not have the same effect as the listed bank here. Increasing the CAR would only entail more suffering upon us Kiwis as the banks struggle to maintain the ratios. We will ultimately be paying for the capital boost through higher fees and interest rates and who is to say the can't and won't do so? So, please RBNZ, thank you for the thought and exercise but let the dog sleep. As it is they are already over the minimum CAR. Thank you. OIA s9(2)(a)

12 May 2019 The Reserve Bank Via Email Capital Adequacy and Impact on SMEs It is my pleasure to provide a short submission on the possible impact the proposed changes would have on the SME economy in New Zealand. Context – Icehouse At the Icehouse we are small business in New Zealand. We were established in 2001 and our customers have over this time created 27,000 new jobs and generated $15.5b total revenue of which $3.8b is international revenue. Our ambition is to see a 10% impact on the country’s GDP by 2021, and we are on track to achieve that. We work with two primary segments of the small business economy, high-tech start-ups (that we support via investment, value-add) & established SMEs (that we support via learning and development and business coaching). Each year, we invest in just over 50 start-ups (representing around 30% of the market) and work with around 1,000 SMEs (representing a very small % of the total SME market, but a relatively high % of the SMEs with > $3m turnover). The focus of our commentary relates to the ‘established SME’ segment of the economy. What we have learned about SMEs in New Zealand is that the single greatest factor that impacts on their performance, both up and down is the mindset, confidence and belief systems of the owners. If they are confident, and they couple this self-belief with a rational, clear plan to progress then they are more likely to succeed, more likely to grow and some of which will go onto becoming material New Zealand and international firms that employ hundreds and thousands of Kiwis. The challenge of SME in thousands upon thousands of firms in New Zealand is that they do not have the mindset to want to grow. For those that do want to grow, they are constrained by their own capability, the size of our markets and the provision of funding, capital that can grow with them and their business over time. Context – New Zealand We have poor productivity, and we also have a reasonable issue with the structure of our economy. We are highly dominated by Government and Large Enterprises who each in their own way, do not create the right conditions for a productive economy. Over 50% of our largest companies ie > 100 FTEs are focused on the NZ market only. Our large firms, are significantly smaller that large firms in comparative countries. Our large firms are way less geographically diversified. Many of our large firms also have high market shares in NZ and are not motivated, incented or required to innovate – most are focused on staying out of court from the regulator. Most ‘dominate’ our markets, dominate access to talent, dominate access to mindshare with Government and often dominate provision and access to funding. We will not address our productivity until we work out how we can improve the performance of our large firms and /or bring new larger firms into the segment. We need to see a ‘shift’ in the segment

curve of micro – small – medium – large to such an extent that we can change the game economically. This means we need to see a greater degree in % and # terms of firms advancing from one phase or segment to the next. We need a shift to the ‘right’ and there will be a number of factors that will determine if we can be successful including trade legislation, technology changes, access to talent and access to capital. There are many things working in New Zealand at the same time – we are exporting more, we are creating a few scaling firms internationally and we are starting to see start-ups come through and be significant – there are some good things happening! Context – Reserve Bank Changes (proposed) Moving to the proposed changes, the question for me is ‘how’ will these proposed changes impact on the SME economy when we need the SME economy to perform, given the ‘large’ economy is not. If the proposed changes lead to a tightening of capital provision that prevents small business growing or worse leads to capital starvation of small business, could the market situation worsen. This is particularly relevant for the SME market because SMEs are price takers, not makers. Large firms will just adjust, use their negotiating leverage with the Banks and will crowd out SME. These comments are not as relevant to the likes of Rocket Lab, Mint Innovation, Halter and the stable of high-tech start-ups that are carving a niche in the world – because their access to capital is from private markets, here and internationally. We don’t see the proposed capital adequacy changes having a material impact on the small number of firms in these niche. Moving to the ‘volume’ part of the SME market - currently, we have a material shift happening generationally with home ownership that is and will change the way SMEs are funded in the future. It is happening now. Banks in NZ are very happy to lend to any business that is fundable and has reasonable security to put against the loan. If home ownership rates continue to decline, what will happen with these changes on capital supply to SME – I believe they will be constrained. We also have in comparison to other markets, such as the UK a lack of innovation and multiple players providing growth funding access to SME. The lack of more variable and innovative funding is another factor that creates context for NZ – we urgently need open banking regulation and we need more options for SMEs to get funding to support their growth. So we have two issues/trends in the market now – change in home ownership and a lack of innovation in the access to funding for firms. With the proposed changes, I cannot see how this will be a good thing for Small Business and I believe it has the risk of it being materially negative for SMEs and ultimately New Zealand. The RB changes will I am sure lead to constraining growth and ultimately success of the small business economy. Small business will be crowded out. What would we like to see from the RB Changes We would like to see a carve out of the changes for SMEs. Large firms can look after themselves – they all have lobbyists, resources and sophistication to work out how they find new sources of funding if there is constraint, and/or they will just pay more for access to finance and/or they will just source from offshore.

SMEs are price takers and they have little leverage – there is no transparency in the market of funding to SME. Their asset bases are fundamentally changing and we have a material market share held in banking by a small number of players. Anything that constrains that market, will in the end have the greatest impact on the ones that have least negotiating power – SMEs. My preference is that the Reserve Bank considers ‘how’ it can contribute to changing the supply of capital for our most critical SME market. Could it contribute to the recommendations that the Small Business Council are making to the Minister of Small Business in July, 2019? Could it recommend changing regulations in NZ? What more could it do to enable a more productive funding environment rather than constraining? Why do we not have open banking in New Zealand? Why do we not have many other types of funding options for small business? Concluding comments At a time when we need the best out of the largest segment of the market, small business it is not a time to strangle the provision of capital to SMEs. We have a generational shift in home ownership, we have a sector that is behind best practice for innovation and this coupled with the RB changes proposed would have a materially negative impact on the growth and survival of many SMEs. Now is the time to encourage, enable and unlock capital to growth minded SMEs in NZ. Now is the time for more innovation coming into the market for providing funding to SMEs, not limiting or controlling it. There is a power imbalance in NZ where unfortunately organisations like the Reserve Bank or Councils or Government tend to make changes that are focused on a few, but impact on many. Please reconsider the changes you are proposing and/or if the Bank still considers they are necessary that there is a carve out or provision of funding of small business is excluded. Regards Andy Hamilton1 ONZM CEO – The Icehouse

1 Andy Hamilton has written separately on micro-economic performance in NZ and how important SMEs are to NZ’s productivity and performance issues. There are two series of articles, first on a Systems Thinking approach to unlocking growth in SME and 2ndly an article on how to enable high growth firms in NZ. These articles can be found on my name at www.medium.com

Angela Sands I personally think it is imperative that consumers are given the utmost confidence in being able to 'save' via the banks. If not the banks, then where? Particularly when one can be shy of share market fluctuations, more to the point, the visibility of them. It's not about head in the sand mentality but IF NZ wants to drive a value of being savers, we need to know that the platforms we save into are not going to crash or start dipping into our hard work

• when we all know how greedy they are at the best of times. IF not the banks, then an entity, a platform that IS guaranteed by the government - at least each investment up to $250k like they did in Australia. Each investment. Not each individual. We don't WANT to have to spread our money around different banks just because we can't have that assurance. We should be able to bank with an entity and simplify things. To put it in perspective, I bank with six banks out of 'fear' of what would happen in a crisis. A bit ridiculous I know, but prudent? It is as it stands because we just don't have that confidence - yet. Please bring in more robust measures. The banks are threatening increased mortgage rates, but really? What about their millions of dollars of profit? Anyway, my thoughts. Thank you. OIA s9(2)(a)

Angus Burrows No the banks should not have to raise their capital requirements. NZ banking environments is already one the safest in the world because of its high capital ratio. The RBNZ has not completed a cost to value/benefit analysis on raising capital levels. ultimately the public will have to incur the expense of raising capital requirements as banks will tighten their lending appetite and hold more money in reserves - this will de-stimulate growth in NZ and reduce inflation. OIA s9(2)(a)

Adrian Banking Separation: Fundamentally, would provide distinction and influence policy. Policy such as MSD/WINZ 'Deprivation of Income' would be influenced in a positive way i.e. separation of banking is a statement that, in many cases hard earned, saved funds should be made safe, and not required to be invested (risked) under such policy. Policy is essentially driving expectation that every individual know investment principle when they just want, indeed believe already that, their money is safe. Banking Separation would provide -

1. a clear understanding and therefore choice for individuals of 'safe vs risk', and
2. a clear distinction of that choice for individuals and policy alike, and
3. a statement that policy, such as MSD/WINZ 'Deprivation of Income', cannot expect risk of 'safe money' lest policy be more clearly defined. Too big to fail: New Zealand's four big banks, is fundamentally, Australia's four big banks. Global crises, or more localised, repercussion is a serious effect, and while holding greater volume of capital is safer than less, separation would provide clear additional benefit and safety from this effect. Banking Separation would allow that particular banking division to focus on their distinction and simplify its structure, making it more transparent. Smaller divisions may allow for more banks to enter the market, which would spread risk (or safety) wider, allow more choice, and negate concern of 'too big to fail'. "How much capital is enough?": An open ended question deserves an open ended answer... And Banking Separation may adjust this question, fundamentally. While I am not a significant responder, these values hold significance to me and many everyday New Zealanders. I request my submission is not publicly released without express permission, but hope it helps somewhat. Thank you for the opportunity to provide feedback on the Review of the capital adequacy framework for registered banks. Anonymous-1

REVIEW OF THE CAPITAL ADEQUACY FRAMEWORK FOR REGISTERED BANKS SUBMISSION 8 May 2019 Anonymous-2

2 1 Introduction The purpose of my submission is to bring to the attention of the Review Team the proposals in my unpublished paper “A Framework to Facilitate the Coordination of Monetary, Macroprudential and Microprudential Policies” (“my paper”) for modifying today’s modern banking systems. A copy of that paper is attached to this submission. My paper was originally written for a British audience, hence the use of £. While the proposals set out in my paper have only been developed to a very high-level and require a multidisciplinary analysis and assessment of how they could be adopted and implemented, they may provide useful “food for thought”. They will also likely require a recalibration of the Basel Capital Accords and a number of other standards set by various committees of the Bank for International Settlements. I believe that the Bank for International Settlements, as the primary global standard setter for prudential regulation of commercial banks, is best placed to fully assess the proposals set out in my paper, a process I suspect will take years and one that I would very much like to be involved with. Central banks, central bank policy frameworks and prudential banking regulation and supervision have largely evolved in response to financial crises, with each structural or regulatory response reflecting beliefs about the factors that contributed to the particular crisis and to the instability of the banking system. However, it seems that the predominant assumption behind virtually all central bank policy frameworks and prudential banking regulation and supervision is that commercial banks act as intermediaries between savers and borrowers, taking deposits and raising debt and equity funding and using the proceeds to make investments and loans. This is simply wrong. Today’s commercial banks create most of the money we use in the real economy. My research shows that: • A stable and quality supply of commercial bank money (deposits) is fundamental to economic activity. • Commercial banks create commercial bank money (deposits) when making investments or loans involving non-banks. • Commercial banks do not use customer deposits, debt or equity to fund (ex ante) investments or loans. • Commercial bank money (deposits) is destroyed when investments or loans held by commercial banks are repaid, redeemed, repurchased or purchased by non-banks. • Commercial banks generally do not use their assets to repay their liabilities rather their customers use commercial bank money (deposits) (a liability) to repay, redeem, repurchase or purchase commercial bank assets (investments and loans). • The real issue is not how to “absorb” investment and loan losses with capital but how and when to destroy the commercial bank money (deposits) that was created when commercial banks made investments and loans that have been written-off.

3 • Commercial bank money (deposits) created by investments and loans that have been written-off needs to be destroyed to reduce a commercial bank’s liabilities to offset the decrease in the commercial bank’s assets caused by the recognition or realisation of investment and loan losses. • The destruction of commercial bank money (deposits) to offset investment and loan losses is achieved by a commercial bank earning income, which reduces the commercial bank’s liabilities (commercial bank money (deposits)) and recognising and realising investment and loan losses, which reduces the commercial bank’s Profit and, therefore, its ability to create commercial bank money (deposits) by paying dividends and other discretionary revenue expenditure, which, if made, would normally result in the creation of commercial bank money (deposits). • Current minimum regulatory capital requirements and the existing accounting equation are likely to increase the pro-cyclical affect of commercial banks during economic downturns when investment and loan losses reduce commercial banks’ regulatory capital, which will cause commercial banks to curtail their investing and lending activities, potentially prolonging any downturn. Because of the above findings I propose that the banking system be modified by: . t The above proposals are explained in more detail in my paper. Both my research and the above proposals for modifying current banking systems are likely to have far reaching consequences for macroeconomics that will need to be worked through. Because I was not intending to make any submissions on the Review or the Capital Adequacy Framework for Registered Banks due to other time commitments and because my research and the proposals in my paper require a fundamental rethink of the legislative frameworks for our monetary and banking systems I only propose making some very high level submissions on the role of capital in regulating banks. Submissions that by their nature are very generalised. 2 The Role of Capital in Regulating Banks Basel III, a global regulatory framework for resilient banks and banking systems, seeks to improve the banking sector’s ability to absorb shocks arising from financial and economic stress, whatever the source, thus reducing the risk of spill over from the financial sector to the real economy.

4 But minimum regulatory capital requirements do not act as shock absorbers allowing commercial banks to both “absorb” investment and loan losses and to continue making new investments and loans. This is because minimum regulatory capital requirements really act as a constraint on commercial bank balance sheet growth. Minimum regulatory capital requirements are a leveraged compliance ratio. With minimum regulatory capital requirements of 8% of risk weighed assets, every $1 of regulatory capital allows a commercial bank to create $12.5 of risk weighted assets. This means that if the ratio of risk weighted assets to total assets is 0.7 then $1 of regulatory capital allows a commercial bank to create $17.85 of assets. A process that operates in reserve when a commercial bank uses a $1 of capital to “absorb” an investment or loan loss. Increasing regulatory capital requirements to 16% of risk weighted assets (6% being a minimum requirement and 10% being a prudential capital buffer) means that for so long as the regulatory capital requirement is binding (i.e. banks do not want any restriction of their ability to distribute earnings) every $1 of regulatory capital allows a commercial bank to create $6.25 of risk weighted assets. So if the ratio of risk weighted assets to total assets is 0.7 then $1 of regulatory capital allows a commercial bank to create $8.92 of assets. While this may prevent fire sales by commercial banks in times of financial and economic stress it may not alleviate the pro-cyclical effect of regulatory capital requirements, largely because it is unknown how commercial banks will respond in a crisis. After suffering significant investment and loan losses that results in a commercial bank’s regulatory capital level falling below the prudential level but remaining above the minimum level will it: • sell assets so that it remains above the regulatory capital level to ensure that it is not subject to any restriction on its ability to distribute earnings; • issue new capital1 to raise its regulatory capital level to remove any restriction on its ability to distribute earnings; • continue to make new investments and loans to the real economy and, therefore, increase the supply of commercial bank money (deposits), even though the commercial bank may be subject to restrictions on its ability to distribute earnings. In the case where the commercial bank continues to make new investments and loans to the real economy but does not sell assets or issue new capital (i.e. the commercial bank increases its level of regulatory capital organically through retained earnings) there is also the issue of the rate at which the commercial bank will make new investments and loans that the real economy desperately needs immediately after an economic shock or financial crisis. Will the commercial bank make new investments 1 In times of financial and economic stress commercial banks are likely to experience difficulty raising additional regulatory capital due to the difficulty investors have assessing the extent of additional losses a commercial bank may suffer, the value of the commercial bank’s remaining viable assets, the forced sale of which may result in further losses, and the consequence that raising additional regulatory capital is more likely to benefit the commercial bank’s creditors, whose claims become less risky after the injection of new capital.

5 and loans at a high rate, risking the potential of being subject to restrictions on the distribution of earnings for a longer period of time, or a low rate, removing any restriction on the distribution of earnings as soon as possible. The unintended consequence of a higher regulatory capital level is that a commercial bank’s management is likely to engage in regulatory arbitrage to pursue short-term goals that suit them and not the wider economy. Higher regulatory capital requirements could also have a disinflationary or deflationary effect as commercial banks seek to decrease their balance sheets and, therefore, the money supply. Although others have commented publicly that higher regulatory capital requirements will cause interest rates to raise. If true, I would have thought that to the extent such higher rates affected economic activity then the solution would be for the Reserve Bank to reduce its short-term target interest rate. However, this may have the unintended consequence of permanently lowering the short-term target interest rate and creating a narrower interest rate band within which the Reserve Bank can conduct monetary policy. As discussed in more detail in section 4 of my paper, with the exception of retained earnings bank capital does not “absorb” investment and loan losses. This is because the various instruments that comprise a commercial bank’s capital do not “absorb” investment and loan losses because these instruments generally cannot be cancelled2. I also believe that it is a mischaracterisation to say that bank capital “absorbs” investment and loan losses first. This is because capital is the last to paid out on the liquidation of an entity using ordinary corporate insolvency laws. Accordingly, I’m not even sure what is really meant when it is said that capital “absorbs” investment and loans losses. Commercial banks are special because they neither use their assets to repay their liabilities, so generally can’t become insolvent in a conventional sense, or use capital to make investments and loans so recapitalising a commercial bank is generally a waste of time, except to comply with an arbitrary leveraged regulatory compliance ratio. A non-bank is insolvent if either it is unable to pay its debts, or its liabilities are greater than its assets. But commercial banks are different. A commercial bank can create commercial bank money (deposits) to pay its debts. A commercial bank generally commits to “settle” its liabilities by issuing an equivalent amount of commercial bank money (deposits), which is irredeemable (“fiat money”). As discussed in Section 5 – Commercial Bank Failure – of my paper, a commercial bank generally does not use its assets to repay its liabilities. Instead its customers use commercial bank money (deposits) (a liability of a commercial bank) to repay 2 Even if the instruments that comprise a commercial bank’s capital could be cancelled to “absorb” investment and loan losses this would result in the commercial bank recognising cancellation of indebtedness income, which would result in an increase in Retained Earnings. This would only result in a wealth transfer from the holders of the capital instruments being cancelled to equityholders.

6 commercial bank assets (investments and loans). Commercial bank money (deposits) once created (when a commercial bank makes an investment or loan) circulates through the banking system as money – a medium of exchange – until destroyed (by the repayment, redemption or repurchase of the applicable investment or loan). Although a commercial bank cannot become insolvent it can, however, become non￾viable if it is unable to destroy more commercial bank money (deposits) than it is obligated to create. Due to preconceptions about a commercial bank’s safety, customers may withdraw their deposits causing a bank run. Commercial banks generally hold a small amount of liquid assets in the form of cash (generally central bank money (deposits and vault cash)) to facilitate a sudden withdraw of deposits. Provided a commercial bank remains viable (i.e. it can destroy more money than it is obligated to create) any money withdrawn from a commercial bank will eventually be returned to the commercial bank and destroyed when interest, investments and loans are paid or repaid according to their contractual terms. Because commercial bank money (deposits) is used as a medium of exchange it can freely move throughout the banking system from one commercial bank to another through the interbank payment system resulting in commercial banks either having a surplus or a deficit of central bank money (deposits). Commercial banks that receive customer deposits of the commercial bank experiencing the bank run will receive an increase of central bank money (deposits) (an asset) and commercial bank money (deposits) (a liability) as a result of the commercial bank money (deposits) being transferred from one commercial bank to another through the interbank payment system. Neither of which can be used by the receiving commercial bank to make new investments and loans. A bank run or liquidation of a commercial bank is only likely to result in the redistribution of the commercial bank’s assets and liabilities amongst the other commercial banks in the banking system even if the “failed” commercial bank ceases to exist. If definitely does not result in commercial bank money (deposits) being repaid. It seems to me that the real issue is not how to “absorb” investment and loan losses using capital. But how and when to destroy the commercial bank money (deposits) that was created when a commercial bank made the investment or loan that has been written-off. This is because that commercial bank money (deposits) will now not be destroyed by the issuer or borrower repaying the applicable investment or loan. This begs the question of whether or not both expected and unexpected investment and loan losses need to be prepaid to such a low probability of default (i.e. once in every 200 years) or whether they could be off-set over the business/financial cycle in part before a time of financial and economic stress and in part over the subsequent recovery. Commercial bank money (deposits) that was created by the investments and loans that have been written-off is destroyed by commercial banks realising income, such as

7 interest income, fees and commissions when customers pay with commercial bank money (deposits), which results in the destruction of commercial bank money (deposits), recognising investment and loan losses, which decreases a commercial bank’s profit, and, therefore, reducing the ability of the commercial bank to pay dividends and other discretionary revenue expenditure that, if made, would normally result in the creation of commercial bank money (deposits). It seems to me that the mechanism for dealing with a commercial bank’s investment and loan losses needs to be separated from the mechanism that constrains the ability of a commercial bank to grow its balance sheet by making new investments and loans. This would also allow the adoption of a legislative framework to permit a central bank to resolve a commercial bank, so long as it remained viable (i.e. it can destroy more commercial bank money (deposits) than it is obligated to create), without the need for taxpayer support. 3 Contact Details 4 Request that Submission be Anonymised I request that my submission be anonymised. 5 Parts of Submission not to be published or released under the Official Information Act 1982 I request that my name and contact details are not published or released under the Official Information Act 1982 (“OIA”) pursuant to section 9(2)(a) to protect my privacy. I request that any information relating to the proposals for the reform today’s banking systems set out in my paper are not published or released under the OIA pursuant to section 9(2)(b) because that information represents approximately 10 years of unremunerated research, which I hope to monetise at some stage. For this reason, I have chosen not to publish my paper. The information that I do not want published or released under the OIA is set out in sections 6 to 13 of my paper and any references in the other sections of my paper to information in those sections. For your convenience I attached a duplicate copy of my submission and paper highlighting in yellow the information that I do not want published or otherwise released under the OIA.

Hi, I'd like all banks to be low risk with plenty of capital reserves. Increase the reserves required, please. I don't want to bail banks out if they get into trouble and I want reckless bankers jailed. I hate the OBR as I fear I will lose hard earned savings if/ when there is a crash. As I don't think the housing market is sustainable in NZ (and elsewhere), I think this is quite possible. And I don't want a "haircut" because others are, imho, greedy and reckless. I would also like financiers and developers stopped from using bankruptcy to avoid their debts...we all know they manipulate the system with lawyers,Trusts and loopholes. (I know this won't ever stop happening, realistically, though!) I work hard for very little at the moment in the hope I will eventually have a worthwhile business. I think good things take time. I am risking my capital - so am able to understand risk -but its my money and my choice. But when it comes to the bank, I can only pray they won't fail! Thanks for letting me have my say. I'm ok with my opinion being made public as long as my name is kept private. OIA s9(2)(a) OIA s9(2)(a) Anonymous-3

I agree with the proposed Capital requirements for NZ based banks. OIA s9(2)(a) OIA s9(2)(a) OIA s9(2)(a) Anonymous-4

I am making this submission in my private capacity. While I am a member of the Institute of Directors (IOD) and Chartered Accountants Australia and New Zealand (CAANZ), and a licenced auditor, a member of the External Reporting Board (XRB) and New Zealand Audit and Assurance Standards Board (NZAuASB), this submission is my private view and may differ from any submission provided by any of those parties, should they make one. I have chosen to make general comments and observations in relation to the capital review paper rather than specifically answer the questions on pages 43 to 45 of the request for submission. Part of this is because many of the questions could be considered technical questions beyond my specific area of expertise but also I have formed an overall view that does not fit necessarily into the range of questions but is valid none the less. Increasing capital Firstly, I would like to start by saying that any increase in bank capital is obviously a sound proposal for a regulator to promulgate. New Zealand was fortunate in the way that it was spared from the more significant impacts of the last Global Financial Crisis (GFC). It would be fair to say that some of this was due to the management by our registered banks, part of it was due to the oversight of the regulators and part of it might have been due to good luck or the fact that the NZ banking system had not entered into some of the products that caused the most damage in other countries at the time the GFC hit. It goes without saying that increasing the capital base of banks will better equip them and the sector to deal with any potential significant economic event, such as a GFC. It will also help to ensure that the depositors, who receive a fixed return, are less likely to be exposed to risk, and more risk will be taken by the capital investors in the entities, who share in the variable equity returns which are significantly higher than a depositors return. Another point worthy of note is that during the last GFC, the New Zealand banks did not require any propping up and indeed were able to continue doing business in a manner that was not that dissimilar to what had occurred before the GFC, and this was all achieved with the current levels of capital. The government did introduce deposit guarantee schemes (wholesale and retail) which undoubtedly provided some comfort at a difficult time and for a period from 2008 – 2011 could have been viewed as implicit support. No bank drew on the deposit guarantee scheme. Those who did were less well governed and less well controlled finance companies did. This in itself would question whether further capital is required. In addition, the RBNZ has, in its various FSR presentations since the GFC, reported that the New Zealand banking system is in good shape, albeit with one or two hot spots that have needed cooling down. The question then becomes whether an increase in capital is warranted at all, or if one is, whether one of the magnitude suggested in the discussion paper is warranted. This question also manifests itself in whether or not the RBNZ should be aiming to insulate the banking sector against a 1 in 200 year event or some other quantum of event. I think it is also important to note that the current New Zealand calculation of capital is done on a more conservative basis than other parts of the world, such that a New Zealand bank calculating capital at say 12% would actually have a higher capital ratio if it calculated its capital under other countries’ rules. Perhaps, as part of this exercise, it would be fair to the New Zealand banking sector to bring the calculations more in line with overseas calculations, such that they were not required to carry additional capital due to our calculation methodologies alone. 26 April 2019 BANK CAPITAL REVIEW Our ref: Anonymous-5

15847979_1.docx 2 In summary, there can be little to argue against increasing capital from a perspective of making the industry or sector safer to protect both the New Zealand depositors and the availability of credit to borrowers in the New Zealand economy generally, in the case of a future negative event. However, the question becomes, when our industry’s past performance is considered, in making the changes suggested, how big will the negative impacts on the general New Zealand economy be from these changes and whether the level of safety built in by insulating against a 1 in 200 year event is appropriately conservative. Tier 2 Capital It would appear from the paper that the admissibility of tier 2 capital is being greatly reduced, if not ultimately eliminated. Is this appropriate and fair and why would we do this in New Zealand when other counties still allow tier 2 or encourage it? Perhaps one option might be that rather than eliminating tier 2 from the calculation process, that the RBNZ allow it to form part of the calculation process but with conditions. Such conditions could be that the RBNZ require that in the event a bank breaches a capital buffer or limit, the tier 2 capital must remain locked within the bank until such time as the breach is remedied, or alternatively permission is required from the RBNZ for the repayment of tier 2 capital. An example of this would be to set a base limit and buffers, and have the banks agree that once any of those were triggered, or indeed if in the opinion of the regulator the capital was trending down to a level that was close to a trigger, the tier 2 capital would be locked in as if it was tier 1 and/or require approval from the RBNZ to repay it at any other time. IRB Rating and Standard Rating Systems I believe that the proposed changes in this area appear reasonable. At present, the smaller banks that are using the standardised models will argue that the IRB banks have a significant advantage and in fact only have to provide approximately 76% of the level of capital that smaller banks do when competing for a loan. It doesn’t really make sense to think that two banks competing for the same mortgage should have to provide different levels of capital simply because of their calculation methodologies. The loan itself is not more or less risky to one bank than the other and it could be argued it gives an unnecessary advantage to the banks using IRB models. If we want the non- Australian owned parts of our banking sector to grow stronger and become a bigger part of the sector and provide more competition, then an advantage like this for the major players does not make a lot of sense. Profit returns over the next 5 years A statement is made in the document and has been made elsewhere that the required proposed capital ratios could be achieved if the banks were only to pay out 30% of their current dividend levels (or put another way withhold 70% of their current dividend levels) for the next 5 years. This statement is somewhat confusing. Recently the RBNZ and FMA have embarked on a significant culture and conduct review and in doing so have repeatedly pointed out that financial institutions dealings with their customers and other stakeholders must be done in a fair reasonable way with their outcomes put at the forefront. I cannot see how arbitrarily deciding it would be acceptable for shareholders to only receive 30% of the normalised dividend they had been used to, for a period of 5 years, is treating them fairly. Regardless of who the shareholders are, it needs to be recognised that they are an important part in ensuring the bank has adequate capital. To so blatantly suggest that one group of stakeholders should be disadvantaged in the level of return they receive is at odds with other statements made in relation to culture and conduct. History of bank failure in New Zealand There has not been, to my recollection, a significant bank failure in New Zealand in recent times. This would imply that the current levels of capital, combined with regulatory oversight and board governance of these entities, has collectively been sufficient. This would then question the need to so dramatically change the level of capital to be carried by the banks. While some increase might be appropriate, I would question whether the quantum of increase proposed is necessary, particularly when it is combined with the additional 1% proposed for the 4 large Australian banks and the effective exclusion of tier 2 capital. In the 1980s the government did bailout the BNZ but this could be considered more in its role as a shareholder and this behaviour has continued with similar bailouts for Air New Zealand and AMI Insurance and others. It will always be the prerogative of the Government to support systemically important business if they feel there is a national need.

15847979_1.docx 3 Practical Ability to Raise Capital For some of the smaller banks, raising this capital may be challenging due for example to their structure (Ex Co￾operatives) and for some of the larger banks, their parents simply may not be able to provide the capital due to home country rules around level of investment in the subsidiary or alternatively they may not have the desire to invest more capital into what could be considered a dilutive investment. It does need to be remembered New Zealand is a young country and not one with large pockets of wealth, and therefore most businesses and families require funding from banks to assist in the formation and development for their business operations or their daily lives. The possible impacts of additional capital One thing that the request for submission paper was rather brief on was analysing the potential impact that any increase in capital will have. No doubt it will strengthen the balance sheets of the sector but the flow on effects of that strengthening must also be considered The big 4 Australian banks are calling this a doubling of the capital requirement, while the RBNZ prefers to call it an increase of somewhere between 20% and 60%. Let’s put the absolute quantum of the increase aside for the moment but accept that there will need to be an increase in capital, and that the increase will be reasonably significant. If bank capital is increased, the first impact will be, with all other things kept equal, a reduction in the return on equity. A reduction in the return on equity will, with all other things being held equal, lower the share price or investment value of that business, and the business will be faced with a number of potential decisions and outcomes: 1 Do nothing and see a reduced share price and reduced attractiveness to investors. If any of these shares are held in investment portfolios, then those portfolios will suffer a decline in value as well. Some of these portfolios will be Kiwisaver portfolios, where many New Zealanders hold investments. 2 Look to increase profitability, so as to restore the return of equity against the bigger capital base to bring it back in line with similar peers. 3 Look to divest of the dilutive investment, or parts thereof, that are capital intensive (in a capital raising area). Assuming that no bank will accept a reduction in return on equity and do nothing, choosing between options 2 and 3 are more likely. If you look at option 2, most banks will make an attempt to increase their profitability such that it is sufficient to restore all or the vast majority of the diminution in return on equity. If they currently have a return on equity of, for example, 15% and the extra capital reduces that to say 10%, they will look to increase their profit accordingly to return the return of equity to 15% or similar. This will result in them having to look at considering a number of things. It is not correct to assert that competition will keep rates etc. where they are, as all participants are affected and those affected the most make up 80+% of the market: 1 The first option available to increase earnings is to decrease deposit rates. This is an undesirable outcome in New Zealand, as a number of super-annuitants and others are reliant on deposit interest to fund their lifestyle. If they are faced with a reduction in the income they are receiving, they may be tempted, as they were prior to the GFC, to invest their money in higher yielding, and therefore higher risk, investments with a greater chance of loss. This was clearly seen in the GFC, where approximately $6bn worth of money left the banking sector and went into the finance company sector. A proportion of this money was subsequently lost when some of these finance companies, who were the recipients of that money, supported what ultimately turned out to be failed property investments. Some of the depositions were spared a loss where the finance company entity was covered by the government guarantee scheme but in many cases this was not the case. New Zealand has a love affair with property investment, and with the tax working group suggestion of a capital gains tax having been recently declined by the Government, the property market will undoubtedly return to being New Zealand’s favourite and probably most lucrative form of investment. Therefore, there will be increased opportunities (that wouldn’t have been there if CGT had been introduced) for financiers to take public money and support property developments. We don’t want a reduction in deposit rates to drive unwise investor behaviour 2 The second option immediately available to increase earnings will be to increase lending rates. New Zealand has enjoyed low lending rates for a long period of time and a number of people have taken advantage of that to

15847979_1.docx 4 become reasonably highly leveraged to increase both their personal and business assets. Despite this they are by and large managing to repay the debt. Our economy has also enjoyed a reasonably strong growth period with unemployment at low levels, and commodity prices and exchange rates supportive of businesses. There is a very real question as to whether or not New Zealand businesses and individuals could, and should, bear the type of interest rate increase that might become necessary to return the banks level of profitability to a suitable level for the sole reason that capital has been increased. Various analysts have estimated the range of the basis point increase in lending rates that could occur. This would have the impact of conducting a real live stress test on the economy. I would hope that this has been carefully modelled to ensure that we don’t inadvertently conduct a live experiment on the economy’s resilience when we don’t need to do – all to increase an already strong sector’s capital level even further. 3 The third area the banks could look at to increase earnings is to reassess their appetite for exposure to lending sectors that are causing them some concern in their lending books. This can either be areas that created an actual losses, or drove significant collective provisioning, as both of these will reduce capital at a time when the bank is being asked to increase it. Alternatively, forms of lending that are not able to provide real security, and therefore are more heavily risk weighted under the capital model, will be less desirable to lend to. The types of industry that come to mind are Dairy, Agriculture, construction but most alarmingly the SME sector. Currently, most banks have exposure to the agriculture sector (in particular dairy), and this is an area where there is reasonably significant levels of provisioning and in some cases even provisioning overlays. These are also sectors that are critical to New Zealand’s economy and export earnings. The construction sector has also suffered recently with a number of companies either having profit downgrades or going into receivership and causing losses. The New Zealand government has a larger agenda for construction ahead of it, whether it be for much required infrastructure, roading and other transport or tourism. The economy needs the construction sector to have appropriate access to credit to enable it to build the assets that our country needs to continue to grow and prosper. Anything that makes it harder to lend to these sectors, such as additional capital requirements generally or the industry exhibiting greater loss characteristics at a time when capital is tight is going to focus lending decision more closely, and tighten lending. Most alarmingly is the potential impact on the SME market from the proposed capital changes and resulting flow￾on impacts. A number of New Zealand businesses start by two or three people in an existing business having an idea that maybe does not compete directly with the business they currently work in and so potentially with or without the current business owner’s blessing, they leave and set up a business that either doesn’t really compete or acts in providing an ancillary service to the business they have left. When these people leave the existing business to set up their own, they have very little to their name other than possibly an idea, some IP, maybe two or three vans which will be on lease, and a premises that will also be leased, or they may even commence their new venture working out of home. They have, in many cases, a ready source of customers and the ideas, but little else. When they go to the bank for a loan for working capital, plant or inventory, that loan will have very little real security to support it, so will require a higher level of capital to lend against than say a less than 80% LVR mortgage. What those entrepreneurs may also have is a partly mortgaged home, and they will therefore end up moving it from below 80% to over 80% to fund the business. As a result they will be required to pay a higher interest rate because the bank requires more capital to fund a higher LVR loan. If the banks are struggling to raise the capital required, these are exactly the types of loans that will no longer be as attractive to them and therefore less freely lent to, thus possibly stunt the growth of the New Zealand economy that is so largely made up of these SME entities. Alternative ownership /divestment issues If the banks do not raise all the capital they require from their parents/ shareholders or the market to comply with the capital regulations and/or cannot increase their profits sufficiently to recover the return on equity reduction due to the additional capital required to be put in place, they will need to look at other options, such as IPO’s/partial IPOs to raise the capital or reduce the level of capital they are required to hold. This could be a partial IPO of the entire business or the packaging up of a particular sector of the loan book. One area that has been identified as a primary target would be the Agri books of the banks being packaged and sold to an overseas party. Agri comes to the fore front because it is an area that is accepted as under stress and one that many banks hold a provision overlay for. My concern would be that

15847979_1.docx 5 for over 100 of years the major banks, in their current forms and predecessor entities, have been very supportive of the Agri sector, and have shown a strong understanding of its cyclical nature, and have stuck beside and managed them during trouble times. I don’t believe that a recently packaged up and sold book of loans that ends up in the hands of an offshore bank, fund or PE house would necessarily get treated with the same level of understanding and patience as a local bank currently does. The purchase would have been made based on investment decisions and return prospects without perhaps the history of relationship of the existing bank. One real risk is that incumbent banks may move out of areas due to the impact of increased capital where they have particular expertise and our nation’s critical industry sectors may be at the behest of foreign banks and other investors, all of whom will be only looking for a return. We have historically seen foreign entities come and go from the New Zealand banking sector at a whim, because their amount invested is not significant for them, or they don’t understand the cyclical nature of the industry, or indeed maybe driven by home country imperatives or economic issues. One other matter that would be worthy of mention is that the current NZ banks have skilled workforces both in terms of experience and numbers who are used to dealing with these credit performance issues. They have built the in￾house expertise to deal with cyclical issues, whereas an investor recently purchasing a portfolio would not necessarily have that skill base. The question that should be asked is do we want the loans to some of our most important sectors in the hands of people who do not necessarily understand the sector, or who may be looking to make a quick return on funds invested, or alternatively may have different cultural responses to lending behaviours. You could look at all of these points that I have made above from a perspective of “these are things that might happen, could happen” or that this is a slightly pessimistic approach. While that comment might be valid, what is potentially alarming is the fact that, to the date of this submission, the RBNZ has not released any economic assessment of what the impact of these capital increases would be on these other areas. I am not an economist and do not have the skills to calculate what the likely impact that the increase of capital to the 16% as suggested would have on deposit rates, lending rates, the level of credit availability and the likelihood of banks being sold /partially IPO’d or packaging up sections of their loan books might have, but what I am concerned about is whether by increasing the capital requirements to the levels proposed, the RBNZ would be congratulated on how safe it has made the banking sector but the making of it safe leads to a whole raft of unintended consequences, harming the rest of the NZ economy to an extent that was not expected . Post implementation review The level of capital that has been suggested as the new “norm“ seems high compared to other countries, and when considered in conjunction with the facts that the capital calculation methods in NZ already generate a high level of base capital and the likelihood that tier 2 will be considered inadmissible, these three factors make the increase all the more harsh. Therefore it is important that there is a staged implementation as suggested. The five year proposed lead in time seems the minimum that would be appropriate, but I also wonder whether, in light of the unknown territory we may be moving into, there should be some sort of post implementation review. Such a review could be undertaken in, for example, years 2 and 4 to assess whether or not the increase in capital and the benefits from it, from a banking sector security perspective, are outweighed by any unexpected or unintended side-effects such as those discussed above. If there were any unintended consequences developing, these could be addressed by a mid-term post implementation review. The New Zealand economy is one that tends to grow at steady but lower % of growth than many others in the world, and that path could be quite susceptible to being slowed or reversed by some of the impacts of the increased capital. It would be a shame if the economic pain that the rest of the world has felt for the last 10+ years since the GFC that we have not suffered was inflicted here by a policy primarily aimed at strengthening the market/economy against such a shock event. Governance The final matter that I would like to raise is that there seems to also be a slight governance weakness in the process. The RBNZ has promulgated a proposed increase in capital and the methods by which it will occur, submissions will be sought, received and reviewed and the RBNZ will make a decision. If such an important decision was being made in the corporate world, management would bring a paper to an external board who would be involved in approving it and may ask for further supporting information. If I was a member of that board, the information that I would be seeking,

15847979_1.docx 6 which to my knowledge is currently not available, would be what the impact on the wider NZ economy was going to be from this capital increase, as while I would undoubtedly want to take any opportunity to strengthen the capital base of the sector, I wouldn’t want to do so if it harmed other areas of the economy to a greater extent that the benefit of a bigger capital base. At present, there seems to a very strong and rightful focus on strengthening the banking sector against a 1 in 200 year shock but there doesn’t seem to be the same level of assessment of what the other impacts on the NZ economy will be. Ensuring there is no unintended consequence, such as those discussed above, is in some ways as important as ensuring the sector can withstand the 1 in 200 year shock. In conclusion I would like to reiterate that I make this submission in my personal capacity and not as a member of CAANZ or the IoD, nor in my role(s) on the External Reporting Board or the NZAuASB. Yours sincerely

Anthony Everitt I support the RBNZ's proposal to raise capital requirements for banks. Under the OBR system, bank shareholders are currently accepting too little risk (responsibility) relative to their return. They are transferiing too much risk to depositors (me) relative to our return. I expect NZ banks to be safe, and currently I am not sure they are, especially with growing real estate market concerns and exposure. One would hope the market would sort all this out without regulation, but unfortunately the market is not always effective (see Global Financial Crisis). The requirement to have a liveable minimum wage is a good analogy where regulation is necessary to put an acceptable floor in the market. OIA s9(2)(a)

Anthony John Gavigan I believe that the OBR Open bank resolution system should be addresses at the same time as the issue of capital adequacy. They are related in that if capital proves inadequate, eg in the case of a complete collapse of economic confidence as predicted by https://www.armstrongeconomics.com for end of 2032. If such a collapse occurs and an OBR haircut is required on depositor liability the haircut deposits should automatically convert by statute into new voting transferable ordinary shares and the issuer should be the authorised entity delivering the haircut. OIA s9(2)(a)

From: Anurag Shetty Subject: Capital Adequacy Framework Review Hi RBNZ, I support RBNZ’s proposal on imposing stronger capital standards on banks that match the public's risk appetite. I believe lower leverage is the key to the resilience of the New Zealand's banking sector. The bank shareholders must definitely have more skin in their investment. It's prudent the RBNZ is taking steps to protect depositors by enforcing tougher capital requirements. Regards, Anurag OIA s9(2)(a)

1 17 May 2019 Ian Woolford Financial System Policy and Analysis Department Reserve Bank of New Zealand P O Box 2498 Wellington 6140 By email: CapitalReview@rbnz.govt.nz Dear Ian ANZ Bank New Zealand Limited submission on the Capital Review Paper 4: How much capital is enough? Thank you for the opportunity to provide feedback to the Reserve Bank of New Zealand (Reserve Bank) in response to the fourth consultation of its review of the capital adequacy framework for locally incorporated banks: ‘Capital Review Paper 4: How much capital is enough?’ (consultation paper), released on 14 December 2018. ANZ Bank New Zealand Limited (ANZB) appreciates the opportunity to provide feedback on the capital ratio proposals, and would welcome further discussion with the Reserve Bank on such complex and important changes to the capital framework. ANZB believes the current capital adequacy ratio requirements and voluntary capital buffers maintained by banks are appropriate for New Zealand. Additional conservatism is unnecessary, as the current framework already produces adequately conservative capital positions. This is illustrated by international comparisons, stress testing results over several years, and the Reserve Bank’s own past analysis. PricewaterhouseCoopers’ (PwC) updated analysis1 concludes that New Zealand’s major banks are well capitalised on an internationally comparable basis, standing in the top quartile of international banks, and above the “unquestionably strong” Common Equity Tier 1 (CET1) level of 10.5% required by the Australian Prudential Regulation Authority (APRA). With regards to the proposals presented in the consultation paper, ANZB’s concerns are the absence of cost-benefit analysis of the proposals (which, if undertaken, should be subject to independent review and public consultation), and the degree to which the Reserve Bank has under-estimated the potential impact of its proposed capital regime on both the New Zealand economy and New Zealand bank customers. This submission presents analyses to illustrate our concerns, which include:  In terms of the wider economy, ANZB estimates that the long run cost of the proposals, in present value terms, would be approximately 20% of GDP versus the Reserve Bank’s estimate of 4%-12% of GDP.  The Reserve Bank has not estimated the transitional impacts of the proposals, which ANZB estimates at GDP being 1%-3% lower over 10 years. 1 ‘International comparability of the capital ratios of New Zealand’s major banks – update paper’, PwC, May 2019

2 ANZ BANK NEW ZEALAND LIMITED  For customers, ANZB believes potential increases in loan pricing and reductions in loan capital would particularly affect the agricultural and commercial sectors, and potential reductions in deposit pricing would adversely affect savers.  The potential reallocation of bank capital towards the housing sector would increase the concentration risk on this sector and leave banks more exposed to any sharp correction in residential property prices, potentially increasing financial stability risks. ANZB considers the analyses presented in the Reserve Bank’s consultation paper and supporting documents to be inconclusive and to fall short of presenting a convincing case for the far-reaching measures proposed. The Reserve Bank acknowledges it has no definitive basis for using a 1 in 200 year event (1/200) as its starting assumption2 . ANZB agrees with other published criticisms of the Reserve Bank’s conclusions on the optimal quantity of capital. The Reserve Bank’s estimate of the required increase in capital to meet its 1/200 criterion is overdone. Given the levels of uncertainty that clearly exist around risk appetite, optimal capital levels and the costs and benefits of the Reserve Bank’s proposals, ANZB’s view is that the Reserve Bank should reconsider the proposals, in terms of: i. the method of aligning internal ratings based approach (IRB) banks to standardised banks (the scalar); ii. the increase in the amount of capital deemed necessary; and iii. the composition of the ‘capital stack’ between CET1, and Basel compliant Additional Tier 1 (AT1) and Tier 2 instruments, with the objective of determining a capital outcome which provides for bank stability without the excessive costs that ANZB and others have identified. As stated in previous submissions, ANZB’s view is that regulatory capital requirements for New Zealand incorporated banks should align more closely with those of APRA, particularly with respect to AT1 and Tier 2 instruments, which would provide greater flexibility in the cost and source of bank capital. This submission comprises detailed explanations of ANZB’s position on this consultation, our concerns regarding the Reserve Bank’s proposals, as well as responses to the questions from the consultation paper, structured as follows: Part 1 Submissions regarding the Reserve Bank’s proposed capital framework. Part 2 Responses to the Reserve Bank’s questions from the consultation paper. Appendix 1 Examination of the analysis underlying the Reserve Bank’s proposed changes to capital adequacy. Appendix 2 Modelling the potential impacts of the Reserve Bank’s proposed changes to capital adequacy. Publication of submission ANZB consents to the publication of this submission, with certain aspects withheld to protect the privacy of natural persons or on the grounds of commercial sensitivity, under the Reserve Bank’s publication of submissions policy. ANZB considers that the redacted information should be withheld if a request for it is made under the Official Information Act 1982 on the basis that:

2 ‘An outline of the analysis supporting the risk appetite framework’, Reserve Bank Capital Review Background Paper, April 2019.

3 ANZ BANK NEW ZEALAND LIMITED

1. The information has been supplied to Reserve Bank in connection with a consultation on the possible exercise of its powers under Part 5 of the Reserve Bank of New Zealand Act, and therefore may not be disclosed pursuant to sections 105(1)(c) and 105(2) of the Reserve Bank of New Zealand Act 1989.
2. Making available the information would be likely to unreasonably prejudice ANZB’s commercial position, and therefore the information may be withheld under section 9(2)(b) of the Official Information Act 1982. Contact for submission Please contact or email Once again, we thank Reserve Bank for the opportunity to provide feedback on the consultation paper. Yours sincerely Sir John Key Chair ANZ Bank New Zealand Limited

4 ANZ BANK NEW ZEALAND LIMITED Part 1 ANZB’s submissions regarding the Reserve Bank’s proposed capital framework Summary: 1 ANZB submits that the current regulatory capital framework is adequate and appropriate. 2 ANZB submits that the Reserve Bank should reconsider its proposals in light of the potential impact on particular sectors of banks’ customers, including agricultural and commercial borrowers, and depositors.

1. ANZB submits that the current regulatory capital framework is adequate and appropriate. ANZB submits that the current capital requirements are adequate, and that New Zealand banks are well-placed in the global context. ANZB’s review of the Reserve Bank’s analysis is outlined in Appendix 1 of this submission, and concludes that the Reserve Bank’s case for proposed increases in capital significantly above current levels is not proven. When adjusted for differences in regulatory treatment across countries, New Zealand banks’ capital positions are already at, or close to, levels consistent with the international studies on which the Reserve Bank has based its proposals. Stress testing results over several years, as well as the Reserve Bank’s own analysis in 20123 , also suggest that capital increases of the magnitude proposed are not warranted. The Reserve Bank’s proposals are not accompanied by a cost-benefit analysis. ANZB’s estimate of the economic costs of the proposals greatly exceeds the Reserve Bank’s estimate. Other analysts, including Sapere Research4 and Tailrisk Economics5 also conclude that the costs of the Reserve Bank proposals may exceed the benefits. PwC conducted a study in 20176 that positioned New Zealand bank capital ratios against capital ratios in other jurisdictions on an internationally comparable basis. The study was updated in May 20197 , to address feedback from the Reserve Bank, and to assess the implications of the Reserve Bank’s capital proposals on internationally comparable capital ratios. PwC concludes that the 2017 CET1 capital ratios of New Zealand’s four Australian owned banks would increase by 520 basis points to 15.5% when measured on an internationally comparable basis. With the Reserve Bank proposals taken into account, the internationally comparable CET1 capital ratio for New Zealand banks would be 27.1%, an increase of 1,680 basis points on 2017 levels.

3 ‘Regulatory Impact Statement on the application of Basel III capital requirements in New Zealand’, Reserve Bank, September 2012. 4 ‘How much capital is enough – a review of the Reserve Bank Tier 1 capital proposals’, Sapere Research Group , May 2019. 5 ’The 30 billion dollar whim: A review of the Reserve Bank consultation paper: ‘How much capital is enough’’, Tailrisk Economics, March 2019. 6 ‘International comparability of the capital ratios of New Zealand’s major banks’, PwC, October 2017. 7 ‘International comparability of the capital ratios of New Zealand’s major banks – update paper’, PwC, May 2019.

5 ANZ BANK NEW ZEALAND LIMITED Source: ‘International comparability of the capital ratios of New Zealand’s major banks – update paper’, PwC, May 2019. 2. ANZB submits that the Reserve Bank reconsider its proposals in light of the potential impact on particular sectors of banks’ customers, including agricultural and commercial borrowers, and depositors. ANZB believes the Reserve Bank’s proposals would have a greater impact on the pricing and availability of lending than the Reserve Bank believes, particularly in the agricultural and commercial sectors. Also if the OCR was reduced to offset increases in lending margins as a result of the proposals, the burden of the proposals would transfer to depositors. ANZB believes that the large corporate segment and the commercial property market would become increasingly reliant on offshore banks for funding, which may have higher risk of withdrawal in periods of stress. In general we would expect a shift in lending away from higher risk weight business towards the housing market, given that housing lending would require a third to a half as much capital as other lending. ANZB believes (see Appendix 2) that the costs of the proposals to the economy over both the transitional period and the long term would be greater than the Reserve Bank believes. In the absence of a compelling cost-benefit analysis, ANZB considers that the Reserve Bank should reconsider its proposals.

6 ANZ BANK NEW ZEALAND LIMITED ANZB has estimated the potential impact on annual GDP growth over the transition period of the Reserve Bank proposals. Estimated impact on annual GDP growth over the transition period:

8 ANZ BANK NEW ZEALAND LIMITED achieved a 15% return. Impact of the proposals on New Zealand banking market competitive dynamics 12. We think it is likely that banks subject to the proposals would pursue lending margin increases. The degree to which banks can achieve these will depend on the degree of competitive pressure in each sector. Large corporate sector 13. Australian subsidiary banks that cannot lend to large corporates (i.e. generally, listed companies or subsidiaries of offshore multinationals) via branches of their parent banks would find themselves at a competitive disadvantage relative to branches of foreign banks operating in New Zealand, and also to offshore bank lenders9 . While large corporates understand the benefits of long standing relationships with local banks (as the Reserve Bank noted in the recent industry forum on the proposals), our experience is that large pricing differentials will generally win out as long as sufficient capacity from offshore banks is available. 14. 15.

16. The large corporate market could be banked mainly offshore in this scenario. As a result, large corporates would be exposed to a greater risk that offshore lenders would reduce lending to New Zealand customers if stress developed, either in New Zealand or in their home markets.

9 While it is possible, we wouldn’t expect the capital markets to provide a materially larger proportion of funding to these customers than at present. The flexibility and pricing of bank facilities has value to corporate customers.

9 ANZ BANK NEW ZEALAND LIMITED 17. The Reserve Bank suggested in the recent industry forum that offshore lenders may choose not to exit in these circumstances. However, our experience through the GFC is that they tend to do so and that the risk of a withdrawal of capital would be increased. 18. In a similar manner, the commercial property market would have access to debt funding sources other than New Zealand banks, potentially moderating pricing increases in this sector, but also increasing its reliance on offshore banks. Agricultural sector 19. 20. We note that ANZB’s share of the agricultural sector has reduced from ~39% in 2010 to around 28% currently. 21. 22. Offshore banks may increase agricultural lending. However, it would require substantial increases in participation by offshore banks/branches to offset the potential response of local banks subject to the capital proposals. 23. In summary, we believe the Reserve Bank proposals would have a material impact on lending to the agricultural sector, one of the economy’s productive sectors SME sector 24. SMEs have relatively high risk weights, depending on security levels. The prospective ROEs of SMEs would be materially reduced by the proposed framework. Housing sector 25. The housing market would experience some upward margin pressure given that virtually all current mortgage lenders would be affected by the capital proposals. 26.

10 ANZ BANK NEW ZEALAND LIMITED 28. An increase in housing lending margins may attract other entrants, such as securitisation vehicles, into the market. However, they are unlikely to have a material effect over the transition period and the underlying cost efficiency of the major banks may constrain the returns available to these entrants, as is the case now. Unsecured personal lending 29. Unsecured personal lending is not a material part of ANZB’s balance sheet (~1.5%). However, we would expect margins in this area to also be under upward pressure. Impact on customers will fall mostly on agriculture and SME sector 30. 

31. Given these potential market dynamics we think the impact on margins and availability of loan funding to the productive sectors of the economy will be significantly larger than the Reserve Bank expects.
32. In Appendix 2 we model the impact on the economy of our views on the market. In summary, the results of this modelling are: • • As a result, GDP might be 1-3% lower over 10 years (i.e. an annual impact of approximately $3bn to $9bn in today’s dollars, abstracting from price impacts which haven’t been modelled). This is consistent with international evidence that points to significant transitional impacts on both lending spreads and the volume of credit of increases in required capital ratios. • GDP would be 1% lower permanently as a result of the changes (approximately $3bn in current annual GDP terms, abstracting from price impacts). Contributing to this, as discussed above, the proposals would encourage structural adjustment in the allocation of capital away from business and farm lending in favour of housing, with negative implications for New Zealand’s productive potential.

11 ANZ BANK NEW ZEALAND LIMITED • We expect that the Reserve Bank would consider reducing the OCR to offset the impact of increased lending margins on borrowers. We would offer the following comments on this: − − Even with an offset from monetary policy, negative impacts of the proposals on business investment and farm lending cannot be avoided. − Offsetting the economic impacts of the proposals would significantly reduce available monetary policy ‘firepower’. − Reducing the OCR would benefit borrowers, but deposit rates would reduce accordingly. • We would expect a movement of deposits into other, generally more risky, asset classes. Whether this would result in funding pressures on banks would depend on the path of asset growth. • Overall, we expect the costs to the economy in terms of increased costs/lower availability of credit to fall most heavily on sectors with few other funding options and higher risk weightings, Relevance of Australian subsidiary banks’ ROEs 33. The Reserve Bank has referred to the relatively high statutory ROEs being earned by Australian bank subsidiaries as possibly unjustifiably high. High ROEs have been used in support of the joint aims of ‘levelling the playing field’ and improving competition. 34. We submit the following points: • New Zealand is currently experiencing a very benign credit environment. Current credit provisions are around a quarter of expected through-the-cycle levels. At expected through-the-cycle levels of credit provisions, ANZB’s ROE would be ~1.0% lower. We expect that credit losses will inevitably increase to more normal levels and returns will reduce accordingly. • Secondly, large New Zealand banks are efficient on a global comparison, with average cost-to-income ratios (CTI) of ~37%.

14 ANZ BANK NEW ZEALAND LIMITED 40. This is depicted in the diagram below. 41. The Reserve Bank’s calibration of the capital framework is set out in its memorandum of 13 November 2018 entitled ‘Capital ratio calibration’. In the first chart in the memo, the New Zealand ‘base case’ is set out. 42. The probability of a bank crisis decreases as the capital ratio increases. However, it is not linear: as Tier 1 capital increases the marginal reduction in probability decreases significantly. In ANZB’s view, analytical and empirical support for the proposals appears weak 43. ANZB’s analyses of the risk appetite and calibration of the capital framework are provided in Appendix 1 (‘Examination of the analysis underlying the Reserve Bank’s proposed changes to capital adequacy’) and Appendix 2 (‘Modelling the potential impacts of the Reserve Bank’s proposed changes to capital adequacy’). 44. Appendix 1 discusses the Reserve Bank’s analysis of the proposed changes and concludes that:  The Reserve Bank’s risk tolerance framework is incomplete and the chosen risk appetite is not well justified.

15 ANZ BANK NEW ZEALAND LIMITED  Taking into account New Zealand’s relatively conservative regulatory regime, international studies do not support the Reserve Bank’s risk assessment. The risk assessment is also inconsistent with modelling by major banks, stress tests and New Zealand’s historical experience.  In addition, in portfolio modelling in a New Zealand context, implausible PD and LGD assumptions have to be made in order to find that higher capital ratios are needed.  Based on international studies and our analysis, the Reserve Bank’s assessment of the costs of the proposals appears too low. In particular, transitional impacts of the changes are not considered.  Overall, analytical and empirical support for the Reserve Bank’s proposed level of capital ratios appears weak. The international literature supports capital ratios close to where they are now, taking into account cross-country differences in risk weighting. 45. Appendix 2 concludes that the economic impacts of the proposals would be larger than the Reserve Bank estimates:  

 We estimate that as a result of the proposals GDP might be 1-3% lower over 10 years (an annual impact of approximately $3bn to $9bn in today’s dollars, abstracting from price impacts), consistent with international evidence that points to significant transitional impacts on both lending spreads and the volume of credit from increases in required capital ratios.  We expect GDP would be 1% lower permanently as a result of the changes. This would be a cost of approximately $3bn in current annual GDP terms, abstracting from price impacts. Contributing to this, the changes would encourage structural adjustment in the allocation of capital away from business and farm lending in favour of housing (as mortgage lending requires relatively less capital to be held against it), with negative implications for New Zealand’s productive potential. The 15%/16% Tier 1 capital requirement 46. ANZB submits that:  The proposed RWA floor of 85% of the Reserve Bank’s version of the standardised approach (Reserve Bank standardised approach - BS2A) is too high.  Similarly, setting the scalar at 1.20 to achieve an average 90% of the Reserve Bank standardised approach is too high.  The 16% Tier 1 minimum is too high.

16 ANZ BANK NEW ZEALAND LIMITED Reserve Bank standardised approach 47. ANZB agrees that standardised risk weights should be conservative when compared to IRB risk weights, because they provide limited differentiation for risk. ANZB has compared BS2A, Reserve Bank’s standardised approach rules, and the Basel framework. ANZB found BS2A to be more conservative than the underlying Basel framework. 48. The Reserve Bank is proposing a risk weight floor for IRB banks of 85% of the Reserve Bank standardised approach. The Basel Committee on Banking Supervision (BCBS) suggests a floor of 72.5% of Basel standardised RWA for IRB banks. 49. ANZB has confirmed that the 1.20 scalar would lead to an IRB risk weight that would be ~90% of the Reserve Bank standardised approach, which would imply that the 85% standardised floor would never become binding given the current benign credit environment. 50. ANZB considers that calibration of the scalar and the floor to the Reserve Bank standardised approach should be based on a “Through the Cycle” average. 51. ANZB has modelled the impact of the proposal Given the sensitivity of IRB models, ANZB’s risk weights with a 1.20 scalar would exceed 100% of the Reserve Bank standardised approach risk weights in an economic downturn. That is, ANZB’s risk weights would be more conservative than the standardised model in an extreme downturn.

Standardised Banks 52. The standardised approach to calculating capital requirements acts like a quasi￾leverage ratio, as the Reserve Bank has noted13. Applying a risk weight floor to IRB banks that is tied to a standardised outcome may constrain the ability of IRB banks to increase lending as risk reduces. 53.

13 ‘Capital review design options for an output floor for IRB banks’, P6, paragraph 17, Reserve Bank, October 2018.

17 ANZ BANK NEW ZEALAND LIMITED Tier 1 requirement 56. The Reserve Bank is proposing a 6% minimum Tier 1 ratio, with a 10% buffer, made up of a 7.5% CCB and 2.5% of DSIB/CCyB buffers. 57. ANZB’s view, together with other analysts, is that the proposed total capital buffers are too large:  The conservatism in the Reserve Bank’s current risk weightings means that New Zealand banks’ current capital ratios are materially understated against other jurisdictions’ capital ratios that have been referenced by the Reserve Bank in setting the proposed capital ratio minimums;  New Zealand banks’ current capital ratios are effectively already close to levels consistent with the Reserve Bank’s 1/200 risk appetite;  The Reserve Bank’s choice of a 1/200 risk appetite is, in any case, arbitrary;  A 16% Tier 1 minimum ratio and increased risk weights that lead to at least 50% increases in CET1 capital would lead to far larger impacts on the economy, concentrated in particular sectors, than the Reserve Bank expects. 58. PwC conducted a study in 201714 with the aim of comparing New Zealand bank capital ratios with capital ratios in other jurisdictions on an internationally comparable basis. 59. The study was updated in May 201915 to address feedback from the Reserve Bank, and to assess the implications of the Reserve Bank’s current capital proposals on internationally comparable capital ratios. 60. PwC’s revised conclusion on its original study is that the 2017 CET1 capital ratios of New Zealand’s four Australian-owned banks are 520 basis points higher at 15.5% when measured on an internationally comparable basis. 61. If the Reserve Bank proposals were implemented, the internationally comparable CET1 capital ratio would be 27.1%, an increase of 1,680 basis points.

14 ‘International comparability of the capital ratios of New Zealand’s major banks’, PwC, October 2017. 15 ‘International comparability of the capital ratios of New Zealand’s major banks – update paper’, PwC, May 2019.

18 ANZ BANK NEW ZEALAND LIMITED Source: ‘International comparability of the capital ratios of New Zealand’s major banks – update paper’, PwC, May 2019. 62. ANZB accepts that it is difficult to make comparisons with the capital requirements of other jurisdictions, however, the task is not impossible. The Reserve Bank’s assessment of an adjustment of 100-200 basis points is out of step with both PwC’s and APRA’s assessments. Impact of proposals on borrowing costs Impact on bank funding costs 63. The Reserve Bank refers to the Modigliani-Miller (MM) Theorem as a reason why banks’ overall funding costs should not increase by {(the increase in equitythe current cost of equity)-(the reduction in debtthe current cost of debt)}. The Reserve Bank assumes a 50% offset due to the MM effect. 64. The Reserve Bank therefore expects the overall increase in bank funding costs would be small enough to ‘come out in the wash’ and the consequential impact on customer pricing would not be material and could be easily offset, if necessary, by adjustments to the equilibrium OCR setting. MM effect 65. ANZB’s view is that the MM effect on ANZB would not be material. ANZB is a privately-owned subsidiary of an Australian listed company. (ANZB is the most

19 ANZ BANK NEW ZEALAND LIMITED significant New Zealand subsidiary among the Australian subsidiaries constituting around a quarter of ANZ’s earnings. Other Australian subsidiaries are ~10% of their overall groups.) 66. The impact, if any, on our parent-bank shareholders’ view of the riskiness of the parent bank following the proposed capital increases is likely to be minor. (This would be even more so for other Australian subsidiaries which are much less relevant for their parents’ shareholders.) 67. Therefore, we expect little impact on the cost of equity used for our business decision-making purposes. Debt funding costs 68.

 

71. Overall, we would expect no material MM effect offsetting the funding cost impact of the capital proposals. D-SIB banks
72. ANZB accepts that it qualifies as a domestic systemically important bank (D-SIB). We acknowledge that a D-SIB buffer would conform to the Basel framework.
73. We will respond to the question on identifying D-SIB banks in our submission on the separate D-SIB consultation. Counter-cyclical buffer
74. The Reserve Bank’s proposed deployment of the CCyB is inconsistent with the BCBS Basel III requirements16 .
75. ANZB submits that the CCyB should be set to zero in normal conditions and only deployed by the Reserve Bank “when excess aggregate credit growth is judged to be associated with a build-up of system-wide risk to ensure the banking system

16 ‘Basel III: A global regulatory framework for more resilient banks and banking systems’, BCBS, December 2010 (rev June 2011), paragraphs 136 to 150. Refer also to: ‘Guidance for national authorities operating the countercyclical capital buffer’, BCBS, December 2010; ‘Frequently asked questions on the Basel III Countercyclical Capital Buffer’, BCBS October 2015 and ‘Implementation – Range of practices in implementing the countercyclical capital buffer policy’, BCBS, June 2017.

20 ANZ BANK NEW ZEALAND LIMITED has a buffer of capital to protect against future potential losses” in accordance with the BCBS and international best practice. 76. The BCBS further notes that jurisdictions are likely to only need to deploy the buffer on an infrequent basis, rather than the Reserve Bank’s proposed “normal conditions”. Should there be an option to set the CCyB to zero in the exceptional circumstances following a crisis? 77. Yes. Is 1.5 percent an appropriate calibration for the ‘early set’ component of the CCyB? 78. Refer to response provided in paragraphs 74-76. Tier 2 capital 79. Tier 2 capital provides loss absorption when a bank has failed, protecting senior creditors, including depositors, from loss. 80. The advantage from banks’ viewpoint is that it is lower cost than Tier 1 capital. 81. The disadvantage for Australian subsidiaries of the Reserve Bank’s proposed Tier 2 instruments (including redeemable preference shares) is that they do not conform to APRA’s requirements. Therefore, any Tier 2 debt issued at the New Zealand level would not feed ‘upstream’ to the parents’ capital ratios and would require the parent to issue an equal amount of capital itself. To avoid doubling the cost of this amount of capital, Australian subsidiaries would probably need to issue any Tier 2 debt to the parent. 82. ANZB reiterates its submissions on the Reserve Bank’s ‘Capital Review Paper 2: What should qualify as bank capital? Issues and Options’ (July 2017) with regard to both AT1 and Tier 2 capital criteria including:  AT1 and Tier 2 capital criteria should align with the international standards and APRA criteria to enable capital to be raised externally and be recognised by both the Reserve Bank for the New Zealand subsidiaries and APRA for the Australian parents; and  AT1 instruments should be able to be redeemed as redemption features do not diminish the capital value of the AT1 instrument. 83. If the Reserve Bank’s proposals proceed, we see no need for a Tier 2 capital requirement, as the chances of a bank failure would be insignificant. 84. If the Reserve Bank reduces the proposed increases in CET1 capital as a result of the consultation process, ANZB’s view is that Tier 2 capital may be useful as a ‘backstop’ to Tier 1 capital holdings. Escalating Supervisory Response 85. ANZB expects that, in a situation where capital ratios are declining rapidly due to deteriorating credit conditions, net dividends would be voluntarily constrained. 86. ANZB would also expect closer and more active supervisory action by the Reserve Bank in this scenario.

21 ANZ BANK NEW ZEALAND LIMITED Leverage Ratio 87. ANZB agrees with the Reserve Bank that the extra value of a leverage ratio for New Zealand banks is not material. By far the major risk in New Zealand bank balance sheets is credit risk in the loan portfolio. 88. Banks in other jurisdictions (e.g. the US) hold much higher proportions of their balance sheets in highly rated securities with associated low risk weightings. In that case, a leverage ratio may provide another useful lens on overall risk. 89. In ANZB’s case (and, we assume, similarly for other banks in New Zealand), highly rated securities are held almost entirely to meet regulatory liquidity ratios and, in ANZB’s case, amount to around 13% of total assets. 90. We have not found that the lack of a reported leverage ratio negatively affects offshore investor perceptions of ANZB. 91. We do not think that minimum leverage ratio requirements should be part of the capital framework or disclosure requirements. Transition Period 92. Completing this over a five year period is excessively ambitious in ANZB’s view. It would involve material changes to balance sheet settings over a relatively short time frame and assumes that the current benign credit conditions would continue. 93.

95. In ANZB’s view there would be no material harm, and it would provide flexibility to issuers, to allow issuers’ outstanding AT1/T2 instruments to have full capital treatment until their first call date, at which time capital treatment would fall to zero.

22 ANZ BANK NEW ZEALAND LIMITED Appendix 1: Examination of the analysis underlying the Reserve Bank’s proposed changes to capital adequacy Appendix 1 examines the Reserve Bank’s analysis underlying its proposed changes to capital adequacy.17 This should be read alongside Appendix 2, which models the impacts of the Reserve Bank’s capital proposals, including the magnitude of the capital raise, effects on spreads and lending volumes, and the ultimate economic impacts. Key points:  Absence of cost-benefit analysis The standard international approach of cost-benefit analysis to determine optimal bank capital ratios would be a more appropriate framework for analysis than the Reserve Bank’s idiosyncratic ‘risk appetite’ approach. The latter uses a number of unknowable assumptions to generate a capital ratio ostensibly delivering “soundness” and then asymmetrically asks only if this number could be higher without generating unacceptable expected output costs (“efficiency”). It does not also ask whether it should be lower, i.e. whether the anticipated costs of the proposed ratio may exceed the expected benefits.  Risk appetite not well justified The assumed societal risk appetite underlying the Reserve Bank’s proposals (1 financial crisis on average every 200 years, i.e. a 0.5% risk of a financial crisis in any given year) is not well justified. But taking it as a given, for an increase in capital to be required under this framework, it must be the case that a financial crisis is likely to occur more than once every 200 years at current capital ratios – indeed, significantly more often, given the scale of the increase proposed. This is inconsistent with rating agency assessments, the international literature (once cross-country differences are appropriately adjusted for), modelling by IRB (big 4) banks, the Reserve Bank’s own stress tests, and New Zealand’s historical experience.  Implausible assumptions in portfolio modelling After appropriately adjusting for New Zealand’s relatively conservative regulatory regime, the international studies quoted by the Reserve Bank do not support the assertion that the proposed level of capital is required to bring the risk down to 0.5% per year. Implausible assumptions have to be made in order for portfolio modelling to generate this result in a New Zealand context.  Cost assessment is too low The Reserve Bank’s assessment of the costs of the proposals appears too low. Both international studies and our own analysis suggest long-term impacts on lending spreads and GDP that, while uncertain, are considerably larger than the Reserve Bank assumes. Additionally, the Reserve Bank does not take into account likely impacts on the availability and allocation of credit, which may have significant long-term costs.

17 Proposed changes are outlined in the Reserve Bank’s consultation paper, ‘Capital Review Paper 4: How much capital is enough?’, December 2018; updated January 2019.

23 ANZ BANK NEW ZEALAND LIMITED  Transitional costs are not considered Transitional costs of the proposals are significant but are not considered. International literature finds significant impacts on both credit supply and lending spreads during a transition to higher capital ratios, consistent with our modelling. It also supports our view that there would be a structural adjustment in the allocation of capital away from business and farm lending in favour of housing, with negative implications for New Zealand’s long-run productive potential.  Support for the proposals appears weak Overall, both analytical and empirical support for the Reserve Bank’s proposed level of capital ratios appears weak. The international literature supports capital ratios close to where they are now, taking into account cross-country differences. And plausible assumptions from the international literature suggest the costs of the proposals are likely to outweigh the benefits.

1. The risk appetite framework lacks a cost-benefit analysis The basis of the Reserve Bank’s proposals is an idiosyncratic “risk appetite” framework. They argue that this is consistent with the literature in the sense that people are generally considered to be risk averse. However, it is not consistent with the literature regarding optimal bank capital ratios, which uses cost-benefit analysis. Under Reserve Bank’s approach, the chosen regulatory minimum for bank capital is decided based on an assumed desire to avoid a 1-in-200-year financial crisis (i.e. reduce the annual probability of crisis to less than 0.5%). Then, given a capital ratio that “at least” delivers that, as a second stage, the question is asked whether the ratio could be made higher still without incurring “material” expected losses to output. The Reserve Bank’s process does not ask whether the expected benefits of the occasional averted financial crisis outweigh the ongoing costs associated with the proposed capital ratio, and thus whether it might be sub-optimally high. The Reserve Bank has indicated that they will not do a full cost-benefit analysis until after submissions have been received. The key shortcoming of the risk appetite approach is that while it acknowledges that people are indeed risk averse, it does not acknowledge the fact that risk-averse individuals still have regard for costs – they may prefer a different point on the trade-off curve versus a more risk-loving individual, but costs still matter to them. Cost-benefit analysis that incorporates risk aversion, as seen in the literature, would be a more appropriate way to assess optimal capital ratios, taking this into account and better acknowledging the preferences of risk-averse individuals. The Reserve Bank suggests that cost-benefit analysis is too difficult for non-specialist audiences to understand and assert that maximising economic outcomes is not necessary.18 But we would argue that non-specialist audiences understand costs and benefits, and that incomplete information makes responding to the proposals even more difficult for these individuals. For the public to be able to provide useful feedback on their preferences, there needs to be something meaningful to choose between. A clear understanding of the costs of the proposals is required, along with information about

18 ‘Risk appetite framework used to set capital requirements’, Reserve Bank Banking Steering Group internal note, 30 October 2018, and ‘An outline of the analysis supporting the risk appetite framework’, Reserve Bank Capital Review Background Paper, April 2019.

24 ANZ BANK NEW ZEALAND LIMITED how the Reserve Bank is assuming that society will weigh these against potential benefits. Rather than being a more “transparent” approach, as the Reserve Bank suggests, the risk appetite framework omits crucial information. Given the Reserve Bank’s chosen approach, the choice of risk appetite for the capital review is necessarily subjective, but in theory should be representative of the public’s risk tolerance – since society will ultimately bear the costs and benefits of the changes. There is no evidence the Reserve Bank has attempted to gauge how often the public would prefer financial crises to occur, for obvious reasons (“never, thanks”), but the Reserve Bank provides no justification as to why the particular 0.5% risk tolerance they have chosen is the right one. The Reserve Bank assert that they have been delegated the authority to make this assessment on behalf of the public.19 Yet despite being at the very heart of their analysis, the Reserve Bank’s assumed appropriate level of risk tolerance was decided late in the process and became more stringent with what appears to be little justification. The Reserve Bank states that it assessed a 1-in-100-year risk tolerance to be “inefficient”, with reference to an internal note. But the internal note that accompanied that assessment was in fact supportive of a 1-in-100-year risk tolerance (consistent with capital ratios close to current levels) and provided no analysis supportive of the claim that this risk tolerance might be considered inefficient.20 A more recent paper, released in April 2019, attempted to retrospectively show that this level would be inefficient. However, this analysis relied on implausibly high parameters for key inputs in order to generate this result.21

Perhaps most importantly, there is little evidence to suggest that the Reserve Bank’s assumed societal risk tolerance is not already satisfied at current capital ratios. That is, for an increase in capital to be required, it follows that a financial crisis must be likely to occur more than once every 200 years at current capital ratios. This is inconsistent with rating agency assessments, the international literature (appropriately adjusting cross￾country comparisons), modelling by IRB (big 4) banks, the Reserve Bank’s own stress tests, and New Zealand’s historical experience.22 Of course no risk assessment is infallible, and history may not be a good guide to the future. But the Reserve Bank does not provide any new evidence that the risks are greater than the weight of existing evidence suggests. Likewise, there is little acknowledgement that financial crises are not random events – and that financial policy has evolved over time to reduce the chance of one occurring. While the catalyst may be hard to predict, the underlying cause of bank failures can be sustained poor lending decisions in the years preceding, which an effective bank regulator might hope to be able to mitigate. Alternatively, broader economic developments, regulation or incentives affecting private decisions may be the underlying cause, which higher levels of bank capital do not address.

19 ‘An outline of the analysis supporting the risk appetite framework’, Reserve Bank Capital Review Background Paper, April 2019. 20 In November, a paper was presented to the Reserve Bank’s internal policy committee proposing capping the probability of crisis at 1% or less, consistent with a 1-in-100-year risk tolerance. See ‘Risk appetite framework used to set capital requirements’, Reserve Bank Banking Steering Group decision paper, 30 October 2018. The Reserve Bank claimed that this paper informed an assessment that a 1-in￾100-year risk tolerance would be “inefficient” (see ‘An outline of the analysis supporting the risk appetite framework’, Reserve Bank Capital Review Background Paper, April 2019). However, while the alternative 1-in-200-year risk tolerance was mentioned in a footnote, there was no accompanying analysis to support an efficiency assessment or indicate that one was undertaken, and no justification for why the proposed 1-in-100-year proposal was abandoned. 21 In particular, underlying assumptions for both average system-wide probability of default and loss given default used to generate this result are implausibly high, as discussed in the following section. 22 The four major banks are rated AA-, consistent with a bank failure occurring 1 in every 1,250 years.

25 ANZ BANK NEW ZEALAND LIMITED Along these lines and others, much has been done in New Zealand since the GFC to guard against crisis events occurring: previous increases in bank capital (banks’ Tier-1 capital has increased from ~8% in 2007 to ~13% currently), the introduction of a macro-prudential policy toolkit (LVR restrictions), increased use of stress testing, strengthening of the bank resolution regime, and changes to liquidity policy (such as the introduction of the core funding ratio) to insulate banks to some degree from stresses in global financial markets. There is also little acknowledgement that the New Zealand banking system is simpler than in other jurisdictions, which means the risks associated with complex and opaque systems, complicated financial instruments and the like are less relevant here. As mentioned, the Reserve Bank discounted their own stress test findings when assessing the current soundness of the New Zealand banks. It is true that it is very difficult to predict both the triggers and effects that lead to crisis events, and the complex interplay of various financial and real variables during such a period. This makes stress testing fallible and by necessity oversimplified. But nonetheless, the tests do suggest that current levels of capital would allow the banks to withstand very significant stress events that history tells us would be very rare indeed. Stress tests have consistently shown that New Zealand banks would weather an extreme downturn well (i.e. the most severe downturn modelled), with current capital ratios. One internal Reserve Bank paper presented to the policy committee in August 2017 (based on more severe assumptions for losses than those in published stress tests) advised that a capital ratio of 12-13½% CET1 was appropriate – close to or a touch above where they are now, but much less than the Reserve Bank is now proposing.23

2. International comparisons and modelling International studies support NZ capital ratios, appropriately adjusted, at around current levels An important contextual issue underpinning all of the Reserve Bank’s analysis is a lack of comparability between capital ratios across jurisdictions. The Reserve Bank does not take this into account in their presentation of results from international studies, but all results need to be adjusted to reflect these differences. This has significant bearing on the findings. The New Zealand regulatory approach is conservative relative to international peers, due to both definitional and risk-weighting differences. A 2016 Reserve Bank paper suggested that New Zealand banks capital ratios might be 1-2%pts higher when measured on an internationally comparable basis.24 In 2017, PwC was commissioned to do a report looking at the comparability of New Zealand bank capital levels internationally.25 That report found that if New Zealand banks’ capital ratios were restated using international measurement conventions, they would be 6%pts higher and were already relatively high by international standards. The Reserve Bank discounted

23 ‘Implications of stress tests for calibration of capital requirements’, Reserve Bank MFC information paper, 28 August 2017. 24 ‘Capital review: current standards and international comparisons’, Reserve Bank FSO information paper, 7 September 2016. 25 ‘International comparability of the capital ratios of New Zealand’s major banks’, PwC, October 2017.

26 ANZ BANK NEW ZEALAND LIMITED this analysis, citing a number of sources of potential upward bias.26 A second PwC report27 has revised its previous findings in light of the Reserve Bank comments to an increase of 5.2%pts. However, rather than making adjustments to the PwC estimates to reflect their concerns or producing their own estimates (as APRA did for Australia, finding a 3%pt adjustment is appropriate there), the Reserve Bank disregarded the PwC report and used a Standard & Poor’s cross-country comparison instead. The Standard & Poor’s estimates were used in the Deputy Governor’s recent speech to argue that capital ratios in New Zealand are not particularly high relative to other countries.28

Reflecting possible upward bias in the original PwC estimate above, Tailrisk Economics suggest an adjustment of 4%pts might be appropriate.31 This suggests that 22% is the appropriate benchmark for comparing to the Reserve Bank’s proposal of a 16% Tier-1 (going concern) capital minimum, plus a buffer of 2%pts.32

We use a 1-3%pt adjustment (similar to the Reserve Bank’s own previous work, and APRA’s findings). There is also uncertainty about the buffer banks might be willing to hold over the regulatory minimum. Banks might be willing to hold a smaller buffer, of say 1%pt. The Reserve Bank assumes a buffer of around 0.8%pts (0.5%pts unweighted).33

Based on uncertainty about both the bias and likely buffer, we take a conservative approach for international comparisons, using a 2-4%pt adjustment. This implies a comparable range of 18-20% for the Reserve Bank’s proposal of a 16% Tier 1 capital minimum. In places where studies refer to CET1, we compare to a range of 16½-18½%, corresponding to a 14½% CET1 ratio in New Zealand, assuming the new proposals are met through common equity. Similarly, the current level of Tier-1 capital in New Zealand

26 ‘International comparison of capital ratios (PwC report)’, Reserve Bank FSO information paper, 6 December 2017 & ‘2017PwC (NZ) study’, Reserve Bank FSO information paper, 8 May 2018. 27 ‘International comparability of the capital ratios of New Zealand’s major banks – update paper,’ PwC, May 2019. 28 ‘Safer banks for greater wellbeing’, Reserve Bank speech by Deputy Governor Geoff Bascand, 26 February 2019.

31 ‘The 30 billion dollar whim’, Tailrisk Economics review of the Reserve Bank consultation paper: ‘How much capital is enough’, March 2019. 32 We assume a smaller buffer than the banks currently hold (3%pt), since the new capital regime would be conservative and disciplinary action for minor breaches is not expected to be stringent. Nonetheless, we expect that banks would wish to maintain some kind of buffer to avoid adverse reactions from rating agencies and market as a consequence of potential breaches. 33 Where applicable, we have converted the Reserve Bank analysis of capital ratios using unweighted assets to be in terms of risk-weighted assets for comparability.

27 ANZ BANK NEW ZEALAND LIMITED of 11½% taking into account the Reserve Bank’s proposed changes to risk weights, or 13% without, is broadly equivalent to an adjusted 14-16% ratio when comparing to international studies. Once appropriately adjusted, international evidence suggests that a minimum Tier-1 capital requirement of 16% is too stringent relative to the Reserve Bank’s intended objective of achieving an annual probability of crisis of 0.5%. The Reserve Bank leans on three key international studies throughout their analysis in assessing the relationship between the level of bank capital and the probability of a financial crisis:  A 2017 Federal Reserve paper found that a 14-17% capital ratio was associated with a probability of crisis of 0.7-1.0%.34 However, they assume a shorter sample for their time series analysis than the other papers, eliminating the relatively benign period from 1970-1989. Their estimates using another method found a 17% capital ratio was associated with a probability of crisis of 0.5%, suggesting a slightly higher capital ratio than currently imposed might be consistent with the Reserve Bank’s assumed risk tolerance. But it also suggests the Reserve Bank’s proposed increase is far too stringent. In general, this paper is US-centric and less relevant to New Zealand. The paper notes that a range of conservative assumptions are made because financial crises are more common in the US than in other jurisdictions.  A 2010 BCBS study (adjusted to be on a Tier-1 basis by the Bank of England) suggests that the current level of capital in New Zealand (of 14-16%, adjusted) is associated with a probability of crisis of 0.8-1.3%, higher than the Reserve Bank’s 0.5% target.35 However, based on the underlying studies, the Reserve Bank’s proposed minimum is too stringent and might be accompanied by a probability of crisis of 0.2% or less, consistent with avoiding a 1-in-500-year event. Overall, this paper suggests that an adjusted 17% capital ratio might be desirable based on the Reserve Bank’s risk tolerance – a little higher than current regulations, but below the Reserve Bank’s proposals.  In 2015, the Bank of England conducted a study that was superior in that it took into account changes that have reduced the probability of crisis post-GFC.36 This paper found that a 14-16% international capital ratio was associated with a probability of crisis of 0.3-0.5%. This suggests that the current level of capital is consistent with the Reserve Bank’s stated risk tolerance. An adjusted ratio of 18-20% comparable to the Reserve Bank’s proposals would be accompanied by a probability of crisis of less than 0.3%. In short, taking cross-country differences properly into account, none of these studies supports the notion that capital ratios need to be as high as the Reserve Bank is proposing. Indeed, the Bank of England study, which appears to be the most relevant to New Zealand, suggests that the current level of bank capital is about right. Leaving aside

34 ‘An empirical economic assessment of the costs and benefits of bank capital in the US’, Federal Reserve Board discussion paper, March 2017. 35 ‘An assessment of the long-term economic impact of stronger capital and liquidity requirements’, BCBS report, August 2010. 36 ‘Measuring the macroeconomic costs and benefits of higher UK bank capital requirements‘, Bank of England Financial Stability Paper No. 35, December 2015.

28 ANZ BANK NEW ZEALAND LIMITED the fact that results from the former two studies appear less relevant to New Zealand, simply aggregating results from these studies suggests that New Zealand capital levels of around 10-15% would be consistent with the Reserve Bank’s risk tolerance – close to where they are now. This assessment is based on the Reserve Bank’s risk tolerance framework, which assesses soundness without taking into account costs. However, the papers used by the Reserve Bank to justify this assessment are actually cost-benefit studies. As discussed in the final section, Tier-1 capital ratios of 11-13% are justified in New Zealand based on cost-benefit analysis – again, around where they are now. Unusual assumptions are required to justify 16% Tier 1 capital ratio Even if a 1-in-200-year risk tolerance were the appropriate level, and evidence of risk exceeding this at current capital levels could be provided, the Reserve Bank still has to make very unusual, conservative assumptions in their portfolio modelling for a 16% capital ratio to be justified in the New Zealand context. And taking into account that we expect banks will look to hold a buffer above the minimum, implied levels of capital under the proposals look even less justifiable. The Reserve Bank used two pieces of internal analysis to justify this level. These use portfolio modelling of New Zealand bank risk, as an acknowledgement that the New Zealand context matters for determining the appropriate level of capital here.37 The first of these was presented to the internal policy committee in November 2018; the second was produced a month after the Reserve Bank’s proposed changes had already been announced. More recently, the Reserve Bank has released a background paper, which provides a third version of this analysis.38

The parameters that were used changed between all three of these analyses and there are issues with each. The implausible parameter assumptions below result in estimates of the probability of crisis that are improbably high.  The first piece of work assumed a long-run average system-wide probability of default of 2.8% with a risk tolerance calibrated to avoid a 1-in-200-year event. This probability of default was not justified, and is inconsistent with historical data on non-performing loans (which historically have averaged 1.5%) and the calibration used in IRB bank model outcomes (1.1-1.4%). The 2.8% value is consistent with default rates from stress tests, but these are based on stress periods by definition, and are therefore not suitable estimates for the long-run average. Similarly, a loss given default of 40% was used, which appears too high based on both stress tests and banks’ own modelling (29-37%).  The second piece used more plausible estimates of the probability of default (1.5%) and loss given default (35%). It is clear from the explanatory evidence in this note that the previous assumptions were implausibly high. However, in order to generate a 16% result consistent with the previous paper, a risk tolerance calibrated to avoid a 1-in-333-year event was used –

37 ‘Capital ratio calibration’ Reserve Bank FSO decision paper, 13 November 2018, and ‘Explanatory note on portfolio risk modelling in the New Zealand context’, Reserve Bank note, 25 January 2019. 38 ‘An outline of the analysis supporting the risk appetite framework’, Reserve Bank Capital Review Background Paper, April 2019.

29 ANZ BANK NEW ZEALAND LIMITED much more stringent than the Reserve Bank’s stated objective. Justification for this was the notion that the distribution of rare events is uncertain and not necessarily “normally distributed”. While that might be true, the assertion does nothing to support such a large reduction in the modelled risk tolerance versus the stated objective.  The third paper used a risk tolerance calibrated to avoid a 1-in-200-year event, consistent with the Reserve Bank’s stated objective. However, their baseline assumptions for system-wide probability of default (2.25%) and loss given default (40%) were again implausibly high. This is despite it being acknowledged in the second piece of work that lower estimates for these parameters would be more appropriate. Not only does the Reserve Bank use these implausible parameters to justify a 16% capital ratio from a soundness perspective, it also uses them to argue that a 16% capital ratio is “efficient” (i.e. generates no material extra output costs) relative to a capital ratio of 14.5%. This assertion requires even more implausible assumptions, in addition to those described above, including an estimate for the cost of crisis that looks too high and very low estimates of economic costs of higher capital ratios (both discussed later). 3. Reserve Bank cost estimates are too low Based on the Deputy Governor’s recent speech, the Reserve Bank expects the impact of the proposals on credit spreads to be in the range of 20-40bps, split between lending and deposit margins. We estimate the impact will be larger than this, and seen primarily in higher lending spreads.

We produce larger estimates than the Reserve Bank for the impact on credit spreads for several reasons:  The Reserve Bank’s estimates imply that that return on equity would fall to levels that we consider would likely not be acceptable to shareholders, given the (unchanged) cost of capital and dilution of returns for existing shareholders. This reflects the unique ownership structure of the New Zealand banking system.  The Reserve Bank expects that increased competition will encourage banks’ required return on equity to fall. However, we think that meaningful changes in the structure of the financial system could take a very long time indeed. − Smaller banks would also need to raise capital in response to the proposals; − Non-banks have shown no sign of resurgence to fill other lending gaps (e.g. in response to macro-prudential policy);

30 ANZ BANK NEW ZEALAND LIMITED − Foreign banks may be put off by how much capital is required to operate at scale here as locally-incorporated banks (given opportunity costs) and may find cost-efficient opportunities in our relatively small economy limited when operating as a branch; and − It would take much more structural change for corporates to move at a meaningful scale to alternative funding sources, given the shallow nature of capital markets in New Zealand.  The Reserve Bank assumes that debt funding costs will fall in response to the changes. However, New Zealand banks can fund relatively cheaper in international wholesale markets versus their standalone credit rating, reflecting the implicit backstop of the Australian parents. The standalone credit rating would be relevant only in unprecedented circumstances (parent banks getting in trouble independent of the New Zealand subsidiaries) that historical borrowing costs suggest don’t appear to play a big part in lenders’ thinking. Given this, we expect only a small impact on debt costs as a result of higher capital ratios. It seems likely that bank shareholders would indeed bear some of the cost of the changes in the form of a lower return on equity, but in order to maintain returns at acceptable levels, lending spreads would need to increase more than the Reserve Bank anticipates. Alternatively, a smaller impact on credit spreads is possible, but this could be accompanied by a reduction in credit availability if the returns on equity on the New Zealand businesses are not attractive relative to alternative uses of the capital. The Reserve Bank makes a conservative assumption for the impact on deposit margins, though they highlight that some increase in margins could come via this channel, rather than lending spreads. We think deposit margins are likely to be little affected because banks would probably still need to maintain competitive margins to attract deposits (for assets to grow). While we do not expect to see much of an impact on deposit spreads, depositors would nonetheless bear some of the cost, since the OCR would need to be lower than would otherwise be the case. Any fall in deposit rates adds uncertainty to the economic impacts of the changes. Generally, loan and deposit margins move in the same direction, and that will tend to be captured in economic modelling that is based on historical data. But in this case, deposit margins would be flat or falling while lending spreads would be rising. This means that income from deposits would be lower than would usually be seen when lending spreads rise – potentially increasing negative economic impacts. On the other hand, lower deposit margins could encourage less saving, helping to hold up consumption. The net impact of these effects is uncertain. However, swamping all this, the fact that the OCR would need to be lower would inevitably lead to a chunk of the burden from higher capital ratios falling on savers, many of whom tend to be older. If long-term deposit rates fell from 4% to 3.5% as a result of the OCR being persistently 50bp lower, then this would result in a person with a $100,000 nest egg earning $20,000 less over the course of a 20-year retirement. For households and businesses, higher lending spreads directly affect borrowing costs and cash flow – and we think the magnitude of this impact is not well understood.

31 ANZ BANK NEW ZEALAND LIMITED Tailrisk Economics39 estimates that borrowing costs might increase by $1.7bn per year, based on the Reserve Bank’s estimated impact on lending spreads of 40bps and a capital raise of 5%pts. In net present value terms, this is a total cost of $300bn, assuming a discount rate of 5%. For individual borrowers, this means that a business with a loan of $5m could pay $50,000 extra per year, assuming an increase in lending spreads of 100bps. Likewise, a household with a $400,000 mortgage could pay an extra $1,000 per year, assuming an increase in mortgage spreads of 25bps. A 50bp lower OCR would help to offset these effects for mortgage borrowers, but would not be sufficient to offset the impact for farms and businesses. We believe higher lending spreads would be experienced disproportionally by businesses and farms (as discussed both in our modelling note and later), though the extent of this differentiation and exact impact on sectoral margins is hard to predict. However, as we estimate in Appendix 2 the capital raise could be closer to 6%pts (assuming a buffer of 2%) than the 5%pts the Reserve Bank estimates. This implies that the total costs and impacts on individual borrowers could be larger than those estimated above. Higher borrowing costs have flow-on effects to GDP. In their consultation paper, the Reserve Bank said they expected a 1%pt increase in capital ratios might reduce the level of trend GDP by 0.03%. Based on our calculation of a 6%pt capital raise, this implies a ~0.2% permanent impact on GDP from the Reserve Bank’s proposed change. More recently, the Reserve Bank has suggested an impact of 0.3% of GDP (with 0.6% as an upper bound).40

The literature suggests the latter estimate (0.3-0.6%) is more realistic.  The 2010 BCBS study finds permanent effects of 0.09% per 1%pt capital increase. Based on these estimates, the Reserve Bank’s proposals would amount to the level of GDP being lower by 0.6%;  The 2017 Federal Reserve paper finds permanent effects of 0.04-0.07% per 1% capital increase. Based on these estimates, the Reserve Bank’s proposals would amount to the level of GDP being lower by 0.3-0.4%;  The 2015 Bank of England paper finds permanent effects of 0.01-0.05% per 1%pt capital increase. Based on these estimates, the Reserve Bank’s proposals would amount to the level of GDP being lower by 0.1-0.3%. While this study points to a GDP impact consistent with the Reserve Bank’s (original) cited estimate, it argues for optimal capital ratios that are well below the Reserve Bank’s proposals. A simple average of these results suggests a permanent effect on GDP of around 0.4% (with a range of 0.1-0.6%). However, we find that the permanent effect of the proposed changes is likely to in fact be closer to 1% or larger still. Our modelling of the economic impacts of the proposed capital changes in Appendix 2 focuses on the next ten years, including impacts on the cost, availability and allocation of

39 ‘The 30 billion dollar whim: A review of the Reserve Bank consultation paper: ‘How much capital is enough’’, Tailrisk Economics, March 2019. 40 ‘An outline of the analysis supporting the risk appetite framework’, Reserve Bank Capital Review Background Paper, April 2019.

32 ANZ BANK NEW ZEALAND LIMITED credit, and the structural adjustment that the transition to higher capital ratios would entail. We find an impact on GDP of 1-3% after 10 years. The degree to which this would be expected to result in a permanent impact on GDP is subject to a lot of uncertainty; it is possible that impacts dissipate as the financial system adjusts, as it inevitably will. But based on our modelling, we would expect a permanent impact on GDP of 0.4-2.0%.41

In light of these issues, it appears the Reserve Bank has chosen an estimate of the long￾term GDP impact that is at the lower end of what might be plausible in practice – even assuming their worst-case scenario of 0.6% of GDP. While 1% of GDP might sound like a small permanent impact, this implies that the level of GDP – the net value of total production in the economy every year – is lower in perpetuity. In present value terms (i.e. taking into account the fact that we care more about today than tomorrow), this is equivalent to a total impact of at least 20% of GDP (assuming a discount rate of 5%). By contrast, the Reserve Bank’s estimates of 0.2- 0.6% would imply a total net present value impact of 4-12% of GDP. We believe this is too low, taking into account the studies above and the effects that those studies do not consider. 4. Transitional impacts and structural changes in the economy None of the long-term studies discussed above is designed to consider the transitional impacts of capital changes. The papers above estimate permanent impacts on GDP in

41 These estimates assume that at least half of the impact on lending spreads is permanent, at least a third of portfolio rebalancing effects (away from business and agriculture) are permanent, and up to half of any disintermediation effects are permanent.

33 ANZ BANK NEW ZEALAND LIMITED order to weigh up the very long-term costs and benefits of higher capital requirements. However, transitional impacts on the economy are also important. The Reserve Bank does not discuss these issues. We have subsequently found estimates in the literature consistent with this estimate. A 2010 BCBS paper released alongside the one cited above aggregated estimated transitional impacts from a range of models across a range of different countries.43 This study only considers the impact over 4.5 years (not 10 as in our work) and finds significant impacts on lending spreads in the transition period. A 1%pt increase in capital ratios is found to increase lending spreads by 15-17bps. This implies that the Reserve Bank’s proposals would be consistent with a 90-100bp increase in aggregate lending spreads in the transition. As a result of higher lending spreads, the study finds that a 1%pt increase in capital ratios reduces GDP by 0.16% in the transition. This would amount to the level of GDP being 1% lower in response to the Reserve Bank’s proposals. In addition to this, large impacts on lending volumes were also found. For example, in the UK, half of the move to higher capital ratios was met through a reduction in risk￾weighted assets. Based on the studies that captured these effects, a 1%pt increase in capital ratios was estimated to reduce credit volumes by 1.4-1.9% as a result of credit rationing. The Reserve Bank’s proposals would therefore correspond to an 8-12% fall in lending volumes (in line with our modelling). Taking into account impacts on both lending spreads and volumes, a 1%pt increase in capital ratios was found to reduce GDP by 0.32%. This would amount to the level of GDP being 2% lower in total in response to the Reserve Bank’s changes. This impact is consistent with our modelling. Over the same time period, we find an impact of close to 1% without a pull-back in lending volumes. But if credit supply is affected, GDP is lower by 2%. These additional impacts could be particularly persistent if the Australian parent banks permanently retrench. The BCBS paper also highlighted the importance of possible portfolio rebalancing effects. They note that banks are likely to respond to higher capital requirements by changing the composition of their assets to reduce credit risks, reducing “riskier” lending in favour of borrowers that are considered more creditworthy. In response to the Reserve Bank’s proposals, banks would be incentivised to steer lending towards categories with lower risk weights – away from commercial and agricultural lending and towards residential mortgages. Transitional impacts can affect the long-term structure of the economy in ways that may not be desirable. Our estimates point to a significant impact on business investment and farm transactions in response to the changes in capital requirements, as banks’ portfolios are rebalanced towards lending at lower risk weights. This is not considered in the Reserve Bank analysis, but it could have significant effects on the structure and long-term productive potential of the economy. We find that business investment and farm activity are structurally lower, in favour of housing activity.

43 ‘Assessing the macroeconomic impact of the transition to stronger capital and liquidity requirements’, BCBS report, August 2010.

34 ANZ BANK NEW ZEALAND LIMITED There has already been a significant repricing in loans and reduction in credit availability for these segments following the GFC, which appears to have contributed to weak business investment. Further moves in this direction could affect the long-term productive potential of the economy if there were permanent impacts on the capital stock and productivity. All of our modelling assumes a benign economic environment, but the economic context during the transition is also relevant. We think the following issues need to be taken into account when considering the impacts of the proposals, particularly when deciding the length of the transition and the possibility of making the proposals conditional on the economic environment.  We are already late in the economic cycle and a recession over the period of transition to higher capital ratios is a very real possibility. Higher costs and reduced availability of credit could significantly worsen the economic impacts in such an event. In particular, the economic impacts of the proposed changes could be greater if accompanied by increased risk aversion by investors and/or higher bank funding costs. Should economic conditions deteriorate, it is imperative that credit is freely available to cushion a downturn and stimulate the economy.  We expect that the neutral OCR would be lowered by the capital changes over a meaningful economic horizon. The lower the level of the neutral OCR, the greater the risk that monetary policy runs out of conventional ammunition and unconventional measures need to be employed, given the OCR starting point of just 1.50%. The costs of such policies could be significant in this context. 5. Cost-benefit analysis Given all of the issues outlined above, analytical support for the Reserve Bank’s chosen level of bank capital appears weak. The Reserve Bank considered “efficiency’ (including the safety/output trade-off) only in the second stage of their assessment as a possible justification to move ratios higher, rather than using cost-benefit analysis to consider whether their chosen ratio is optimal. Nonetheless, the key features of the cost-benefit analyses that make up the literature on optimal capital are discussed. For context, previous cost-benefit analysis by the Reserve Bank found an optimal Tier-1 capital ratio (including a buffer) of 13% (i.e. close to current levels).44 A similar conclusion comes out of assuming more standard parameters than the very conservative assumptions the Reserve Bank used in their portfolio modelling (as discussed above), or alternatively with a risk tolerance to avoid a 1-in-100-year event, as the Reserve Bank initially envisaged. International cost-benefit studies give wide ranges for the optimal levels of bank capital. A key uncertainty that underpins the literature is the cost of financial crises. The Reserve

44 This was outlined in the Reserve Bank’s 2012 Regulatory Impact Statement on the application of Basel III capital requirements in New Zealand.

35 ANZ BANK NEW ZEALAND LIMITED Bank acknowledges that cost estimates are sensitive to the discount rate chosen, and that costs would be lower if they used a discount rate similar to that employed in New Zealand Treasury analysis. Moreover, estimates of the cost of crises in the literature are estimated in a manner that assumes costs are invariant to the level of prevailing capital ratio. However, higher capital ratios reduce both the probability and costs of crises and given that capital ratios in New Zealand already look quite high, the costs of a crisis here is likely to be smaller than international estimates from other jurisdictions would suggest. There is also considerable uncertainty about the degree to which the scars of financial crises are permanent. Given the significance of such events, persistent effects certainly seem reasonable, in light of impacts on the capital stock, long-term unemployment and so on. But it is not clear why the economy should never fully recover and potential growth should be permanently affected, especially since any associated credit crunch would be expected to subside. The Reserve Bank cites the cost of financial crises at 63% of GDP, based on the BCBS paper mentioned above that focused on long-term impacts. This estimate appears too large. If effects are not assumed to be permanent, then the cost is estimated to be significantly lower (19%). The Bank of England paper discussed above takes into account post-crisis reforms, including resolution regimes (especially on the fiscal position and loan spreads) and bail-in of creditors (which reduces contagion) and finds a cost of financial crises of 43% of GDP (assuming some permanent effects). Notwithstanding uncertainty about the appropriateness of this estimate, something around this level seems a sensible middle ground. Related to this, an important assumption underlying the Reserve Bank’s analysis is that bank failure is the prime driver of the costs associated with financial crises and that those costs can be avoided through higher bank capital preventing bank failures. While financial crises are indeed costly, it is not necessarily the case that all of the associated costs can be avoided, since often the economic costs are due to misallocation of capital before the downturn – sustained unwise lending, in short, that directs resources to non￾optimal uses. It is also in practice very difficult to separate out the cost of the bank failure from the cost of the event that precipitated it. These issues are evidenced by the fact that countries that avoided bank failures still endured significant economic costs during the GFC.45

Bearing in mind these issues, the Reserve Bank includes a range of results from the optimal capital literature in their consultation paper. Issues with cross-country comparisons are again important here. As previously mentioned, we consider the appropriate benchmark for comparing the Reserve Bank’s proposed minimum of 16% to international studies is an adjusted Tier-1 ratio of 18-20% (16½-18½% for CET1). It is similarly reasonable to assume that Tier-1 capital held by New Zealand banks currently is equivalent to an adjusted ratio of 14-16% when comparing to international studies. With this in mind, support for higher capital ratios is weak:  The 2015 Bank of England paper finds optimal Tier-1 capital ratios of 10-14% in its baseline results – well below the Reserve Bank’s proposals. Alternatively, taking into account permanent transitional impacts (which we think are likely to

45 ‘Benefits and costs of higher capital requirements for banks’, Peterson Institute for International Economics, March 2016.

36 ANZ BANK NEW ZEALAND LIMITED be very important), they find optimal ratios of 7-11%. Adjusting for cross-country differences, current levels of capital in New Zealand already exceed their baseline estimates.  The Federal Reserve paper finds optimal Tier-1 capital ratios of 13-26%. This wide range includes the Reserve Bank’s proposals. However, this study assumes estimates for the cost of financial crisis and probability of crisis that look implausibly large for New Zealand, with an implausibly low discount rate to boot (2.7%). It is noted in that paper that conservative assumptions are used relative to the Bank of England paper because crises tend to be more frequent in the US. We are inclined to dismiss these results, but include them in our assessment below nonetheless.  The 2010 BCBS paper finds optimal Tier-1 ratio of 16-19% (adjusted to be comparable by the Bank of England) – below the Reserve Bank’s proposals, though a little above current levels in New Zealand after adjusting for cross￾country differences.  Other papers produced optimal CET1 ratios of 16-20%, 10-15%, 10% and 12- 16%.46 On average, this points to an optimal ratio of 13-14% – well below the Reserve Bank’s proposals and broadly consistent with where capital ratios are now.  Another paper produced optimal Tier-1 ratios of 12-14% – well below the Reserve Bank’s proposals.47 Adjusting for cross-country differences, current levels of Tier￾1 capital in New Zealand already exceed this. Leaving aside the fact that some of these studies are likely to be less relevant for New Zealand, simply averaging the baseline model results produces an optimal Tier-1 ratio of 15% (after adjusting results from those studies that consider CET1). In the New Zealand context, this would imply an optimal Tier-1 capital ratio of 11-13% including buffers. This is consistent with current capital levels, even taking into account the Reserve Bank’s proposed changes to risk-weighting of assets. Another way to think about this is by comparing the costs and benefits of the changes. Making reasonable assumptions based on the literature, the Reserve Bank’s proposed changes might reduce the probability of crisis from 0.4% to 0.1%. Assuming a plausible cost of financial crisis of 43% of GDP, this might yield a long-term expected benefit of 13% of GDP. Relative to an estimated cost of 20% of GDP outlined above, this simply does not pass the cost-benefit analysis test. And that’s without taking into account New Zealand specific issues, which suggest that the long-term costs of capital increase could be greater and that transitional impacts on the structure of the economy could be significant. Overall, the proposed changes amount to very expensive insurance against very rare, uncertain events – rarer than the Reserve Bank analysis suggests. The international literature on optimal capital ratios supports capital levels around where they are now.

46 ‘The long-term economic impact of higher capital levels’, BIS discussion paper, December 2015, ‘Optimal bank capital’, Bank of England Discussion Paper, January 2011, ‘A cost-benefit analysis of Basel III: evidence form the UK’, Loughborough University Discussion Paper, 2015, and ‘Welfare analysis of implementable macro-prudential policy rules: heterogeneity and trade-offs’, ECB Working paper, August 2015. 47 ‘Benefits and costs of higher capital requirements for banks’, Peterson Institute for International Economics, March 2016.

37 ANZ BANK NEW ZEALAND LIMITED The costs of the proposed changes would be borne by bank shareholders, borrowers, and depositors (since the OCR would need to be lower). It would result in lower GDP and tilt the structure of the economy away from business investment and agriculture towards housing, to the detriment of productive activity generally.

38 ANZ BANK NEW ZEALAND LIMITED Appendix 2: Modelling the potential impacts of the Reserve Bank’s proposed changes to capital adequacy Executive summary  Estimating the macroeconomic impacts of the proposed changes is prone to extreme uncertainty. Three key unknowns include: a. the impact on the required return on equity, b. the impact on the cost of debt/deposit funding, and c. the extent to which banks are able to recoup returns via net interest margins.  

 All up, we estimate that GDP might be 1-3% lower over 10 years, not allowing for an offset from more expansionary monetary policy. In our view, impacts on the upper part of this range are most plausible. We find that most of the adverse effects would be concentrated in the five-year transition period,  However, even with an offset from monetary policy, negative impacts on business and rural lending cannot be avoided. Moreover, given real￾time uncertainties and partial information, even if this is an accurate estimate, the monetary policy response may be smaller or later than modelled.  In our view, the most likely outcome is a lower OCR than otherwise (say ~50bps) and somewhat slower GDP growth for a period of time. As a lower bound, we estimate that a 30bps adjustment in the OCR is the smallest monetary policy response that could be required in the transition period, given the range of possible impacts on the cost of credit and taking into account portfolio rebalancing effects.  This analysis implicitly assumes that the macroeconomic environment remains benign throughout the transition period. There is a risk that banks’ funding costs could increase more than we assume due to some external (global) event. It would also be unusual for the New Zealand economy to go 15+ years without a recession. Either of these factors could exacerbate the impacts in practice, though the impacts would be difficult to isolate in the data.

39 ANZ BANK NEW ZEALAND LIMITED

1. Impacts on bank capital For our analysis we use Reserve Bank estimates for the amount of capital required to meet the proposed requirements.48 The major banks currently have Tier-1 (going concern) capital of 13.4% (13.6% for smaller banks) of risk-weighted assets (RWA). However, once the Reserve Bank’s proposed changes to the RWA are taken into account, the Reserve Bank estimates that this amounts to 11.6% for the major banks (unchanged at 13.6% for smaller banks). To meet the new 16% Tier-1 requirement (with 14.5% from CET149), the major banks would need to increase their capital by $12.8bn ($13.7bn for the banking system as a whole). In addition, the majors would need to replace $6.2bn of non￾compliant AT1 capital, while smaller banks would need to replace $0.1bn. This is held by parent banks so would require a shift in the structure of capital holdings, rather than new capital. We therefore ignore this aspect of the capital changes for our analysis. The amount of new capital required in practice will depend on how much buffer the banks choose to hold on top of the regulatory minimum. Currently, the banks hold a buffer of ~3%. The desired buffer depends on the variability of capital and the regulatory consequences of breaches of the minimum. Given that the new regime is more conservative, we assume that a 2% buffer would be sufficient. This would require an extra $5.8bn for the four major banks and $0.8bn for smaller banks. In total, the Reserve Bank’s proposed changes to capital adequacy could require New Zealand banks to hold up to $20.3bn more Tier-1 capital over the next five years ($18.6bn for the major banks) – or 6% of newly-defined RWA (6½% for majors). This capital raise calculation assumes no growth in assets, for simplicity. We do think that bank assets would still grow; But we are implicitly assuming that capital can be raised to cover growth in assets through the associated growth in earnings. The total macroeconomic impacts could be smaller or greater than we find here depending on the rate of asset growth, the speed of loan repricing, or other practicalities that are beyond the scope of this paper.
2. Impacts on return on equity The Reserve Bank assumes that the capital increase to meet the new requirements can be met over the next five or more years (seven or more for the smaller banks) through retained earnings. However, scope to raise capital (via retained earnings or other means) will depend in practice on shareholders’ required return on equity (ROE) and the marginal cost of funds (taking into account risk, including cycles in risk appetite).

48 Table 9 from ‘Capital Review Paper 4: How much capital is enough?’, Reserve Bank, December 2018; updated January 2019. 49 CET1 comprises ordinary shares, retained earnings and some reserves.

40 ANZ BANK NEW ZEALAND LIMITED

The return on the book value of equity for the banking system as a whole was around 14% over the year to September 2018. Without a change in loan pricing, an increase in capital would reduce this return both as the value of capital is diluted and as the cost of funds increases (due to a greater share of liabilities coming from more expensive equity rather than debt funding). Potentially providing a partial offset, the Reserve Bank assumes that investors’ required return (on both equity and wholesale and depositors’ bank debt) would fall as a result of the investment being less risky (on account of reduced income variability for capital holders and reduced probability of bank failure, of most concern to those providing debt funding).50 The Reserve Bank estimates that this would result in a 50% offset in terms of the increase in the weighted-average cost of funds faced by banks.51

A 50% offset could be perceived as a generous assumption, for reasons discussed below. But taking this at face value, we estimate that the required rate of return on marginal equity might fall from 14% to 12%, compared with an assumed minimum cost of equity of 11%. The Reserve Bank 50% offset estimate is taken from international empirical studies. In practice, however, it likely overstates the fall in New Zealand banks’ costs of funds. This is for two reasons:  First, on the equity side, for the major banks ‘the shareholders’ boil down to the Australian parent bank boards. They have a perceived cost of equity for their New Zealand operations that is unlikely to be greatly affected by the increase in capital holdings.  Second, while the increased capital would improve the standalone credit rating of the major New Zealand banks, this is not, in practice, the rating at which New Zealand banks borrow in international wholesale funding markets. The implicit backstop of the Australian parents means that New Zealand banks are able to fund themselves in debt markets markedly more cheaply than if they were standalone operations. An increase in the capital holdings of New Zealand banks would likely bring the two credit ratings closer together, but unless the Australian parents banks’ credit ratings were to fall below those of their New Zealand subsidiaries (in which case funding costs are likely to go through the roof in any case), then the increased capital holdings of the New Zealand subsidiaries are not the be-all and end-all for lenders. It could similarly be argued that depositors implicitly assume that the Australian parents would provide liquidity if required, as indeed they did during the GFC.

50 The Modigliani-Miller theorem suggests that a firm’s cost of capital will be entirely unaffected by changes in the balance between debt and equity finance (in the absence of real-life frictions such as tax rules), since debt loadings affect required returns on equity inversely. 51 Relative to a 100% offset implied by the Modigliani-Miller theorem.

41 ANZ BANK NEW ZEALAND LIMITED This would also suggest that the impact of higher New Zealand capital holdings on perceived risk – and hence required returns – may in practice be small Given that net interest margins represent the lion’s share of bank profit, and banks would probably still need to see growth in deposits, it seems reasonable to assume that the vast majority of the adjustment to recoup the required return would need to come through adjustments to lending margins. We find that there would necessarily be some impacts on lending growth and/or net interest margins in order to restore ROE to acceptable levels, even if the required ROE were to fall as the Reserve Bank assumes. 3. Long-run impacts on the cost of credit In the long run, the price of credit is likely to adjust to compensate for the higher average cost of funds faced by the New Zealand banks, taking into account any reduction in required ROE. We include a small reduction in the cost of wholesale and deposit debt funding, as discussed in the previous section. This impact on the cost of credit and the neutral interest rate would have long-term persistent impacts, but these could dissipate somewhat over time. Over the course of many years, potentially decades, the financial system would adapt and these effects are likely to dissipate accordingly. Our range of estimates is wide, due to the many uncertainties that exist. However, our mean estimate is considerably larger than cited by the Reserve Bank, which states that the cost of bank credit might increase by 6-8bps for each 1% increase in capital as a share of RWA – amounting to an increase in the cost of credit of around 20-50bps in aggregate, depending on buffers. The single biggest contributor to the discrepancy is that the Reserve Bank’s estimates implicitly assume that ROE can fall to levels that we believe would be unacceptable to shareholders, especially after taking into account dilution of returns for shareholders of existing equity. 4. Transitional impacts on the cost of credit

It is possible that the required – and actual – ROE eventually falls further than this (which is one reason why long-run price impacts might be smaller). One way this could occur is if the opportunity cost – the expected returns on alternative

42 ANZ BANK NEW ZEALAND LIMITED investments – were to fall. However, as discussed above, required ROE may not fall from current levels at all, and in particular not at least until the five-year transition is complete, as investors are likely to demand some compensation for the extra uncertainty that the changes would generate. Impacts would be expected to be largest for agriculture and commercial loans, which require more capital and where impacts on return on equity are expected to be larger. Conversely, price impacts are likely to be smaller for residential mortgages, since less capital is required and competition is greater. In the large corporate/institutional space, foreign competitors may acquire market share if interest margins are widened – they are very likely to have a higher proportion of debt funding, and hence lower costs. We expect very little loan price adjustment in this space.

In addition to pricing adjustment, it is likely that some degree of portfolio reallocation would occur, with banks incentivised to steer lending towards those categories with lower risk weights and hence better returns. This means that residential mortgage lending would be favoured over commercial and agricultural lending, which would impact farm sales and investment in both the business and agricultural sectors. This reallocation would mute the overall impact on credit growth, but only slightly. And to the extent that it steers lending away from funding productive assets, this offset is not particularly positive from a longer-run GDP and productivity perspective.

43 ANZ BANK NEW ZEALAND LIMITED

5. Economic impacts The following results are based on vector auto-regression (VAR) analysis, whereby one statistically estimates how macroeconomic variables typically affect each other in a systematic fashion and then ‘shocks’ one (or more) variables, tracing the implied impacts on all the variables over time. The background business cycle context may well be very important in practice but is necessarily abstracted from in this kind of analysis.52 We model the transition to higher capital requirements over a five-year period, with an assumption that banks would make changes in a smooth fashion in anticipation of well-signalled changes to requirements. We then consider the economic impacts over a 10 year horizon. Charts of responses can be found in the Annex. All responses are relative to a baseline counter-factual. We do not show or explore what this baseline might look like. Over the long run, we estimate that GDP might be 1% lower permanently, as was discussed in Appendix 1.

52 The variables in the VAR are mortgage rates, house sales, house prices, residential investment, consumption, farm sales, ANZBO agriculture investment intentions, ANZBO business investment intentions, and business investment.

44 ANZ BANK NEW ZEALAND LIMITED

45 ANZ BANK NEW ZEALAND LIMITED

46 ANZ BANK NEW ZEALAND LIMITED

47 ANZ BANK NEW ZEALAND LIMITED



48 ANZ BANK NEW ZEALAND LIMITED

50 ANZ BANK NEW ZEALAND LIMITED

Page 1 of 23 ASB Bank Limited Submission to the Reserve Bank consultation Capital Review Paper 4: How much capital is enough? May 2019

EXECUTIVE SUMMARY Page 2 of 23 EXECUTIVE SUMMARY

1. ASB Bank Limited (ASB) fully supports the role of the Reserve Bank of New Zealand (the Reserve Bank) in promoting the maintenance of a sound and efficient financial system. ASB is proudly committed to its purpose of accelerating the financial progress of all New Zealanders, and accordingly supports the calibration of New Zealand bank capital settings to a level that promotes a stable and resilient financial system.  
2. However, ASB does not support the level and composition of the proposed capital requirements as detailed in the Reserve Bank consultation “Capital Review Paper 4: How much capital is enough?” (the Review Paper). In its current form, the proposal is at significant risk of ‘over insuring’ against a systemic banking crisis in New Zealand and will have a significant impact on many New Zealanders and the broader economy.
3. The proposal does not have sufficient regard for the Reserve Bank’s efficiency objective, and as such the proposal risks being inconsistent with the purpose of the Reserve Bank of New Zealand Act 1989 (the Reserve Bank Act), that is, “to promote the prosperity and well‐being of New Zealanders, and contribute to a sustainable and productive economy… .”1
4. The Reserve Bank states that the long‐run macro‐economic impact of the proposed increased capital requirements is likely to be modestly negative2 , and argues this is a relatively small price to pay for a significantly more stable banking system. In our view, the Reserve Bank analysis overstates the likelihood of a New Zealand systemic banking crisis and underestimates the negative macro‐economic impacts in coming to this conclusion.  
5. The proposal overestimates the risk of a New Zealand systemic banking crisis by: (a) setting New Zealand bank capital levels to be the highest in the developed world, without taking sufficient account of the simple and more traditional nature of banking in New Zealand; (b) focusing on capital levels in isolation and not taking sufficient account of the combination of existing and anticipated Reserve Bank financial system stability mechanisms, such as liquidity requirements, Open Bank Resolution, macro‐prudential tools and interventions, and depositor protection; and (c) using overly conservative modelling inputs, and assigning little value to New Zealand stress test results and rating agency assessments. These concerns are detailed in Section 1, with supporting analysis in Attachment 1.  
6. The agriculture, small business and first home buyer sectors are important to the financial prosperity and well‐being of New Zealand. Lending to these sectors is expected to reduce as banks re‐evaluate their business models as a result of the proposed higher capital levels. These impacts have been underestimated by the Reserve Bank due to optimistic assumptions regarding the effect of the proposal on: (a) the availability of credit; and   (b) customer interest rates. These concerns are detailed in Section 2.  
7. Section 2 also outlines the fundamental dependence New Zealand has on offshore capital and funding, which to date has been primarily sourced by the four largest New Zealand banks. If New Zealand’s capital minimums become significantly higher on a relative global scale, it will reduce

1 The Reserve Bank Act, Section 1A(1). 2 Reserve Bank: Capital Review Paper 4: How much capital is enough? (January 2019), Paragraph 75; 1 percentage point Tier 1 capital increase equates to a 3 basis point decline in GDP.

EXECUTIVE SUMMARY Page 3 of 23 the attractiveness of investing in New Zealand and thereby restrict credit availability to New Zealanders. To the extent the proposal results in credit rationing, non‐bank financial institutions or smaller banks may not be able to access offshore markets to the same level as that provided by the four largest New Zealand banks today. Where other large foreign banks enter the New Zealand market, it is questionable whether they would support low‐return, high capital intensive sectors. In addition, the long‐term, through‐the‐cycle commitment of these foreign banks could be challenged. 8. In Section 3 we present an alternative capital mix and structure which would enable banks to better support New Zealanders and the broader economy, while continuing to maintain a strong, stable and resilient financial system. The alternative capital structure:    increases the current minimum Common Equity Tier 1 (CET1), Tier 1 and Total Capital requirements for New Zealand banks;  places New Zealand bank CET1 ratios above the 90th percentile when compared on an internationally comparable basis;  recognises the Reserve Bank’s preference for higher‐quality capital;    provides greater transparency and international comparability of the strength of New Zealand bank capital ratios;  retains the positive risk management disciplines, efficient and effective capital allocation and more accurate risk‐based pricing that Internal Ratings Based (IRB) modelling facilitates;  recognises the systemic importance of large banks;  provides the Reserve Bank with flexibility to moderate future excessive credit growth; and  delivers a more efficient cost of capital which will ultimately benefit all New Zealanders.   9. Attachment 2 summarises our response to a number of the specific questions posed in the Review Paper.   10. ASB strongly recommends that, prior to finalisation of any proposed revisions to the quantum and composition of bank capital, a full macro‐economic cost benefit analysis (including an assessment of industry loss rates incurred through previous financial crises) should be completed in conjunction with the New Zealand Treasury. These findings should be made public. This analysis should be a key input in deciding whether the final capital settings support a sound and efficient financial system, and promote the prosperity and well‐being of all New Zealanders.  

SECTION 1 Page 4 of 23 SECTION 1 – THE PROPOSAL OVERESTIMATES THE LIKELIHOOD OF A NEW ZEALAND SYSTEMIC BANKING CRISIS

1. ASB is supportive of measures to ensure that New Zealand banks are well capitalised. However, given the relatively simple banking models that operate in New Zealand, the Reserve Bank’s proposed capital settings are extreme.  
2. The Reserve Bank proposes to increase capital requirements for IRB banks via significant increases to both risk‐weighted assets (RWA) and the CET1 ratio. The combination of these settings overstates the likelihood of a New Zealand systemic banking crisis and would set New Zealand bank capital levels to be the highest in the developed world.
3. In addition, the calibration of the proposed capital settings appears to be largely dependent on an empirical modelling approach, which does not take sufficient account of other existing Reserve Bank financial stability settings, uses overly conservative inputs, and assigns little value to New Zealand banking system stress tests and rating agency assessments. Simplicity of New Zealand banking models
4. New Zealand banks have relatively simple banking models with a traditional retail and business lending focus, predominantly in the domestic market. In addition, these banks typically do not engage in complex investment banking activities, including complex structured credit products which have been significant contributors to previous global banking crises in the United States and Europe.  
5. New Zealand banks are not subject to the same risks and inter‐connectedness common to larger and more complex banking models that justify higher capital requirements.   The proposal would set New Zealand bank capital levels to be the highest in the developed world
6. On the basis of information collated by PricewaterhouseCoopers (PwC) for its report “International comparability of the capital ratios of New Zealand’s major banks” published in May 2019 (the PwC Report), the Reserve Bank’s proposed CET1 ratio of 14.5% translates to an internationally comparable CET1 ratio of approximately 24.5% for New Zealand IRB banks. The higher internationally comparable ratio is primarily the result of the conservatism built into the proposal with respect to RWA, and the impact of the higher IRB scalar and output floor.
7. Figure 1 shows PwC’s estimates of the weighted average internationally comparable CET1 ratios for the four largest New Zealand banks compared to the weighted average by country of 97 major banks in Australia, North America, Europe and Asia.3 For the New Zealand banks, it shows the weighted average internationally comparable CET1 ratio under current Reserve Bank capital rules,4 and what this ratio would be under the Reserve Bank’s 14.5% CET1 ratio proposal. It further illustrates, that not only are New Zealand banks already well capitalised on a comparative basis by being in the top quartile internationally, the Reserve Bank proposals will make New Zealand the most highly capitalised banking jurisdiction in the developed world. The New Zealand banking system is not exposed to risks that demand such a high capital setting on a comparative basis, particularly given the relative simplicity of New Zealand bank balance sheets.  
8. We also highlight that the proposed New Zealand CET1 ratio shown in Figure 1 is likely understated, as it does not include any capital buffers above the regulatory minimum. Banks typically hold a capital buffer above minimum capital requirements to ensure that any short‐ term capital fluctuations do not inadvertently result in a breach of minimum capital requirements.

3 The underlying bank capital data used in the analysis was the most recently published full or half year‐end data as at 31 July 2017. 4 Capital ratios for these New Zealand banks have increased since this date. For example, ASB’s CET1 ratio has increased from 9.7% as at December 2016 to 11.5% as 31 December 2018 (calculated in accordance with BS2B “Capital Adequacy Framework (internal models based approach)”, (November 2015)).

Figure 1 Zealand major ba SOURCE: (May 201 9. APR 2014 min The com this com 10. We histo pop How too perf coun chan Capital l 11. Ther cons sett mac Reso prom Act 12. Give take than : The weight d banks unde anks in Austr PwC “Interna 19) A completed 4 Australian imum CET1 r PwC report mparable bas on an intern mplex and mo accept that orically been ulation grow wever, the pr high especia formance is n ntries, partic nge.   levels have b re are a num sidered whe ings. These i cro‐prudentia olution, as d mote financi Review inclu en the inter‐r e into accoun n setting cap ted average r the Reserv ralia, North A ational compa d an extensiv Financial Se ratio require equates the is. The Reser nationally co ore exposed New Zealand n more volat wth, and grea roposed CET ally when con not a certain cularly given been conside mber of other n determinin nclude exist al interventio iscussed in A al soundnes uding deposit relatedness nt all these m ital requirem internationa ve Bank prop America, Eur arability of the ve capital rev rvices Inquir ements of the APRA requir rve Bank’s pr mparable ba to volatility d is exposed ile compared ater depende 1 ratio differ nsidered in t n determinan the unknow ered in isola r Reserve Ba ng the proba ing Reserve ons such as t Attachment 1 s (for examp tor protectio of these fina mitigating fac ments in isola SECTION 1 Page 5 of 23 ally compara posal compar rope and Asi e capital ratio view in 2017 ry. As a conse e largest ban rement to a roposal is ap asis, despite in offshore f to its own id d to Australia ence on agric rential betwe erms of rela nt of future r wn impact of ation nk tools and ability of a sy Bank financi the high loan 1), as well as ple, the liquid on). ancial stabilit ctors when d ation. ble CET1 rati red to the we ia s of New Zeal 7 in response equence of t nks to 10.5% CET1 ratio o pproximately the Australia funding mark diosyncratic a due to its s cultural prod een the two tive risk. We relative GDP major extern d mitigants th ystemic bank al system sta n‐to‐value (L s anticipated dity review a ty measures, determining ios for the fo eighted aver land’s major b e to recomme their review, , effective fr of 15.4% on a y 9 percentag an banks bei kets. risks, and its smaller size, duction and o countries as e also note th performanc nal factors s hat should h king crisis and ability measu LVR) restricti reviews des and Phase 2 o we believe capital requi our largest N rage by coun banks – updat endations fr APRA increa rom 1 Januar an internatio ge points hig ng larger, m s GDP growth more cyclica overseas tra s shown in Fi hat historic G e between t uch as climat ave also bee d associated ures (for exa ons and Ope signed to furt of the Reserv it is importa irements, rat New try of 97 te paper” om the ased the ry 2020. onally gher than ore h has al de. gure 1 is GDP he two te en capital ample, en Bank ther ve Bank nt to ther

SECTION 1 Page 6 of 23 Modelling approach 13. The Reserve Bank has placed strong reliance on empirical evidence and modelling using relatively narrow data sets when calibrating its proposed capital settings. Moreover, it has assigned little value to stress tests results and rating agency assessments specific to New Zealand. In addition, overly conservative model assumptions have resulted in much higher proposed capital settings than required to protect against a 1 in 200 year loss event.   14. This is discussed in detail in Attachment 1.  

SECTION 2 Page 7 of 23 SECTION 2 – TO MEET THE HIGHER CAPITAL REQUIREMENTS, BANKS WILL RESHAPE THEIR BUSINESSES WHICH WILL HAVE A NEGATIVE IMPACT ON THE NEW ZEALAND ECONOMY

1. Increased bank capital requirements will negatively impact GDP by hampering two important drivers of economic growth, namely the availability and price of credit. The Reserve Bank assumptions in relation to how these drivers impact GDP draw heavily on international academic research and modelling. ASB is of the view that these assumptions are not directly applicable in a New Zealand context.  
2. In addition, the Review Paper does not address how the GDP impact of the proposal will be distributed across the New Zealand economy given that the rational economic response from New Zealand banks will be to increase pricing and restrict credit availability to low‐returning, capital‐intensive sectors.
3. The proposals will significantly increase the amount of capital New Zealand banks are required to hold against various lending portfolios and consequently will considerably increase the cost to New Zealanders. Banks must achieve a sufficient level of return on capital deployed in order to continue attracting capital. Higher regulatory capital requirements and lower returns in certain exposures or sectors will inevitably reduce their attractiveness relative to alternative lending opportunities.  
4. This is also the case where lower capital requirements exist for exposures of equivalent risk elsewhere within a banking group. A recent example of this is the Commonwealth Bank of Australia (CBA) Group’s Institutional Banking & Markets business. After a period of relatively low risk‐adjusted returns, it optimised RWA and increased pricing in the institutional loan book in order to improve low shareholder returns. The business actively reduced capital intensity of its exposures and focused on targeted impairment management, which saw its credit RWA drop by 31% over circa two years (from A$106bn as at June 2016 to A$73bn as at December 2018).
5. Any significant increase in capital requirements will similarly result in ASB focusing on optimising low‐returning sectors, leading to re‐pricing and/or reduced risk appetite. This will impact sectors that are important to New Zealand including agriculture, small businesses and first home buyers.  
6. The concerns around credit availability and price increases are addressed below. Assumptions regarding availability of credit  
7. The Review Paper is not explicit in how potential credit rationing has been factored into the GDP impact assessment. International studies have been used by the Reserve Bank to benchmark ‘on average’ impacts on GDP based upon increased capital levels. This benchmarking includes generic assumptions around a reduction in the availability of credit at a certain price, rather than explicitly modelling this from a New Zealand perspective.  
8. We have two key concerns with respect to this approach, as outlined below. Impact on customers
9. The proposal will result in material impacts for certain customers, namely first home buyers, small business owners and agri‐businesses, via reduced availability of credit and/or increased borrowing costs. These customers are economically and socially important to New Zealand. These sector impacts do not appear to have been specifically modelled in the Review Paper or other material that the Reserve Bank has released. We believe these restrictions on credit availability will be significant, and will detrimentally impact the overall well‐being of many New Zealanders.

SECTION 2 Page 8 of 23 10. According to Reserve Bank data,5 approximately 59% of high‐LVR borrowers (>80% LVR) are first home buyers. Under the proposals, this type of lending will attract a significant capital increase amongst the four largest New Zealand banks. For example, capital requirements are five times as much for home loans with LVRs between 80% and 90% compared to home loans with LVRs below 60% under ASB’s IRB accredited home loan model. When coupled with the proposed significantly higher capital requirements, this produces a capital impost that will make loans to this sector unattractive without significant repricing. Also, over 40% of ASB’s small business borrowers have a LVR greater than 70%. Higher capital will likely lead to negative downstream impacts to first home buyers and small business customers, due to increased pricing and/or the reduced availability of credit. 11. This situation is similar for the agricultural sector6 except that regulatory capital intensity is significantly higher than home lending and most small business lending, and returns on capital in this sector are on average already well below banks’ cost of capital. The proposed capital requirements are excessive and would further reduce returns in this sector, leading to credit rationing and increased pricing to recover the cost of the additional capital allocated to these exposures. We estimate the pricing impact for agri‐business loans will be on average more than twice that for other business loans. These changes would be implemented at a time when the sector is facing a number of challenges including increasing risk from climate change, and higher costs associated with environmental protection and water quality. 12. Where banks reduce their risk appetite in these important sectors, it is likely that banks will implement these changes early in the transition period. This heightens the risk of a more immediate negative impact on customers and the economy. Indeed, ASB is already reviewing and adjusting its risk‐appetite settings in certain sectors in anticipation of increased capital requirements. 13. In addition, the capital proposals for IRB banks result in particularly punitive and irrational capital outcomes for both agricultural and high‐LVR home lending in arrears  even those loans with adequate security. For example as shown in Table 1: (a) A farm loan of $100,000 with a 65% LVR that is 90 days in arrears with a 5% loan loss provision results in $100,000 of capital and provision allocated in addition to the underlying security value; and (b) A $100,000 home loan with an 85% LVR (typical first home buyer) that is 90 days in arrears with a 1% loan loss provision results in $88,000 of capital and provision allocated in addition to the underlying security value.   14. Table 1 shows that the capital requirements and provisions for these loans provide for an unrealistic reduction in security values. We cannot find evidence of any financial crisis in the developed world that precipitated such extreme reductions in security values. 15. These irrational capital outcomes are a product of the conservative Reserve Bank IRB assumptions around loss given default (LGD), the proposed increase in the IRB scalar and the significant difference between Basel framework loss‐absorbing capital parameters.7 We believe that these capital outcomes are disproportionate to the risks inherent in these exposures and will likely result in IRB banks reducing their risk appetite for these types of loans.   

5 Reserve Bank: New Residential Mortgage Lending by Borrower Type – C31 (26 March 2019) 6 In 2018, agriculture constituted 4% of production‐based New Zealand GDP and was New Zealand’s largest export sector, accounting for over 50% of New Zealand’s merchandise exports and over a third of New Zealand’s total goods and services exports. 7 Basel framework assumes an 8% Total Capital ratio, compared to the Reserve Bank’s proposed Total Capital ratio of 18%.

SECTION 2 Page 9 of 23 Table 1: Examples of dollar amount of capital required under current and proposed Reserve Bank Total Capital minimums for loans in arrears (IRB banks)   Security Value Total Capital and Provision $ Change Reduction in security value covered by capital and provision (under proposed requirements) Current Proposed $100k of rural lending   65% LVR and 90 days in arrears, 5% provision $154,000 $54,000 $100,000 +$46,000 100% $100k home loan 85% LVR and 90 days in arrears, 1% provision $118,000 $46,000 $88,000 +$42,000 90% Participation of other lenders in the New Zealand economy 16. New Zealand is fundamentally dependent on offshore sources of capital and funding. The four largest New Zealand banks currently have $35bn of capital and $66bn of funding that has been sourced offshore, which has provided significant support for historical GDP growth. As at September 2018, these banks had funded over $373bn of lending assets to New Zealanders, representing 86% ($432bn) of all customer lending by New Zealand registered banks. Over the past decade, these banks have increased invested capital by a total of $15bn, which has provided circa 90% of domestic lending growth8 that has helped support the New Zealand economy. New Zealand capital markets do not have the capacity to provide this level of support, either via domestic savings or wholesale markets. 17. The Review Paper has indicated that it expects other lenders to become more active in New Zealand to support continued GDP growth. This may be a significant challenge for non‐bank financial institutions and smaller banks as they may not have the scale, or adequate credit rating to be able to access cost‐effective capital and funding to support growth to the same level as that provided by the four largest New Zealand banks, which benefit from AA‐ rated parents. Where large foreign banks enter the New Zealand market, they may focus on high‐return, low‐ risk sectors which may not be conducive to supporting important sectors. In addition, the long‐ term, through‐the‐cycle commitment of these foreign banks could be challenged. As an example, during the Global Financial Crisis a number of foreign banks exited their Australian operations when their parents experienced stress within their home jurisdiction. 18. The four largest New Zealand banks receive long‐standing, through‐the‐cycle support from their respective Australian parents who historically have had, and will continue to have, strong economic and geopolitical ties to New Zealand.   19. We also note that shareholders’ commitment to their New Zealand businesses is not limited to the amount of physical capital deployed in New Zealand. Significant franchise value has been built up by the Australian major banks over many decades. Further, the complimentary nature of their investments in the closely‐linked New Zealand economy provides a strong incentive for the Australian banks to support the four largest New Zealand banks in times of stress.   Assumptions regarding customer interest rates 20. The Reserve Bank takes the view that the impact of the higher bank capital requirements on interest margins will be minimal (in the vicinity of 20 to 40 basis points). This analysis relies heavily on the assumption that a higher capital level and the perceived increased soundness of the bank will reduce its cost of funding. We do not believe that debt investors will accept a lower return to the extent the Reserve Bank has assumed (refer to Attachment 1 for further information).  

8 Approximately $135bn.

SECTION 2 Page 10 of 23 21. ASB economists9 estimate that the higher capital requirements will likely result in a 50 to 75 basis point increase in customer lending rates, which is significantly above the Reserve Bank estimate. This will have significant downstream consequences for the economy. The regulatory capital that is required to be held at product and sector levels is a major driver influencing pricing decisions. Capital‐intensive sectors will be disproportionately impacted by repricing. For example, ASB’s analysis suggests that for a borrower with a $500,000 mortgage with a 15% deposit (a typical first time buyer), borrowing costs will increase by over $3,000 per year. A farm loan of $3 million could face increased borrowing costs of $35,000 per year on average. In many cases, however, the reduced availability of credit may have more significant well‐being impacts. 22. We highlight that the pricing impacts outlined above assume no reduction in the Official Cash Rate (OCR). The Reserve Bank has stated that reductions in the OCR could offset the impacts of higher borrowing costs as a result of the proposal. It is important to stress that such a reduction would negatively impact a greater number of savers and retirees who are heavily reliant on interest income from bank deposits. In ASB’s case, we have 30% more customers that earn interest from a deposit product than pay interest on a lending product. The impact on savers of an OCR reduction will be even more apparent in an environment where interest rates are already historically low.

9 ASB: What to make of proposed bank capital requirements (April 2019).

Figure 2 Zealand country SOURCE: (May 201 4. We appe theo as d 5. In ad and         6. We Credit R 7. The Zeal betw 2: The weight d banks unde of 97 major PwC “Interna 19) consider the etite needed oretical basis iscussed in A ddition to pr productive e increases th banks; places New comparable recognises t provides gre Zealand ban retains the p and more ac recognises t provides the delivers a m set out belo RWA scalar a four largest land. One of ween the am ted average r ASB’s alter banks in Aus ational compa e alternative d to limit the s), given the Attachment 1 roviding a be economy, as e current mi Zealand ban basis; he Reserve B eater transpa nk capital rat positive risk ccurate risk‐b he systemic e Reserve Ba ore efficient w our recom nd output fl New Zealan the Reserve mount of cap internationa rnative capit stralia North arability of the capital struc e probability level of cons 1. etter balance s an integrate inimum CET1 nks above the Bank’s prefe arency and in ios; managemen based pricing importance nk with flex t cost of capi mmendations loor d banks are e Bank’s state ital required SECTION 3 Page 12 of 23 ally compara tal structure h America, Eu e capital ratio cture is withi of a systemi servatism inc e between fin ed capital pr 1, Tier 1 and e 90th percen rence for hig nternational nt disciplines g that IRB m of large ban bility to mod tal which wi s on each com the only ban ed objectives to be held f 3 ble CET1 rati compared t urope and As s of New Zeal in the range c banking cr corporated i nancial syste roposal, the a Total Capita ntile when co gher‐quality comparabili , efficient an odelling faci nks; derate future ll ultimately mponent of nks with curr s is to reduce for a particul ios for the fo o the weight sia   land’s major b of the Reser isis to a 1 in nto the Rese em resilience alternative c al requireme ompared on capital;   ity of the str nd effective c litates; e excessive c benefit all N the capital s rent IRB accr e the degree ar asset type our largest N ted average banks – updat rve Bank’s ris 200 year eve erve Bank mo e and a susta capital struct nts for New an internati ength of New capital alloca credit growth New Zealand structure. reditation in e of discrepa e between th New by te paper” sk ent (on a odelling, inable ture also: Zealand onally w ation h; and ers.   New ncy he IRB

SECTION 3 Page 13 of 23 and standardised banks. This is based on the assumption that IRB banks hold less capital than standardised banks. The Reserve Bank has sought to address this discrepancy by requiring that IRB banks scale up RWA to approximately 90% of RWA if calculated using a standardised approach (a combination of increasing the current IRB scalar from 1.06 to 1.20 and introducing an output floor of 85%).   8. In fact, under the current rules, IRB banks already hold higher capital against certain exposures compared to standardised banks. As an example, ASB’s agriculture portfolio RWA as at March 201810 was higher than had it been calculated under a standardised approach. The Reserve Bank’s proposal to increase the IRB scalar and the introduction of the 85% output floor will require ASB to hold even more capital against these portfolios than under current requirements, and compared to a standardised modelling approach. 9. The proposed significant increase in the IRB scalar and the introduction of the 85% output floor will result in IRB banks converging to less risk‐sensitive standardised RWA. This will result in IRB banks pricing their lending based on less risk‐sensitive standardised RWA, thereby devaluing the importance of granular risk modelling. This risk modelling is foundational to better understanding relative asset risks and customer pricing, and enables more efficient and effective allocation of bank capital across the New Zealand economy. Properly‐applied IRB modelling does not create an unfair advantage but rather reflects a better understanding of credit risks and allocation of capital to those risks. 10. Table 3 below reinforces IRB risk sensitivity for loans 90 days in arrears relative to standardised banks. It highlights that under current capital rules IRB banks already hold 3.1 and 4.3 times the amount of loss‐absorbing capital for rural and high‐LVR home loans, respectively. Under the proposal, capital levels are further increased to levels that are extreme in both absolute and relative terms.   Table 3: Examples of dollar amount of capital required (excluding provision) under current and proposed Reserve Bank Total Capital minimums for loans in arrears (IRB banks)   $100k of rural lending 65% LVR and 90 days in arrears (assuming 5% loan loss provision) $100k home loan   85% LVR and 90 days in arrears (assuming 1% loan loss provision) Capital treatment Current   Proposed Current Proposed IRB bank $49,000 $95,000 $45,000 $87,000 Standardised bank $16,000 $27,000 $10,500 $18,000 IRB to Standardised multiple 3.1x 3.5x 4.3x 4.8x 11. To further reinforce the effectiveness of their advanced risk‐based modelling in managing credit risk, the Review Paper highlights that the impaired asset ratios of the four largest New Zealand banks have been significantly lower than the overall market over the last 30 years. The Review Paper cites an impaired asset ratio (which shows the percentage of impaired and past due assets over total assets) of 3.2% on a simple average basis across all New Zealand banks.11 When this average is weighted by each banks’ relative size, the ratio falls to 1.4%. Given their significant share of total lending, this reduction reinforces that the large banks in New Zealand have materially lower impaired assets than the smaller banks. This highlights the critical role that the advanced risk‐based modelling plays in risk assessment, thereby strengthening the overall

10 The date of the Reserve Bank Quantitative Impact Study in relation to bank capital. 11 Reserve Bank: Capital Review Paper 4: How much capital is enough? (January 2019), Paragraph 58.

SECTION 3 Page 14 of 23 resilience of the banking system and supporting the efficient and effective allocation of capital across the New Zealand economy. 12. ASB believes that a better alternative to narrowing the capital outcomes between standardised banks and IRB banks is to have standardised banks adopt the Basel III framework for residential exposures. Residential mortgage lending represents approximately 65% of the total credit risk of the three largest New Zealand standardised banks (excluding Rabobank). Adoption of the Basel III framework for residential exposures (or the similar approach proposed by APRA), would result in RWA measured under the revised, more risk‐sensitive standardised approach reducing to levels more closely aligned to residential mortgage RWA for IRB banks. 13. The application of higher relative conservatism proposed by the Reserve Bank through RWA adjustments (rather than increases to the headline capital ratio) also reduces transparency and comparability of a bank’s capital adequacy levels on a global basis. Given the reliance that New Zealand’s banking sector has on offshore funding, maintaining transparency is particularly important so investors can understand the relative strength of New Zealand’s capital settings when compared to investment alternatives in other jurisdictions. ASB recommendations:  Maintain the IRB scalar at the existing 1.06; and  Introduce an output floor of 75%. 14. Our recommended RWA adjustments for IRB banks are still more conservative than the Basel framework12, where the 1.06 scaling factor has been removed and an output floor of 72.5% has been introduced. D‐SIB buffer 15. The introduction of a D‐SIB buffer will recognise the systemic importance of D‐SIBs in the New Zealand context and will bring New Zealand in line with global practice. Setting the D‐SIB buffer at a level greater than 0% provides for higher loss absorption in a clear and transparent manner, and will strengthen the New Zealand banking system as a whole. ASB recommendation:  Set the D‐SIB buffer at 1.0%. CCB 16. ASB agrees that the CCB should be increased from its current level and recommends this be increased to 6.5%. This increase is premised on the principle that New Zealand banks should be well capitalised and is more reflective of a 1 in 200 year loss event in the New Zealand context. APRA has determined that for Australian D‐SIBs to be considered “unquestionably strong”, they should maintain a CET1 ratio greater than 10.5% (consisting of a 5.0% CCB, a minimum CET1 ratio of 4.5% and a D‐SIB buffer of 1.0%). Setting a CCB buffer 1.5% higher than APRA recognises the increased idiosyncratic risk of the smaller New Zealand economy, and in our view would more appropriately insure against the risk of a systemic banking crisis without unduly hampering economic growth.   ASB recommendation:  Set CCB to 6.5%. CCyB 17. The original intent of the CCyB was to allow banking regulators to raise bank capital requirements during periods of excess credit growth. The Reserve Bank has proposed a formal introduction of a CCyB to be set at 1.5% (as an ‘early set’), in a period of slow credit demand,

12 BCBS: Basel III: Finalising post‐crisis reforms (2017).

whic envi Rese 18. A ke Supe econ a CC New the Figu SOU 19. Sinc prud redu de‐r 20. Furt has ban 21. Ther Aust cons RWA ASB    AT1 and 22. ASB com requ cont

13 BCBS: Gu 14 ASB subm ch is contrar ironment ris erve Bank to ey indicator o ervision (BCB nomy. The B CyB of betwe w Zealand ec Basel III fram ure 3: New Ze RCE: The Rese ce 2013, the dential tools uce New Zea risking of ban thermore, AP private secto ks’ regulator refore, the in tralian paren sumption for A by limiting recommend Introduce CC Develop a ro the Reserve Include a mi meet the hig d Tier 2 capit has previou mpliant AT1 a uirements.14 tain non‐viab

uidance for natio mission on the Re y to the orig ks unnecessa o increase bu of excessive BS) and utilis CBS suggest een 0% and 2 onomy has b mework wou ealand Credi erve Bank Ma Reserve Ban and sector‐s aland bank ba nking sector PRA requires or credit exp ry capital min ntroduction nt banks. Thi r New Zealan lending to c dations: CyB initially s obust metho Bank’s decis inimum pre‐a gher capital tal sly provided and Tier 2 los As discussed bility triggers

onal authorities op eserve Bank Capi ginal intent o arily dampen ffers during credit expan sed by APRA s that a gap 2.5%. Figure been below 0 ld consider s it‐to‐GDP Ga cro Prudentia nk has succes specific RWA alance sheet exposures th s banks to inc posures. The nimums incr of a CCyB in s in turn will nd exposure capital intens set at 0%; odology inclu sion to set a announceme levels. d a submissio ss‐absorbing d in our prev s, conversion perating the coun ital Review Paper SECTION 3 Page 15 of 23 f the CCyB.13 ning econom an expansio nsion, propos , is the moni of between 3 below illus 0% since Ma setting a CCy ap al Indicators M ssfully addre A overlays. T t risks and po han would h clude propor introduction easing in pro New Zealan l result in hig s, providing sive sectors. uding key ind CCyB greate ent period o on in support capital secu vious submis n and redem ntercyclical capit r 2: What should 3 3 Introducing mic growth, a onary credit p sed by the B itoring of the 2 and 10 per strates that t rch 2010. Th yB of greater March 2019 (F ssed excessi hese tools h otentially hav ave been ac rtional CCyB n of a CCyB w oportion to t d will increa gher required further moti dicators (e.g. er than 0%; a f 12 months t of the cont urities in New sion, these s ption featur al buffer (2010). qualify as bank c g a CCyB in th and limits the period.   asel Commit e Credit‐to‐G rcentage poi the Credit‐to his is well bel r than 0%. igure 1A) ve credit gro ave been use ve resulted i hieved via im for each jur will result in A their New Ze se the cost o d returns on ivation to red Credit‐to‐GD and to allow ban inued inclus w Zealand ba securities sho res to ensure capital? dated 8 S he current e options for ttee on Bank GDP Gap of a ints could eq o‐GDP Gap fo low the zone owth through ed effectivel in a more tar mplementing isdiction in w Australian p ealand expos of capital for regulatory c duce New Ze DP Gap) to in nks sufficient ion of Basel nk capital m ould continu e appropriate September 2017. r the king an quate to or the e where h macro‐ y to rgeted g a CCyB.   which it arent sures. r the capital ealand nform t time to III minimum ue to e

SECTION 3 Page 16 of 23 investor demand and diversification. The ability to issue these securities to third‐party investors diversifies sources of loss‐absorbing capital, reducing singular capital reliance on parent banks. 23. The inclusion of these types of securities in a bank’s capital structure has the effect of materially reducing a bank’s cost of capital. For example, ASB estimates that AT1 capital provides for loss absorption at a cost approximately one third of CET1, with Tier 2 at one quarter of CET1. 24. ASB acknowledges the Reserve Bank’s concerns regarding the complexity of these securities and distribution to retail investors. These issues have been successfully addressed in other jurisdictions. ASB would welcome discussions with the Reserve Bank and the Financial Markets Authority to work collaboratively to ensure appropriate treatment. 25. AT1 levels should be increased by 0.5% to 2.0%. This results in a more conservative proportion of AT1 (2.0%) to Tier 1 (assuming a Tier 1 ratio of 14.0% as per ASB’s alternative proposal) of just over 14.0%15 compared to the Basel III framework’s AT1 to Tier 1 proportion of 15.7%.16   ASB recommendations:  Set the AT1 and Tier 2 capital requirements each to 2.0%; and  Maintain existing non‐viability triggers and conversion features for contingent capital in line with global practice and the Basel III framework. Leverage ratio 26. The Reserve Bank is seeking views on the leverage ratio as another mechanism to strengthen financial system resilience. The BCBS introduced the leverage ratio as a mechanism to manage the build‐up of excessive balance sheet leverage not captured through risk‐based capital ratios. This is especially relevant in jurisdictions where IRB banks operate with relatively low risk weights (e.g. United Kingdom, Switzerland and the Netherlands) where the leverage ratio serves as a backstop to risk‐based capital requirements preventing excess build‐up of leverage in periods of economic expansion.   27. ASB views the introduction of a leverage ratio requirement in the New Zealand context as unnecessary given the already conservative RWA settings and the proposed higher Tier 1 minimum. The combination of these factors would make it unforeseeable that the leverage ratio minimum would be breached prior to breaching the capital minima. The calculation and reporting would unnecessarily create additional administrative burden for banks and the Reserve Bank with little or no prudential benefits. ASB recommendation:  No explicit leverage ratio requirement should be implemented for New Zealand banks.

15 2.0% of AT1 divided by 14.0% Tier 1. 16 1.5% of AT1 divided by 9.5% Tier 1 (4.5% CET1 minimum plus 2.5% CCB plus 1% D‐SIB plus 1.5% AT1).

ATTACH BY THE R Calibrat

1. The requ cons outl Calibrat financia
2. The to e insu typic have even
3. As a desi resil (a) Figu SOU (b) (c) (d) (e) HMENT 1: FU RESERVE BA ion of minim Reserve Ban uirements. W sidered whe ined below. ion of New Z l stability me Reserve Ban stimate the urance regula cally entirely e almost no nts occurring a banking reg gned to redu lience in the Macro‐prud macro‐prud deleveraging implementa borrowers.   ure 4: Stock o RCE: Reserve Open Bank R manner, the Liquidity: via proposed Re financial inte Monetary po contributes Payment sys and this sou URTHER ANA ANK mum capital nk has consid We believe th n determinin Zealand bank easures nk has applie capital requi ators must se y outside the ability to imp g.   gulator howe uce the likeli New Zealan ential toolkit ential measu g in the resid tion of high‐ of high‐LVR m Bank Macro P Resolution: a ereby reducin a Open Mark esidential Mo ermediaries olicy: the ab to a system stem regulat rce of system AT ALYSIS ON TH requiremen dered a num here are othe ng the proba k capital sett ed a probabil ired for a 1 i et capital lev eir control (s plement syst ever, the Res ihood of a sy d financial sy t: the Reserv ures (i.e., hig dential prope ‐LVR restricti mortgage len Prudential Ind allows a dist ng systemic ket Operatio ortgage Obli supply suffic ility to contr c banking cr tion: this has mic risk; TTACHMENT Page 17 of 23 HE ECONOM nts ber of factor er important ability of a sy tings does no lity of ‘ruin’ m n 200 year lo vels by estim uch as earth temic mitiga serve Bank h ystemic bank ystem. These ve Bank was gh‐LVR restric erty sector a ions and incr nding dicators – Mar ressed bank risk and mai ns (OMO) an gation secur cient high‐qu rol the risk th isis; s significantly T 1 3 IC ASSUMPT rs when sett t factors that ystemic bank ot sufficiently methodology oss event. Un ating the like quakes and nts that can as a number king crisis occ e tools includ one of the fi ctions). Figu s a direct res reasing equit rch 2019 (Figu to be kept o ntaining con nd on‐going e ritisation pro uality and liq hat unwarran y reduced re TIONS AND M ing the prop t should hav king crisis and y take into a y used by ins nlike banking elihood of ev pandemics). influence th r of mitigatin curring, ther de: irst regulato re 4 below i sult of the Re ty requireme ure 7A) open for bus nsumer conf enhancemen ogramme des quid assets to nted asset pr liance on en MODELLING posed capital e also been d capital sett ccount other surance regu g regulators, vents that ar They theref he likelihood ng tools that reby increasi rs to introdu lustrates eserve Bank’ ents for inve ness in an or dence;   nts such as t signed to en o the OMO; rice inflation d‐of‐day set USED tings, as r ulators , re fore of such are ng uce ’s stor rderly he sure n tlement

ATTACHMENT 1 Page 18 of 23 (f) Sector level capital requirements: the ability of the Reserve Bank to adjust sector‐specific capital risk parameters (for example, minimum LGD for home lending and agriculture); (g) CCyB: the ability to increase this buffer during periods of excessive credit growth; and (h) Disclosure: greater transparency through reporting, including bank disclosure statements and bank financial strength dashboard reporting. The importance of transparency and disclosure is that it contributes to strong governance, risk management and financial discipline. 4. In assessing the appropriate level of capital, the Reserve Bank should assess the reduced likelihood of a banking crisis given the existence and availability of these prudential tools. Overly conservative inputs in the underlying modelling 5. ASB is concerned that the proposal is significantly more conservative than the probability of a systemic banking crisis occurring once in every 200 years. This level of conservativism built into the Reserve Bank’s modelling overstates the risks within the banking sector, particularly so when there is already a strong regulatory and operating environment that seeks to reduce the probability and consequence of a systemic banking crisis within New Zealand. 6. The Reserve Bank has relied on two portfolio risk models in their analysis of capital requirements. We believe that there is a significant level of conservatism incorporated into the key assumptions (e.g. probability of default, the LGD, and model confidence intervals) of both models.   7. One of the models includes a 99.7% confidence interval chosen for the Review Paper which equates to a 1 in 333 year event of a systemic banking crisis. Recalibrating this confidence level to a 1 in 200 year event (which implies a 99.5% confidence interval) reduces the required capital levels by approximately 1.9 percentage points (to approximately 13.6%) of Tier 1 capital.   8. The second model estimates capital through sampling distributions of assumptions. The ranges from which the parameters are sampled are conservative. For example, the average probability of default from the models of the four IRB banks is 1.1%. This estimate is an average of the modelling of the four IRB banks, each of whose model has been approved by the Reserve Bank. The range of assumptions adopted by the Reserve Bank was 1.5% to 3.0%. A more reasonable range would be to include the IRB bank 1.1% average probability of default as a data point. The upper end of the range appears to be heavily influenced by the experience of small banks. 9. Similarly, the range for LGD of 35% to 50% does not include the LGD average of the approved modelling for the four IRB banks (29%) or the results of the 2017 Reserve Bank stress test which indicated a weighted average LGD rate of 31%. The range is overly influenced by the 42.5% LGD required by the Reserve Bank for high‐LVR farm lending and does not give sufficient weighting to the substantial volume of low‐LVR residential lending. The LGD calculation also does not take into consideration the Reserve Bank’s ability to influence lending quality through macro‐ prudential regulation (e.g. high‐LVR restrictions) and the ability to specify conservative model parameters when appropriate. 10. There was very limited information provided on the choice of correlation parameter which is a critical model input in determining the severity of loss in a crisis (and therefore the capital required to absorb that loss). The Reserve Bank’s own modelling shows the significant sensitivity of this parameter in determining the quantum of bank capital required. For example, the difference between a 24% and a 28% correlation parameter was 1.8 percentage points of Tier 1 capital.17 This equates to approximately $5bn additional capital in the New Zealand banking system.

17 Reserve Bank: Explanatory note on portfolio risk modelling in the New Zealand context (January 2019).

ATTACHMENT 1 Page 19 of 23 Assigns little value to stress tests and rating agency settings 11. New Zealand banks undertake regular stress tests to assess their capacity to withstand severe macro‐economic, operational and idiosyncratic downside shocks. These tests, set by the Reserve Bank (and APRA for Australian‐owned New Zealand banks), are well beyond any economic conditions New Zealand has experienced. For example, the types of assumptions used in these tests have included GDP falls of 8.8%, unemployment peaking at 16%, commercial and residential property values falling 44% and 36%, respectively.18   12. Similarly, an independent stress test undertaken by the IMF as part of its 2017 Financial System Stability Assessment of New Zealand concluded that the New Zealand banking system is adequately capitalised against adverse macro‐economic shocks.19 The stress scenario assumed a 7.5% decline in real GDP over the period, negative inflation, and a 35% fall in house prices. 13. Banks demonstrated the required resilience to operate and recover without external support during these scenarios. ASB believes that stress test results are a relevant and important input into the assessment of capital settings and system resilience, and is concerned that the Reserve Bank has given these results little consideration in its capital calibration.   14. Rating agencies assign ratings to entities to reflect their relative credit worthiness. Typically, ratings reflect an assessment of how likely it is that an entity will default.20 Standard and Poor’s (S&P) currently assigns a standalone credit rating of BBB to the four largest New Zealand banks. S&P’s assessment takes into account, amongst other things, the current amount of loss‐ absorbing capital that is available to protect against default. 15. Rating agencies assess the ability of their ratings to reflect relative default risk by observing the actual default rates that occur across their rating scale. For example, S&P conducts an annual global default study21, with its latest report indicating that over the past 36 years, the probability of BBB rated entities failing is equivalent to a 1 in 600 year event. This indicates that rating agencies consider that the probability of a New Zealand bank failure even under current capital settings is significantly less likely than the Reserve Bank proposed calibration of a 1 in 200 year event.   Price of credit 16. The economic theory known as the Modigliani‐Miller offset suggests that there is an offsetting reduction in funding costs that comes from the perceived safety of increased capital holdings. The Review Paper assumes around a 50% offset; that is, around half of the increase in a bank’s average funding costs that would be implied by a change to a higher share of capital funding would be offset by a lower required return on a bank’s capital and non‐capital funding. It is likely that this offset is overstated within a New Zealand context because: • The Australian‐owned New Zealand banks trade at risk premia of 15 to 20 basis points higher than that of their parents’ issued debt, largely as a result of reduced liquidity and a higher country risk premium. We do not consider that holding high levels of capital will materially change liquidity and country risk premia differentials. • There is a limit to how small a risk premium investors would be willing to accept on New Zealand bank debt before substituting to bank debt from other larger more liquid economies such as Australia, the US and Canada.

18 Reserve Bank Reverse Stress Test (2016). 19 IMF: IMF Country Report No. 17/119 ‐ New Zealand Financial System Stability Assessment (May 2017). 20 S&P and Fitch ratings both take this approach. 21 S&P: Default, Transition and recovery: 2017 Annual Global Corporate Default Study and Rating transitions (April 2018).

ATTACHMENT 1 Page 20 of 23 • Our discussions with rating agencies to date indicate that they will not revise New Zealand bank issuer ratings if the Reserve Bank proposal is implemented as currently proposed. In fact, Fitch ratings recently noted that parental support uplift may actually be downgraded if the Reserve Bank's final capital proposals are so onerous that there is a significantly increased chance of the parents divesting their New Zealand operations.22 17. The above points support our view that holding higher levels of capital are unlikely to lead to a reduction in bank funding costs.

22 Fitch Ratings: Fitch Affirms New Zealand Four Major Banks at AA‐ (3 April 2019).