Capital Review Principles

Ref #6758869 v1.2
MEMORANDUM FOR FSO Committee
FROM Financial Policy
MEETING DATE 2 November 2016
SUBJECT Capital Review Principles
FOR YOUR Decision
It is recommended that the Committee:
1. Note we have been asked to develop principles for the capital review. These build on
previous principles.
2. Note the main components of the forthcoming capital review papers.
3. Agree to the principles outlined below, or suggest refinements.
Background
1. Financial Policy has been asked to bring a paper to FSO Committee to establish
some principles for the Capital Review. So far, we have presented the revised Harrison ‘optimal’ capital model, a literature review on capital, and some cross country capital level comparisons. Agreements reached on the capital principles outlined in this paper will help frame the options presented in subsequent decision papers.
2. At various times in the past principles have been presented to FSO. Some have been
pitched at our overarching regulatory philosophy, and some have been more targeted, such as the principles we used for adopting Basel III.1
3. There is a general theme that runs through the iterations of principles:

• Unless there are good reasons not to we will adopt international standards
  (which usually means minor departures from the Basel framework at the margin, such as not adopting the leverage ratio, rather than a bespoke approach);
• We take a substance over form approach and tailor our requirements to New
  Zealand (in practice this has usually meant adopting the Basel framework, with more conservative calibrations to reflect New Zealand conditions); and
• We will have regard to international comparability, particularly with Australia,
  with the caveat that the preceding principle will take precedence (there is potential for tension here, if for example, under the second bullet point one concluded that the IRB framework was not best for New Zealand whereas Australian-owned banks wanted to continue with IRB).
  4. We believe these principles are uncontroversial and should continue to hold, as
  should the principles used for the Basel III review (see appendix). The purpose of this paper is to develop principles targeted at the broader capital review. 1 See (#4486107) Basel III – quality of capital, a 2011 FSO decision paper.

Ref #6758869 v1.2
5. The paper also updates the Committee on recent Basel developments and our
thinking about some of the options that should be presented to FSO in subsequent decision papers, and raises some communication issues for your consideration. Principles for the Capital Review
6. The principles we propose, drawing in part on previous work and internal discussions
are the following:
a. Banks should be regulated and that, as part of this regulation, banks should be required to have a minimum amount of capital (equity-like funding) which will exceed the economic capital they would maintain in accordance with private incentives. b. Capital should be available and useful before and in a crisis
c. Capital requirements should be sensitive to the riskiness of banks’ exposures.
d. Capital regulation should not unduly benefit or disadvantage some banks more than others. e. Capital regulation should be conservative and fit for New Zealand’s circumstances. f. Capital regulation should be consistent with other aspects of our regulatory approach and practical to administer. g. Where Basel requirements are calibrated to less conservative standards than our current requirements the prior is not to reduce our standards (i.e. no backsliding). Banks should be regulated and that, as part of this regulation, banks should be required to have a minimum amount of capital (equity-like funding) which will exceed the economic capital they would maintain in accordance with private incentives.
7. Banks maintain certain levels of capital in order to protect against a certain level of
losses and meet the expectations of those who invest or do business with them. However the wider costs to the economy that would occur on the failure of one or more banks (systemic risk) are not accounted for by the capital decisions of the individual banks, meaning that banks may privately maintain a lower capital level than is optimal from a societal point of view. Capital regulation seeks to address the market failure that arises from the fact that the individual decision makers (banks) do not fully take into account the societal costs that their actions could impose on the wider public. Capital should be available and useful before and in a crisis
8. One justification – not the only one – for regulatory capital requirements is that banks
which have more capital are less likely to fail in when there is some large unexpected shock, and will impose fewer costs on society when they do fail. For this justification to be valid, providers of capital must actually be “on the hook” for losses.
9. It is reasonably uncontroversial that ordinary shareholders will suffer the first loss, to
the extent of their invested share capital and any retained earnings or reserves not yet paid out as dividends. The likelihood of a (conscious) government bailout of bank shareholders seems low.
10. It is less clear, however, that holders of other instruments that currently count as
capital will suffer losses.
11. Complex instruments that cancel or convert to equity when bank failure is imminent
are relatively untested, particularly in New Zealand. And to the extent that these

Ref #6758869 v1.2 instruments have been purchased by retail investors, governments might face incentives to bail them out, in the same way it has incentives to bail out depositors.
12. In the case of long-term debt, which can count as “Tier 2” capital, there is also a
possibility that a government will feel compelled to make investors whole, particularly if there are concerns about the bank’s ability to attract new funding.
13. Of course, we do not know what current or future governments will do: perhaps the
above concerns are overstated, perhaps not. The nature of these instruments is unambiguously different from common equity and they are untested here at least, that much we know. Capital requirements should be sensitive to the riskiness of banks’ exposures.
14. This principle is based on the following assumptions:

• banks that have riskier portfolios are more likely to fail and cause financial
  instability than banks with safer portfolios; and
• holding more capital reduces the risk of failure.

15. The Basel I standard did not allow for much differentiation of capital requirements on
    the basis of risk. For example, it gave a single risk weight (50%) to all residential mortgages. Under the Basel I standards a bank with a risky portfolio could have the same regulatory capital requirement as a bank with a much safer portfolio. It has been argued that this gave banks incentives to take on more risky loans (which is really only credible up to a point).
16. The Basel II standards were promoted as a remedy for these perverse outcomes.
    Basel II makes it possible to recognise the risk of individual exposures in a much more fine-grained way, either by relying on banks’ internal credit models or by linking risk to credit ratings (implementations in New Zealand and Australia also link to loan￾to-value ratios).
17. Some prominent people have argued that the benefits of additional risk-sensitivity are
    spurious or that there is no real benefit, and that it would be better to rely on a simple – not risk-weighted – leverage ratio as the regulatory instrument, or at least include it as an important complement or backstop (e.g. Andrew G. Haldane, executive director at the Bank of England, and Thomas M. Hoenig, director of the US Federal Deposit Insurance Corporation). Capital regulation should not unduly benefit or disadvantage some banks more than others.
18. APRA, when it was considering implementation of the Basel II framework, noted that
    “[m]ultiple calculation approaches for different institutions may raise potential competitiveness issues”2 .
19. We observe, on average, that the four New Zealand banks using the internal models
    approaches have lower capital requirements than banks using standardised approaches. We have tried to limit the standardised-IRB gap in some portfolios, with varying degrees of success. In some other portfolios the gaps are large and unwarranted. Some gaps may be warranted, such as when underlying risk is lower. However, it is very difficult to know if this is the case, while putting standardised banks at a competitive disadvantage if capital requirements are lower merely because of the regulatory framework. Regulatory neutrality is something we are 2 http://www.apra.gov.au/adi/Documents/cfdocs/Implementation-of-the-New-Basel-Capital-Accord.pdf

Ref #6758869 v1.2 generally concerned about, and it is a factor in both some of the key Basel reforms and the FSI review in Australia.
20. Under the current framework, it would be impossible for standardised banks to
achieve an average risk weight for housing loans which is as low as the average risk weights currently reported by the IRB banks, no matter how prudent their lending.
21. There has been an argument made internationally that banks need a financial
incentive, in the form of reduced capital, to induce them to invest in internal models (which, it is implied, would improve their management of credit risk). There are good and obvious reasons to challenge this argument, but even if one accepts it conceptually, the failure of Basel to estimate the incentive needed was an obvious policy blunder, right from the start. Capital regulation should be conservative and fit New Zealand’s circumstances.
22. New Zealand has a small open economy that is exposed to changes in global supply
and demand. New Zealand is quite reliant on foreign investment and on bank intermediation of investment. The domestic banking sector is concentrated, with a small number of large banks accounting for most of the market and the majority of their lending is also concentrated in a limited number of areas, being to home-owners and farmers. New Zealand is exposed to movements in international commodity prices and house price declines.
23. In such an environment systemic bank crises are more likely unless banks are
financially strong. This suggests capital requirements should be higher in New Zealand than elsewhere where there is greater diversification of risk in the financial system.
24. At the same time, if one accepts the received wisdom that it is costly for banks to
increase capital then New Zealand could be more sensitive to the negative effects of this. Compared to other developed economies, New Zealand is relatively more reliant on the banking sector. In addition, the small number of dominant participants might limit competitive mechanisms which would otherwise limit pass-through of cost increases.
25. We should, however, be quite careful with the above argument. Most recent studies
and modelling work we have looked at suggest that the negative impacts (on activity and so on) of higher capital only become significant (a net cost) at much higher levels than observed here, or possibly even contemplated here even if capital increased by a reasonable amount. The following box recaps the findings from the literature. Box: Effects of capital requirements and optimal capital ratios As discussed at FSO Committee in September (#6665392), a number of recent academic and central bank research papers have quantitatively estimated of the effects of changes in capital requirements on bank’s funding costs, lending rates and steady-state GDP, as well as ‘optimal’ capital ratios that balance the potential costs of capital against societal benefits. Based on our assessment of these studies we made the following observations:

• Recent empirical studies of the Modigliani-Miller theorem suggest that the effect
  holds only at a rate of around half, that is, changes in banks’ funding structures towards costlier sources (equity) pass through to higher lending rates at a rate of around 50%.

Ref #6758869 v1.2

• Recent estimates suggest that a one percentage point increase in banks’ tier 1
  capital ratios leads to around a 5-8 basis point increase in lending rates.
• The effects of higher capital requirements on the steady-state level of GDP (through
  lower credit availability) are modest, at around a 0.01% to 0.05% fall per percentage point increase in banks’ tier 1 capital ratios.
• Across a range of studies, estimates of optimal CET1 capital ratios using this type of
  cost-benefit framework arrive at a range of 10% to 20%, with a number of estimates landing around the 14-18% level depending on modelling assumptions.
• An alternative method based on real-world banking crises suggests that risk￾weighted total capital ratios in the range of 15-23% would have avoided creditor
  losses in around 85% of the 27 advanced economy banking crises experienced in the past 40 years. It is important to note that the empirical estimates use data samples based on ‘status quo’ levels of capital; while we consider that these results are likely to be informative when considering incremental changes to current settings, they may be less relevant to discussions of an order-of-magnitude change to capital levels (e.g. a 30-50% leverage ratio).

26. This supports a conservative bias principle - or to put it another way, at the sort of
    capital levels present in the New Zealand banking system we should be slightly assymetric in that we should worry more about capital falling as a result of regulatory requirements and frameworks, than we should about modest increases. We should recognise, however, that what conservative means in practice is actually quite a slippery concept. This issue was addressed in the previous cross-country comparison paper by Gael. We can aim to be conservative against the international standards, or against domestic risks, or against comparator countries. However, all of these comparators will have different actual risk profiles that change through time and are measured on subtly different, but potentially quite important, applications of the international standards.
27. In practice, we have used ‘relative to the standardised framework’ as a proxy. For
    example, we have targeted our efforts at the exposures most important in the New Zealand context (housing and rural), we have required slightly higher risk weights under our standardised requirements compared to the Basel framework (e.g. housing risks weights for higher LVR mortgages), and we have tried to benchmark the IRB outcomes (with varying success) against our version of the standardised framework. This accounts for much of the ‘conservatism’ our banks report relative to other regimes.
28. We continue to believe a conservative bias is a relevant principle.
29. In terms of framework fit for New Zealand, the core business of New Zealand banks
    is the provision of credit to households and businesses. Except in carrying out this core business – for example, using swaps to hedge interest rate or exchange risk on its lending book – New Zealand banks’ involvement in complex financial structuring or trading is limited.
30. Given the straightforward business model of New Zealand banks, there might be a
    case for a simpler regulatory framework in New Zealand than elsewhere.

Ref #6758869 v1.2
Capital regulation should be consistent with other aspects of our regulatory approach and practical to administer.
31. The Reserve Bank has historically preferred a light-handed regulatory approach that
emphasises self-discipline and market-discipline. We dedicate relatively few resources to bank supervision. (Of course, we have yet to fully reflect on the FSAP findings and any potential implications for our model.)
32. The Basel capital standards, even in the slightly simplified form implemented by New
Zealand, are complex and voluminous. The standards were formulated by countries that have large and complex banking sectors, and which have a hands-on – sometimes to the point of extreme intrusiveness – approach to bank supervision.
33. We have found it difficult to keep pace with the changing Basel standards and banks’
implementation of them. Approval of banks’ internal capital models and capital instruments, which are required in our implementation of the standards, imposes a relatively heavy burden on our small organisation.
34. A simpler approach might be more in line with our generally hand-off regulatory
approach, and our administrative capability (though the latter should be very much a secondary consideration in coming to a policy view). Moreover, it would also be more supportive of the market discipline pillar of our framework, given the opacity inherent with internal models.
35. A simpler approach might also be better suited to New Zealand’s banks, which are
small by international standards and have comparatively few resources (though some foreign-owned banks can draw on the expertise and resources of international parents). Overview of next steps
36. We previously provided the Committee with a list of recent or current proposals for
changes to the Basel capital standards (#6493041). Some of these are noted in our proposed approach below. Recent comments suggest that final revisions of the core capital standards (i.e. the standardised framework, and adjustments to IRB) will still be published “early next year” but there is always the possibility of further delay. Implementation dates for the revisions appear to be no earlier than 2019.
37. Some of the following issues are not particularly dependent on the release of final
Basel standards. Examples would include numerator considerations (qualifying capital and how much), or the leverage ratio. It is possible, but more difficult, to proceed to consultation on denominator-type considerations ahead of finalised Basel standards. However, we will bring the high-level options to FSO (on the IRB framework) for consideration rather than wait for Basel. Proposed approach
38. In an earlier paper to the Committee (#6537247) we proposed a capital work
programme. We now propose some specific items of further work within that programme, for your consideration. In each part of the work, we would have regard to the principles discussed above.

Ref #6758869 v1.2
Credit Risk
39. We propose to reconsider which methods within the Basel framework should be
available to calculate capital requirements for credit risk. Banks can currently use the internal-ratings-based (IRB) method if the Reserve Bank has accredited them to do so, or the standardised method. IRB Approach
40. Our experience of administering the IRB method is that it is difficult to assess
whether or not risk is being properly reflected in models.
41. For instance, the IRB method requires, as a parameter, a probability of default which
is the average across many cycles. While this may be feasible for large, internationally active banks, this parameter is difficult to estimate in practice for banks such as those in New Zealand. Our banks often have historical data which does not include a major downturn. The problem is particularly acute for asset classes that are strongly correlated with the macroeconomic cycle, such as housing loans, since there is no recent experience in New Zealand of a housing downturn of the kind seen in, say, Ireland or Spain after the financial crisis. Banks can attempt to correct for the lack of a sharp downturn in historical data, say by simulating additional hypothetical observations, but this is inevitably a judgement-laden exercise. Estimation is also complicated by changes in banks’ underwriting standards over time, which can mean that historical data is inapplicable to new loans.
42. In practice, because we are not confident that we can assess the true level of risk we
have imposed restrictions on models to keep average IRB risk weights at levels we feel comfortable with (generally below, but not too far below, standardised levels). This impacts on the potential benefits of a more risk-sensitive capital requirement. We are currently conducting a benchmarking exercise that we hope will throw more light on the relative level of risk carried by each bank in housing and dairy exposures, but it is unlikely this will be a panacea. Moreover, rough calibrations and overlays can lose their effectiveness over time.
43. The Basel Committee has raised similar concerns about the reliability of risk
estimates for other portfolios. It found that risk estimates assigned to loans to sovereigns, banks and large corporates varied unjustifiably depending on the internal rating models used by each bank. Due to their opacity, modelled approaches can substantially weaken market discipline without alternative disclosures of capital adequacy on a common basis (e.g. leverage ratio or standardised reporting). It’s arguable whether IRB capital reporting allow market discipline to work effectively.
44. As well as being unsure about the final outcomes, we find the IRB approach time
consuming to administer. We also have limited resources to devote to monitoring and encouraging improvement of banks’ models and associated processes. The Basel Committee recently proposed revisions to the IRB approach which would limit its use (not allowed for bank or large corporate exposures), restrict some input parameters, and introduce an output floor (set relative to the capital requirement under the standardised approach). These are sensible risk mitigants if IRB is to be maintained, but the prior question is whether concerns about the IRB framework (opacity, the difficulty of robustly modelling the parameters, information asymmetries between the banks and the regulator, and so on) argue for removing the framework altogether.

Ref #6758869 v1.2
Standardised Approach
45. The standardised approach is less complex than the IRB approach, though a degree
of complexity seems unavoidable to allow for a thorough treatment of off-balance sheet exposures and credit risk mitigation.
46. In theory, the standardised approach is not as risk sensitive as the IRB approach.
The Basel Committee recently proposed a revised standardised approach which is designed to be more risk-sensitive. In New Zealand we already have a marginally more risk-sensitive standardised approach than in the stock Basel framework, because we link risk to the loan-to-value ratio of residential loans (Basel uses a single risk weight for all such loans), but it is possible it could be made more risk￾sensitive. It seems sensible to wait until the Basel revisions are finalised before making decisions about this part of our capital requirements. We understand the timing is the first quarter of 2017. Proposed work
47. We propose to consider:

• whether or not the IRB approach should be retained at all;
• if the IRB approach is retained:
  o whether it should be retained only for some portfolios (e.g. retail portfolios); o whether internal models banks should be required to calculate and report standardised as well as IRB risk weights and capital requirements, to more easily reveal cases in which models might be under- or over￾estimating risk; o whether parameter estimation should be restricted (e.g. floors for probability of default, prescribed LGDs); o whether there should be output floors, which prevent the average risk weight for a portfolio falling too far below the average risk weight under the standardised approach; o whether to decouple IRB accreditation from minimum capital requirements, i.e. allow banks to signal their superior risk management through their accreditation, but report and require capital on a 100% of standardised basis (the US framework); o whether the standardised approach can be made more risk sensitive; and, o whether there are any other simplifying enhancements that can be made to the standards. Definition of Capital

48. We propose to consider whether or not the current definition of capital is appropriate.
    In particular we will ask whether we should continue to recognise:

• Tier 2 capital (and which kinds)?
• Additional Tier 1 capital (and if so which kinds)?

49. Our consideration of Tier 2 capital will take into account international developments
    which are requiring more funding that is (intended to be) “loss-absorbent” in a resolution but is otherwise not very capital-like (“TLAC”).
50. To the extent that we continue to recognise Additional Tier 1 or Tier 2 capital, we will
    consider whether or not the current regulatory approach – which requires the Reserve Bank to carefully vet every instrument – is the best one. One possibility is to rely more on self-certification by banks that their instruments comply (attestation).

Ref #6758869 v1.2
This will reduce the administrative load before the instrument is issued, but might require changes to supervisory procedures for checking attestations and enforcing any breaches, given our experience to date (non-compliant instruments signed off by bank boards as compliant). Methods for assessing operational risk and market risk
51. Our primary focus will be on capital requirements that relate to credit risk. At present
these account for the majority of the total capital requirement. The treatment of operational and market risks is a secondary consideration. Operational Risk
52. For operational risk, banks can use the Advanced Measurement Approach (AMA) if
accredited by the Reserve Bank to do so, or the standardised approach.
53. The AMA is used by the four largest New Zealand banks. The Reserve Bank does
not hold itself out to be expert in the assessment of operational risk. For this reason the Bank has imposed, as backstops, floors on the amount of capital held for operational risk. In some cases these floors are binding.
54. Internationally, operational risk modelling is regarded as an immature discipline,
relative to modelling of market or credit risk, and the Basel Committee on Banking Supervision has been concerned that AMA models have not been producing reliable results. The Committee has recently proposed that the AMA approach be removed entirely from the Basel standards.
55. The standardised approach is very simple: capital requirements depend only on
assets and income of the bank. The Basel Committee has proposed a revised standardised approach which would allow banks to recognise, partially, their past experience of operational risk events. Market Risk
56. All banks in New Zealand use a standardised approach to market risk. This
approach is based on one developed by the Basel Committee in the 1990s but has not been updated to reflect subsequent changes. The Basel Committee recently released a new market risk standard which allows internal modelling or standardised approaches. At least on a first reading, the standardised approach appears quite complex; it is not clear whether it could be implemented by all the smaller locally incorporated banks.
57. Unusually, New Zealand applies the market risk framework to all market risks,
including interest rate risk in the banking book (IRRBB). APRA requires IRB￾accredited banks to calculate capital for IRRBB, but not using the standard market risk framework. The stock Basel framework allows countries to deal with IRRBB under Pillar II of the three pillars in the framework (i.e. by supervisory adjustment to capital, if felt necessary) or to use a standardised method to calculate capital. Proposals
58. We propose to consider:

• whether the AMA should be retained at all
• if the AMA is retained

Ref #6758869 v1.2 o whether the existing floors should be reviewed o whether the standardised approach should be amended to allow banks to take into account their past experience of operational risk events

• whether the current market risk standard remains fit for purpose (we have no good
  alternative at the moment, but could signal that we plan to develop one) Leverage Ratio

59. The Basel capital framework includes a leverage ratio. This is regarded as a
    backstop which will rarely be binding. However, some commentators argue that a simple leverage ratio is more relevant and reliable than a risk-weighted capital ratio, particularly in times of stress, and suggest it should be given a more prominent role.
60. The Reserve Bank decided not to adopt a leverage ratio at the time of Basel III
    adoption, but has publicly noted it will review this position as the international standards and adoption are finalised. The FSAP will also conclude that we should reconsider the earlier decision.
61. We propose to reconsider whether New Zealand should adopt a leverage ratio and, if
    so, whether it should be the Basel ratio or some other ratio.
    Snapshot: trends in bank capital
    Since Basel III was implemented, banks have issued more than $7bn of new hybrid debt instruments, which qualify as either AT1 or Tier 2 capital. Value of new issues, Year to Dec AT1, $m Tier 2, $m Total, $m 2013 500 0 500 2014 0 539 539 2015 2,253 1,615 3,868 2016^ 2,238 15 2,253 Four years of issuance 4,991 2,169 7,160 ^ this does not include anticipated issues for the remainder of 2016 – for example, a new AT1 issue by Kiwibank. To put that in perspective, common equity has increased (as a result of new issues, retained earnings and valuation changes) by just over $5.0bn, so only marginally more than the total amount raised via AT1. Common equity remains the predominant source of bank capital however.

Ref #6758869 v1.2
The new issues have been sold to a combination of retail investors, parents and offshore parties. While capital levels have been increasing relative to total assets, the risk-weighted exposures3 have been declining as a proportion of total assets for two large banks, ANZ and BNZ. In the case of ANZ, capital is held in effect against half the balance sheet. These two effects – new capital issuance and declining effective risk weights – are reflected in rising capital ratios. 3 ‘Risk weighted assets’ or RWA is measured by risk weighted credit exposures plus a value for i) operational and ii) market risk. - 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 $m Year to June Value of regulatory capital NZ registered banks (source: Financial Statements) T2 AT1 CET1 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% Ratio Risk Weighted Exposures to Total Assets ANZ ASB BNZ WNZL

Ref #6758869 v1.2
Level of capital
62. We propose to consider whether the current minimum capital ratios (Common Equity
Tier 1, Tier 1, Total) remain appropriate.
63. New Zealand’s current regulations require marginally more capital for credit risk than
some other countries’ regulations, after controlling for the composition of bank’s lending portfolios (and undoubtedly less for some other countries). Our previous work estimated this to be within the range of 100-200 basis points in terms of CET1 ratios. But other countries have recently moved, or are moving, to increase capital ratio requirements materially higher than Basel III minima, so it is timely to consider whether levels here remain sufficient. Communications
64. We have given vague indications of the content of the capital review in previous
Financial Stability Reports and industry updates.
65. We propose to provide a more detailed description – though not as detailed as the
one above – of the issues which the review will consider in the forthcoming (November 2016) FSR. (A description has already been drafted for the FSR team). 5.00% 6.00% 7.00% 8.00% 9.00% 10.00% 11.00% 12.00% 13.00% 14.00% 15.00% 2000 2002 2004 2006 2008 2010 2012 2014 2016 Tier 1 capital ratio ANZ ASB BNZ Westpac 8.00% 9.00% 10.00% 11.00% 12.00% 13.00% 14.00% 15.00% Total capital ratio (self reported) ANZ ASB BNZ WNZL

Ref #6758869 v1.2
66. We would make it clear in the FSR that work is at an early stage, that no decisions
have been made, that we will consult with banks in the usual way before coming to decisions, and that we will make reasonable allowances for transition if there are any regulatory changes.
67. However, the proposed work plan is likely to provoke a negative reaction from the
banking sector. There are some issues which banks are likely to be especially sensitive to:

• Consideration of restricting the recognition of additional tier 1 or tier 2 capital.
  Banks have existing AT1 and T2 issues and will be planning more. A public proposal to consider their future may unnerve investors and affect required rates of return.
• Consideration of removing or restricting the IRB approach for calculating capital
  for credit risk. More than changes to operational or market risk standards, removal of the IRB approach would significantly increase capital requirements for the four largest banks.

68. In general, internal models banks will dislike the suggestion of any restrictions on
    their ability to determine their own estimates of risk. All banks will dislike the suggestion of a narrower definition of capital. Standardised banks might be pleased with the possibility of a more risk-sensitive standardised approach, or more parity with internal models banks. OIA s6(a)

Ref #6758869 v1.2
APPENDIX
Principles (from 2011 “Basel III – quality of capital”)

1. The Basel III framework sets out a list of eligibility criteria for each component of capital
   (CET1, AT1, and Tier 2) – 37 criteria in total. In addition a list of (12) regulatory deductions (applied mainly to CET1) are set out.4
2. We propose the following principles to guide our adoption of the various Basel III criteria
   and requirements:
   (i) Our prior is that unless there are good reasons not to, we would adopt the Basel III standards. This will involve tightening some of our existing standards and introducing some new standards. (ii) Where Basel III is less conservative than our existing standards, the starting prior should be to not relax our standards. (iii) Where an existing standard is not impacted by Basel III we will retain it. (iv) We will not generally use Basel III as an opportunity to strengthen our standards in areas unrelated to Basel III. (v) We will take a substance over form approach, such that we may adapt the Basel III standards to ensure a suitable fit for New Zealand conditions. (vi) We will have regard to international comparability and consistency (especially consistency with Australia), subject to principle (v).
3. Where there are trade-offs between satisfying one principle versus other – we will
   recommend solutions to the FSO Committee on a case-by-case basis. However in these circumstances principles (i) and (ii) will carry the most weight. 4 See “Basel III: A global regulatory framework for more resilient banks and banking systems - revised version June 2011”, Basel Committee on Banking Supervision, December 2010, esp. pp12-27 (also available here: http://www.bis.org/publ/bcbs189.htm).