Reserve Bank of New Zealand Consultation on Covered Bonds Regulatory Framework

Ref #4206354
Consultation Document:
COVERED BONDS
Consultation Paper
The Reserve Bank invites submissions on this Consultation Paper by 19 November 2010. Submissions and enquiries about the consultation should be addressed to:
Kevin Hoskin
Adviser, Financial System Policy
Prudential Supervision Department
Reserve Bank of New Zealand
PO Box 2498
Wellington 6140
Email: Kevin.Hoskin@rbnz.govt.nz
Please note that submissions may be published. If you think any part of your submission should be withheld on the grounds of commercial sensitivity or for any other reason, you should indicate this clearly. October 2010

Ref #4206354
Consultation Document: Covered Bonds
1 INTRODUCTION
1.1 Background
1 The Reserve Bank is consulting on a policy for the issuance of covered bonds in New Zealand. 2 A covered bond is a class of corporate bond that is backed by certain assets. The coupon of the bond and the repayment of the principal is “guaranteed” or “covered” by assets in a specific pool of collateral. 3 Covered bonds may have significant benefits for issuers by lengthening the term structures of their funding, diversifying funding by providing access to new investors, and enhancing domestic capital markets. This benefits the financial system by reducing the likelihood of liquidity problems affecting an issuer, and promoting the sound and efficient operation of the system.
1.2 Consultation
4 This consultation paper proposes a formal regulatory framework to support the issuance of covered bonds by New Zealand financial institutions. It sets out our proposed limits on the issuance of covered bonds by New Zealand issuers and discusses the appropriate treatment of residential mortgage-backed securities (RMBS). 5 The Reserve Bank’s initial view is that the formal regulatory framework should include the introduction of a legislative framework. This paper considers what this might mean from a New Zealand perspective, and also the capital adequacy implications and disclosure requirements associated with covered bond issuance. 6 Throughout this paper we have identified a number of specific questions on which we are seeking the views of interested parties. However, stakeholders are also encouraged to provide comments on any other issues that they consider relevant to the development of covered bonds in New Zealand.

Ref #4206354
2 BACKGROUND
7 A covered bond is a debt instrument that is ‘covered’ by a security interest in a specific pool of collateral (the cover pool). The cover or collateral for covered bonds is usually formed from high quality assets (mainly mortgages) that are segregated from the other assets of the issuer. 8 A guarantee over the cover pool is then granted to a trustee in favour of the covered bond holders. The value of the assets in the cover pool is greater than the amount of covered bonds outstanding so that the bonds are “over-collateralised” in order to meet the costs of running the pool. Assets are generally required to meet certain minimum standards to be included within the cover pool, and any non-performing assets are replaced so that the value of the cover pool is maintained. 9 In broad terms, a covered bond is a similar instrument to an RMBS, but the primary difference between an RMBS and a covered bond is that covered bond holders have dual recourse to both the assets in the cover pool and to the issuer. An RMBS only offers single recourse to the assets contained within the RMBS structure. With covered bonds, investors are protected by high quality collateral, over-collateralisation, and a preferential claim on the cover pool, and by an equal claim to any residual assets of the issuer. 10 Covered bonds are generally issued by banks and other credit institutions. The international market for these bonds is currently centred in Europe, and is substantial. The European Central Bank (ECB) reports that the amount of covered bonds outstanding at the end of 2008 was €2.4 trillion.1 11 In the European market initial issuances tend to be large (around €1 billion), with a certain degree of repeat issuance generally required to provide the necessary liquidity to attract a wide range of investors. However, smaller issuance can be possible, particularly through private placements. 12 The primary attraction for issuers entering the covered bond market is the opportunity to gain access to relatively cheap longer term funding. Spreads between covered bonds and senior unsecured debts vary as market conditions change, but recent experience from overseas suggests that a double-A rated issuer might reasonably expect to save up to 50 basis points by developing a triple-A rated covered bond. Question:
1 Will financial institutions in New Zealand be able to generate savings in line with overseas experience? 1 European Central Bank, “Financial integration in Europe April 2010”, www.ecb.int/pub/pdf/other/financialintegrationineurope201004en.pdf

Ref #4206354
13 The creation and expansion of covered bonds has differed between countries, with different structures and features adopted. All covered bonds are designed to achieve the same overall outcome, and whilst there are a number of different terminologies used to describe bonds in different jurisdictions, at a high-level there are two distinct types of covered bonds. These are:
i. Legislative covered bonds, which are created under specific legislation that
ensures that bond holders have a priority claim on assets in the cover pool; and
ii. Structured covered bonds, where legal arrangements such as contracts, trust
arrangements, limited liability partnerships and company structures are used on a case-by-case basis to segregate assets. 14 Many jurisdictions have put in place the necessary legal frameworks to support the issuance of legislative covered bonds. This is especially so in Europe, where legislatively backed covered bonds tend to be favoured in order to comply with the directive on Undertakings for Collective Investment in Transferable Securities (UCITS). 15 To be UCITS-compliant, a bond must meet a number of requirements relating to the nature of the issuing institution, the supervision of the issuer, and the associated regulatory framework.2 Meeting these requirements generates a number of benefits, for example, it allows UCITS investors and certain insurance firms to increase the percentage of their assets invested in compliant covered bonds issued by a single issuer, and allows credit institution investors to reduce risk-weightings. 16 Outside Europe, Canada and the USA are looking to shift to a legislative framework. In Australia, APRA has prohibited Authorised Deposit-taking Institutions (ADIs) from issuing covered bonds, as they are considered inconsistent with the provisions of the Banking Act 1959.3 The Banking Act requires that an ADI’s assets in Australia are available to meet the ADI’s deposit liabilities in Australia prior to all its other liabilities, whereas covered bonds give covered bond holders priority over certain assets. 2 Full details of requirements see Council Directive 85/611/EEC. 3 See ADI Prudential Standard APS 120 Securitisation

Ref #4206354
3 COVERED BONDS IN NEW ZEALAND
3.1 Status Quo
17 At present, there is nothing in the Reserve Bank’s prudential requirements or New Zealand law prohibiting or limiting the issue of covered bonds. While this means there is no legal impediment to issuance, there is also no legislative support for an issue beyond existing contract law. As a result, under the current arrangements, any covered bond issued in New Zealand is a ‘structured’ covered bond. 18 The Reserve Bank had preliminary discussions about the issuance of covered bonds and other securitisation structures with a number of banks in 2008. At that time, we indicated privately to banks that we would be content for them to issue covered bonds at conservative levels. This guidance was intended to allow banks to explore opportunities for new funding, at modest levels, whilst the full implications of covered bonds were analysed. 19 The Reserve Bank is conscious that the majority of large banks in New Zealand are owned by parents that operate in a jurisdiction in which covered bonds are explicitly prohibited. As a result, we considered that it would be prudent to provide a public statement to confirm that the Reserve Bank was supportive of New Zealand banks targeting new funding opportunities through the development of covered bond programmes. 20 BNZ became the first New Zealand bank to issue a covered bond in June 2010 when it issued a 425 million NZD-denominated bond into the domestic markets. We are aware that a number of other banks are in the process of developing their own programmes.
3.2 The development of New Zealand covered bonds
21 The development of a covered bond market for New Zealand banks could have benefits in terms of lowering the cost of long-term funding. However, these benefits have to be traded off against any potential negative repercussions of covered bonds in the event of a banking failure. 22 In carrying out its role as the prudential supervisor of registered banks, section 68 of the Reserve Bank of New Zealand Act 1989 (the Act) requires the Reserve Bank to exercise its powers for the purposes of:

• promoting the maintenance of a sound and efficient financial system; or
• avoiding significant damage to the financial system that could result from the
  failure of a registered bank.
  23 The pricing benefits associated with the issuance of covered bonds can be expected to have a positive impact on the efficiency of the New Zealand banking system. However, the Reserve Bank considers that covered bonds also potentially offer a number of other benefits for the wider financial system. 4 Reserve Bank of New Zealand , Financial Stability Report, May 2010

Ref #4206354
24 First, covered bond investors tend to be different from the type of investors that will generally provide funding to banks through the senior unsecured debt markets. As a result, issuing covered bonds can be expected to provide access to a new investor base, thus increasing the funding diversity of the financial system. 25 Second, the Reserve Bank has, for a number of years, been concerned about the over￾reliance of New Zealand banks on short-term funding. This funding carries significant refinancing risk, and increases the vulnerability of the financial system to downturns. In light of these concerns, the Reserve Bank recently introduced new liquidity requirements for registered banks. One of the key purposes of these requirements is to lengthen banks’ funding profiles. The Reserve Bank is conscious that there is a limit to the extent that retail deposits can grow to meet these requirements and, as a result, banks will need to target additional long-term wholesale funding. The development of a covered bond market will provide an opportunity to meet these requirements. 26 Third, whilst large scale issues might be expected to be placed in offshore markets, the recent BNZ issuance has demonstrated the potential for a domestic market to develop on a smaller scale. This represents a new product for New Zealand capital markets and as such, is consistent with the recommendations of the Capital Markets Task Force report.

27 These benefits provide strong arguments in favour of the development of covered bonds as a new funding source for New Zealand banks. However, there are also a number of negative consequences that could arise from covered bonds. Whilst we do not consider that any of these represent a compelling argument for prohibiting the issuance of covered bonds outright, they are nevertheless important factors that will need to be taken into account when considering whether there needs to be an upper limit on the amount of covered bonds that a bank can issue. 28 First, the issuing bank must maintain the volume and quality of assets within the cover pool, which can potentially have a negative impact on its rating, particularly during a downturn. Rating agencies have indicated that at low levels covered bonds will be treated as a positive from a ratings point of view, as they imply a more diverse funding structure. However, as a bank commits more of its balance sheet to a covered bond programme, the risk associated with the remaining assets increases, which will begin to have a negative impact on the bank’s underlying rating. 29 Second, by subordinating depositors and other creditors below covered bond holders, the value of their claims on the bank’s assets may be eroded. This effectively transfers risk to those investors who are least able to assess and bear that risk (retail depositors). Unless retail interest rates adjust to reflect this increased risk, this may also lead to an inappropriate distortion in the risk-reward trade-off faced by retail depositors. 30 Third, this transfer of risks to depositors could have an impact on the behaviour of depositors and other creditors. In particular, subordinating depositors and other creditors below covered bond holders and leaving them with a claim on a lower quality set of assets may increase the likelihood of a run should a bank find itself facing challenging market conditions. 5 See http://www.med.govt.nz/templates/MultipageDocumentTOC____42335.aspx

Ref #4206354
31 Fourth, ring-fencing a proportion of the bank’s (high-quality) assets in a covered bond structure has the effect of reducing the quality and quantity of assets available to a statutory manager in a failure. As a result, there is a risk that the remaining creditors suffer increased losses compared to a scenario in which all creditors are treated pari passu. This could potentially have a negative impact on investor confidence in the wider New Zealand financial system. 32 The Reserve Bank considers that these risks can be satisfactorily mitigated as long as covered bond issuance is restricted to a relatively conservative level. As a result, when analysed against its requirements under the Act, the Reserve Bank does not consider that banks should be prohibited from issuing covered bonds. The following section discusses the factors that are relevant to considering the appropriate constraint that should apply. Question:
2 Are there any over-riding factors that should preclude financial institutions in New Zealand from issuing covered bonds?

Ref #4206354
4 CONSTRAINTS ON COVERED BOND ISSUANCE
33 As noted in section 2, banks in many jurisdictions have now entered the covered bond market which has grown to around €2.4 trillion. There are, however, only a handful of jurisdictions that have applied formal regulatory limits on the proportion of their assets that each bank can commit to a covered bond programme. These jurisdictions, and their respective requirements, are summarised in table 1 below. 34 The countries that have chosen to place limits on issuance are generally those jurisdictions that have only recently seen the development of covered bond programmes within their jurisdictions. In jurisdictions with a long history of issuance, the wider financial system is likely to have developed around covered bonds.
Table 1: Examples of regulatory limits in overseas jurisdictions
Country Authority Requirements
Canada OSFI Covered bond issuance limited to 4% of total assets at time of issuance. If covered bond subsequently exceed 4%, OSFI must be notified and a plan produced to show how the issuer will eliminate the excess. Italy Bank of Italy Issuers must meet minimum total capital (9%) and tier 1 capital (6%) requirements to issue covered bonds. Limit increases for higher levels of capital from minimum of 25%, to 60% for higher capital ratios, and unlimited issuance allowed beyond 11% total capital and 7% tier 1 ratio. Limits based on eligible assets rather than total assets. UK FSA No longer has explicit limit. Issuers have to discuss all asset encumbrance with the FSA on a case-by-case basis in advance, including covered bonds, securitisations, securities finance transactions and repo financing. FSA can apply additional capital requirements or issue maximum constraints on issuance. US FDIC Covered bond issuance limited to 4% of total liabilities at the time of issuance. 35 One option for New Zealand would be to simply allow the market to constrain issuance. As outlined above, from a credit rating perspective, covered bonds are generally considered to be a positive factor at low levels due to the increased funding diversity that they offer. However, the credit rating of the issuer is focused on the risk faced by investors in the senior unsecured debt. As covered bond issuance rises, the quality of assets remaining to support the interests of those investors is diminished. At a certain point, this negative impact will begin to outweigh the benefits associated with greater funding diversity.

Ref #4206354
36 Rating agencies do not have formal upper limits on the amount that a bank can securitise its assets for any given credit rating. It is our understanding that the level at which the underlying rating of the issuer bank might be at risk tends to vary depending on the business model of the institution in question. However, it would seem likely that ratings of large New Zealand banks might be at risk if overall asset encumbrance were to approach 20 percent of total assets.

37 Simply relying on this market ‘limit’ to constrain issuance to a reasonable level could have a number of positive implications. These might include:

• allowing banks the flexibility to maximise their use of all available instruments to
  adopt the most efficient financial structure given their individual business models;
• protecting competitive neutrality by ensuring that regulations do not preclude
  smaller institutions from accessing new funding streams; and
• minimising compliance costs.
  38 Our understanding is that the rating agencies’ notional limits refer to all products or instruments that result in any asset being encumbered. In practice, this can include a very wide range of instruments including the collateral requirements linked to swaps and derivatives, and exchange rate hedges. Some of these instruments can create volatility in the level of asset encumbrance that a bank will have at any given time. As a result, banks are likely to wish to maintain a reasonable degree of headroom between the level of assets that they encumber in stable, long-term funding programmes such as covered bonds and RMBS, and the upper limit of encumbrance that can be sustained at their desired rating. Question:
  3 In the absence of regulatory constraints, what is the maximum level of covered bond and RMBS issuance that a New Zealand institution might be expected to adopt? 39 The Reserve Bank believes it would be prudent if banks did not issue covered bonds up to the maximum level possible during benign times. First, during a crisis situation some markets may perform better than others. During the recent financial crisis, covered bonds remained an important funding tool for many institutions. Whilst, for a period, this was due to the ECB’s willingness to accept covered bonds, there is also some evidence to suggest that the covered bond market recovered more quickly than others. Although there is no certainty as to how future crises may unfold, from a stability perspective it is desirable for New Zealand banks to be in a position to access as many funding sources as possible at any given time. 6 Moody’s Investors Service, Australian Banking System Outlook – What’s Ahead for 2010?, March 2010

Ref #4206354
40 Second, during the recent financial crisis, many New Zealand banks developed RMBS programmes in order to gain access to the Reserve Bank’s liquidity facilities. Our understanding is that these programmes would be included in an assessment of total encumbered assets by a rating agency. In the event of future crises, this option is unlikely to be desirable if the additional asset encumbrance at the margin generates risk of a rating downgrade. 41 Third, there is uncertainty surrounding precisely how rating agencies would respond to increasing issuance from a New Zealand institution. Ratings are based on many factors, and the 20 percent ‘limit’ represents a level at which the agencies might begin to factor asset encumbrance into their considerations. As such, there is no guarantee that a bank, during favourable market conditions, would not be able to issue long-term funding instruments up to, and beyond, this level and retain its rating. This may be particularly relevant to the large New Zealand banks whose ratings benefit from implicit support of parents that are currently precluded from issuing covered bonds under Australian regulations.7 42 Given these concerns, and the fact that covered bonds represent a new, largely untested product in New Zealand, the Reserve Bank considers that it would be prudent to adopt a relatively conservative initial regulatory limit. As such, we propose to limit the issuance of covered bonds to 10 percent of total assets. The Reserve Bank will review the appropriateness of this limit within 2 years pending an assessment of market developments. 43 Given the need for over-collateralisation and the concerns surrounding the dilutionary effect of encumbering the balance sheet, the Reserve Bank considers that this limit should be applied to the value of assets held within the structure, rather than the value of the securities sold to the market. We consider that this approach strikes the right balance between allowing banks to access cheaper forms of long-term funding, whilst protecting the interests of other creditors, protecting the wider financial system, and retaining headroom to respond to crises. Questions:
4 Is a regulatory limit on covered bonds appropriate?
5 Is 10 percent of total assets an appropriate initial limit for the issuance of covered bonds? 7 The Australian Banking Act 1959 requires that, if a bank is unable to meet its obligations or suspends payment, its assets in Australia are to be available to meet its deposit liabilities in Australia prior to all other liabilities. As covered bonds subordinate depositors’ interests, the Australian Prudential Regulation Authority considers them to be inconsistent with the depositor-protection provisions, and therefore Prudential Standard APS 120 prohibits the issuance of covered bonds by Australian banks.

Ref #4206354
44 The Reserve Bank may consider allowing a higher maximum for a smaller institution where the proposed limit precludes it from entering the market. Any such relaxations would only be granted if the Reserve Bank was satisfied that the bank’s proposals were sufficiently robust. Amongst other things, this might be expected to include consideration of the institution’s capital adequacy, and the overall level of asset encumbrance associated with its existing business profile. As noted above, RMBS programmes are likely to be included in any assessment of asset encumbrance by ratings agencies. As a result, if a bank was to fully exploit the market’s appetite for RMBS in benign times, they would raise many of the same concerns as covered bonds. In principle, there may therefore be a case for introducing a wider limit that covers issuance of both RMBS and covered bonds. 45 In assessing the appropriateness of such an approach, there are a number of factors that need to be considered. First, an effective RMBS structure should result in the transfer of risk away from the issuing bank to other parties, and as such may benefit the stability of the New Zealand financial system.8 Second, during the recent financial crisis, a number of banks used RMBS programmes to access Reserve Bank liquidity facilities; as such it may not be desirable to introduce regulatory limits that would constrain such a response to future crises. Question:
6 Would a regulatory limit on some, or all, forms of RMBS issuance be appropriate? 8 The extent to which this transfer of risk occurs will be dependent on extent to which the bank creates a degree of separation between itself and the structure, otherwise it may, for reputational reasons, feel some “moral” obligation to provide support to an RMBS structure.

Ref #4206354
5 LEGISLATIVE FRAMEWORK
5.1 The need for legislation
46 As noted in section 2, there are two distinct types of covered bonds. These are legislatively backed covered bonds (LCBs) and structured covered bonds (SCBs). Whilst both are designed to achieve the same purpose, SCBs rely on contractual arrangements to protect the rights of bond holders, whereas under an LCB their interests are enshrined in legislation. 47 Within these two broad options, we believe that there is a spectrum of approaches that could be implemented. These are outlined in figure 1 below.
Figure 1: Options for high-level framework
48 A number of factors will need to be considered in determining which of these options represents the appropriate approach for New Zealand. They include:

• Access to potential investors: A legislative solution can be expected to allow New
  Zealand issuers to access a wider range of investors. Discussions with market participants suggest that there are overseas investors that would be willing to hold SCBs issued by New Zealand institutions. However, there are investors that are only willing to invest in legislatively backed bonds, or bonds that have ECB eligibility. Whilst New Zealand covered bonds will not be acceptable to all of

Ref #4206354 these investors, legislative backing may be expected to provide access to a sub-set of them. There also appears to be a trend for jurisdictions to move towards a legislative framework. If the majority of other non-European countries were issuing LCBs, the demand for New Zealand issued SCBs is likely to be reduced. As a result, the benefits of legislation may, in part, be linked to the development of covered bond markets overseas. The extent to which a legislative framework will open up new investor pools is likely to depend upon the specific nature of the framework. Whilst New Zealand bonds will not be UCITS compliant, a regime which effectively replicates those requirements may be seen as a positive.

• Impact on credit rating assessments: The legal and regulatory framework
  underpinning the bond is one of the factors that rating agencies consider when assigning ratings to bond programmes. The precise implications vary from agency to agency, and it is our understanding that some agencies are unlikely to apply significant weight to any new framework in the short-term. However, as the market develops, a robust legislative framework is likely to be viewed as a strong positive from a ratings perspective.
• Pricing versus senior, unsecured debt: The impact of the regulatory framework
  on pricing for New Zealand covered bonds is less clear. However, international benchmarks suggest that a robust legislative framework can result in an LCB generating additional savings of 5-10 basis points compared to an SCB with the same credit rating.
• Financial stability: As discussed above, an LCB framework might be expected to
  provide access to a wider range of investors, which may have a positive impact on system-wide stability. The other potential benefit from a systemic perspective is the removal of the risk that the structure of an SCB proves not to be robust in the event of an issuer default. Whilst New Zealand’s statutory management provisions create challenges in designing the SCB structure to ensure that the bond holders’ claims on the cover pool are bankruptcy remote, it should nevertheless be possible to devise a suitable structure. However, a legislative framework could be beneficial in removing any remaining uncertainty. Alternatively, it may be possible to mitigate some of this uncertainty by creating a standardised ‘NZ bond’ structure, which is adopted by all New Zealand issuers. 49 On balance, the Reserve Bank considers that there are likely to be material benefits from introducing legislative backing for covered bonds in New Zealand. The extent to which these benefits can be maximised will depend on the design of the legislative solution. Various design issues are discussed in the following sections. Question:
  7 Will legislative support for covered bonds generate material benefits for the New Zealand financial system?

Ref #4206354
5.2 Proposals
50 As noted above, the Reserve Bank’s initial view is that a legislative framework should be developed. Figure 1 identified a spectrum of high-level options that could be implemented, although in practice there are a number of more granular factors to consider in establishing the precise design of the framework. The following sections provide more detail on these factors.
5.2.1 Registration
51 Registration would involve the recognition of a covered bond issue by a New Zealand issuer, rather than authorisation of an issue. The effect would be that the cover assets are explicitly protected from the insolvency or statutory management of the issuer. 52 If the framework was to include registration, we envisage that the process for registration would be set out in regulations, which would specify the requirements for admission to the register. Such requirements could involve the provision of information to the Reserve Bank, including the background of the issuer, details of the proposed issue, and independent verification of the transaction documents. The Reserve Bank must be satisfied that the proposals seek to ensure that the cover assets are remote from the issuer. 53 There may be a fee for use of the register, on a cost-recovery basis. Unregistered covered bonds may still be issued, but will not attract the protection conferred by registration.
5.2.2 Structure
54 At a high-level, there are two different ways to segregate the assets of the cover pool from the assets of the issuer. The integrated model legislatively “ring-fences” the cover pool assets, which remain on the issuer’s balance sheet. The SPV model transfers the cover assets from the issuer to a separate legal entity (a special purpose vehicle - SPV) via a “true sale”. The SPV grants security over the cover assets to a trustee in favour of the covered bond holders. 55 The SPV model is based on robust and recognised legal principles, and is used in existing New Zealand structured covered bonds, whereas the integrated model poses some implementation challenges in the context of New Zealand insolvency law. The SPV model is therefore the preferred structure, and the Reserve Bank intends to restrict registration to covered bonds using the SPV structure. This approach supports the existing arrangements, and allows for the grandfathering of existing covered bonds. 56 A possible extension of this approach would be for the Reserve Bank to specify a single structure for registered covered bonds. The industry could design a single structure that could be adopted and marketed as a “New Zealand covered bond”. For an upfront cost, such an approach might be expected to generate a number of benefits. First, it might result in cheaper covered bond issuance over time, as issuers would avoid some of the costs of developing an issue. Second, there may be additional benefits associated with offering, and marketing, a standardised product to investors. Third, if the ‘safe harbour’ approach were to be adopted, a single agreed structure might provide investors with the same certainty as a formal registration process.

Questions:
8 Is it appropriate to require all covered bonds to adopt the SPV structure? 9 Would it be beneficial to develop a single defined structure for a New Zealand registered covered bond?
5.2.3 Cover pool assets
57 Market constraints are likely to have a strong bearing on the assets that covered bond contracts will permit in the cover pool. However, a number of other jurisdictions specify those assets that are “eligible” to be transferred to the SPV and form part of the cover pool. Our understanding is that the inclusion of formal restrictions within the regulatory framework might provide additional comfort to investors regarding the overall quality of the structure, and may potentially be perceived as a positive factor in the credit rating process. 58 In addition, there are a number of restrictions placed on cover pool assets where an issuer seeks to include an issue in the Reserve Bank’s Domestic Markets’ operations. We anticipate that most banks will seek to ensure that any domestic NZD bond programmes are repo-eligible with the Reserve Bank. Furthermore, we consider that repo-eligibility is likely to have positive impact on secondary market liquidity. 59 In light of these factors, we propose to introduce minimum eligibility criteria for the assets that can be included in the cover pool of a registered covered bond. In the first instance, we consider that it would be appropriate to align these criteria with those currently used for the Reserve Bank’s Domestic Markets’ operations eligibility. The prudential requirements for a registered covered bond will be reviewed from time to time, and will not necessarily always align with repo-eligibility criteria. Question:
10 Are the existing domestic market’s operations criteria for asset eligibility an appropriate initial requirement for all registered covered bonds?
5.2.4 Role of the Reserve Bank
60 The role of the Reserve Bank will depend on the option adopted. If the framework includes a registration process, the Reserve Bank would maintain the register of covered bond issues. This would involve receiving applications for admission to the covered bond register, and satisfying ourselves that the application is consistent with requirements, including asset eligibility criteria. We would not be assessing individual For full details, see http://www.rbnz.govt.nz/markets-and-payments/domestic-markets/application-criteria-for-security-eligibility Ref #4206354

Ref #4206354 aspects of the issue, such as the issuer’s oversight of the covered bond, or the ability of the programme to make timely payment on the bonds. 61 Registration would not be a validation of any professional opinions provided to the Reserve Bank by the issuer; issuers and investors must satisfy themselves as to the enforceability of the documents. Furthermore, registration by the Reserve Bank would not mean that the bonds are backed by the New Zealand government. 62 Option 5 in Figure 1 would require the Reserve Bank to carry out additional supervision of each covered bond issue on an on-going basis. However, while the Reserve Bank is comfortable facilitating the development of the market, it is not our role to provide explicit support to investors by way of additional supervision. Issuers are likely to be registered banks and non-bank deposit takers, and as such are subject to general supervision and regulation by the Reserve Bank. As a result, we do not intend to take on a detailed, on-going supervisory role. Any further supervision will be a matter to be determined between the issuer and the investor. Question:
11 Will the absence of additional public supervision have negative consequences?
5.2.5 Legislative changes
63 The precise nature of the legislative changes required will depend on the chosen option. If a registration approach were adopted, the legislation would need to institute a new covered bonds register, and to provide for regulations setting out the process for registration, as well as the requirements for admission to the register. Under a ‘safe harbour’ approach, the legislation would need to provide for regulations that would set criteria to define a legislatively backed covered bond. 64 In order for legislation to provide additional comfort for covered bond investors, the cover pools associated with such covered bond issues need to be explicitly protected from the insolvency or statutory management of the issuer. 65 To achieve this, legislation could be introduced to ensure, upon the insolvency or statutory management of the issuer, that:

• the asset pool remains available for the benefit of the covered bond investors,
  rather than for the benefit of the creditors of the issuer;
• the asset pool is unaffected by the application of the liquidation, voluntary
  administration, or compromises provisions of the Companies Act 1993 to the issuer;
• the asset pool is unaffected by the application of the receivership provisions of the
  Receiverships Act 1993 to the issuer;
• the asset pool is unaffected by the application of the statutory management
  provisions of the Reserve Bank of New Zealand Act 1989 or the Corporations (Investigation and Management) Act 1989 to the issuer.

Ref #4206354

Ref #4206354
66 These changes are designed to formalise in legislation the rights of investors that should already be achieved through the contractual arrangements of the covered bond structure. As such, they are essentially designed to eliminate any remaining uncertainty. Under the registration approach, the investors’ rights to the asset pool of all registered covered bonds would be explicitly protected, whereas under the safe harbour approach investors would be relying on the structure being deemed to be consistent with the criteria at the point of failure. 67 In other jurisdictions there are significant differences in the extent to which further elements are prescribed by legislation. The extent to which additional prescription generates benefits is heavily dependent on the nature of the market in the relevant jurisdiction. 68 We consider that the primary challenge to developing robust covered bonds in the New Zealand market relates to refinancing risk. There are relatively few large institutions operating in New Zealand, and there is presently no secondary market for covered bonds. This creates uncertainty around the long-term future of the bond in the event of issuer default. These concerns may be mitigated over time as the market for covered bonds develops. Alternatively, we could seek to remove some of this uncertainty by developing a legislative process to support the on-going servicing of the bond. 69 In terms of limits on issuance, the Reserve Bank intends to impose the constraints via banks’ conditions of registration, rather than through legislation. Questions:
12 Would the ‘safe harbour’ approach provide sufficient certainty to investors? 13 Should we seek to introduce prescriptive arrangements that help to mitigate refinancing risk?
5.2.6 Timeline and transitional arrangements
70 The Reserve Bank intends to support the existing covered bond arrangements in implementing the regulatory arrangements. We also wish to allow the continued development of the covered bond market in the interim. 71 The first step would be to update banks’ conditions of registration to reflect the limits set out in section 4 of this consultation paper. The conditions of registration for all registered banks will be updated, allowing them to continue to issue covered bonds within formal limits. If a registration framework is adopted, we would seek to ensure that covered bond programmes that are established prior to the introduction of the register can be grandfathered in and added to the register, provided the bond structure is consistent with all requirements set out in the registration process. 72 Changing banks’ conditions of registration is a relatively short process, and we expect that these changes would be included with the other changes set to be made in December. The legislative changes required to institute the register will require more time, and we anticipate that legislation would be introduced later in 2011.

Ref #4206354
5.2.7 Preferred approach
73 As identified in section 5.2.4 above, the Reserve Bank does not consider that it would be appropriate for it to take on the on-going supervision role that would be required under option 5. As a result, it does not intend to proceed on that basis. 74 Introducing a registration framework (option 4) would involve additional development and implementation costs compared to the ‘safe harbour’ approach (option 3). We consider that this cost will be justifiable if a formal register is likely to result in sufficient additional certainty to enhance materially the stability benefits generated for the New Zealand financial system. The extent to which there is a difference between the two options will, in part, depend on the approach taken on a number of the other issues highlighted above.

Ref #4206354
6 CAPITAL ADEQUACY AND DISCLOSURE REQUIREMENTS
6.1 Capital adequacy
75 The capital adequacy rules for banks using the internal models based approach to Basel II are set out in BS2B of the Reserve Bank’s banking supervision handbook.10 Part 5 of BS2B sets out rules for the treatment of securitisation vehicles. Under these rules, banks are required, in most instances, to continue to hold capital against any assets held within securitisation schemes. This represents a conservative approach compared to many overseas jurisdictions, partly explaining why securitisation has not been a significant feature of the New Zealand financial system in the past. 76 In other jurisdictions, it has been assumed that transferring assets into an RMBS structure results in an effective transfer of the risk associated with those assets. This, together with the absence of any requirement to maintain the value of the assets supporting the structure, means that the issuer, in principle, has no on-going commitment to the security. 77 The nature of covered bonds raises some different issues. They are generally issued with an on-going requirement to maintain the value of the cover pool, and as such the risk remains much more explicitly with the issuer, hence the argument for having a consolidated capital treatment is even more compelling. 78 There is also a separate issue regarding the impact that encumbering a proportion of the bank’s better assets has on its overall capital requirement. To achieve a triple-A rating, a covered bond needs to both contain high-quality assets, and include a degree of over￾collateralisation. This over-collateralisation effectively represents the capital within the covered bond structure. 79 This transfer of high-quality assets to the cover pool reduces the quality of assets in the underlying bank. In theory, if one were to analyse these assets on a stand-alone basis, they would incur a higher capital charge on average than the original consolidated asset base. As a result, assessing the issuing bank and covered bond separately might result in an increased capital requirement. Such an outcome would be consistent with achieving a financing structure that maintains the original credit rating of the issuing bank, whilst accessing a portion of funds at a higher rating. In principle, it could be argued that a higher minimum capital requirement should apply for banks that issue covered bonds 80 However, in practice, the extent to which covered bond issuance impacts on the relative riskiness of the remaining assets of the issuing bank is primarily dependent on the level of issuance. As a result, the Reserve Bank does not consider that it is necessary to apply additional capital requirements given the proposal to adopt a conservative upper limit on issuance. Question:
14 Should capital adequacy be assessed on a consolidated basis?
10 See http://www.rbnz.govt.nz/regulation-and-supervision/banks/banking-supervision-handbook

Ref #4206354
6.2 Disclosure
81 The issuance of covered bonds creates a prior claim over the assets of the bank, and as such they have the potential to make the other liabilities of a bank riskier. Whilst the Reserve Bank does not intend to force banks to hold additional capital to mitigate this impact, it does consider that creditors and depositors should have access to sufficient information to fully understand the implications of any bond issuance on the quality of their own claims on the institution. 82 The Reserve Bank is currently undertaking a broad review of its disclosure regime for registered banks.11 The appropriate approach for disclosing the impact of covered bond issuance will be considered further following the completion of this review and will be aligned with the overall philosophy adopted. 11See http://www.rbnz.govt.nz/news/2010/08/disclosure-review-to-reduce-compliance-costs