Response to submissions on the new macroprudential framework

Response to submissions received for the consultation on macro-prudential policy instruments and framework for New Zealand May 2013

2 1.0 Introduction

1. In March 2013 the Reserve Bank of New Zealand issued a consultation document on macro-prudential policy. The consultation document described and sought feedback on the full package of macro-prudential instruments, including the proposed framework for their use.1 In addition to outlining the proposed instruments, the paper outlined the objectives of macro-prudential policy and the proposed decision-making, governance and accountability framework for the conduct of macro-prudential policy.
2. Final submissions on the draft macro-prudential framework closed on 17 April 2013. The Reserve Bank received 27 submissions, including seven from banks, seven from individuals and the remainder from a range of institutions and associations.
3. This document summarises the main issues raised in submissions and provides the Reserve Bank’s response. The submissions are set out in line with the responses received to the questions posed in Section 6.0 of the consultation paper. 2.0 The objectives of macro-prudential policy Submissions received
4. Submissions were broadly supportive of the need for macro-prudential policy to support the stability of the financial system. While submitters typically did not differentiate between the financial stability goals of increased resilience and mitigating extremes in the credit cycle, one submitter argued that minimising systemic risk should be the objective rather than managing the credit cycle. A few submitters argued that the existing prudential regime and moral suasion are already sufficient to manage any excessive credit and asset price growth.
5. Some submitters argued that macro-prudential tools should also be deployed for broader purposes, such as price stability, exchange rate management, or housing affordability.
6. Several submissions also questioned the relationship between macro-prudential policy and the Reserve Bank’s review of bank capital adequacy requirements for housing loans. Reserve Bank’s response
7. The Reserve Bank agrees that the baseline micro-prudential regime is sufficient to manage risks associated with the normal ups and down of the economic and financial cycles. However, macro-prudential policy recognises that a dynamic approach to prudential regulation may be required to address risks arising from procyclicality in the financial system and the tendency for credit and asset price cycles to become mutually reinforcing. Such an approach may also be required to manage serious disjunctures between the New Zealand financial cycle and the international cycle, when, for example, there are major changes in international funding market conditions.
8. It should be noted that under the provisions of the Reserve Bank of New Zealand Act 1989, prudential tools must be deployed for the purposes of promoting financial system soundness and efficiency. It is expected that, in most circumstances, measures undertaken to meet macro-prudential objectives will provide useful support for monetary policy in its role of maintaining price stability, and can be pursued for both purposes if that is the case.

1 Note that the countercyclical capital buffer (CCB) framework was finalised following a consultation in March/April 2012

3 However, the Reserve Bank does not have the powers to deploy prudential instruments to meet price stability or exchange rate management purposes alone. 9. While the Reserve Bank acknowledges the importance of broader government policy goals, such as enhancing housing affordability, it does not have a mandate to pursue these. We note however, that to the extent macro-prudential tools are successful in dampening excessive housing credit and house price cycles, they may indirectly contribute to such objectives, by helping to keep future house price growth more in line with economic fundamentals such as income growth. 3.0 The macro-prudential policy instrument set Submissions received 10. Submissions were broadly supportive of the proposed instrument set: • The countercyclical capital buffer (CCB) • Adjustments to the core funding ratio (CFR) • Adjustments to sectoral capital requirements (SCR) • Quantitative restrictions on the share of high loan-to-value ratio (LVR) loans to the residential property sector. 11. One submitter argued that broad-based macro-prudential interventions risk punishing banks that are not driving the risky behaviour. With regard to the CFR, a bank argued that the CFR should only be used in times of broad-based and strong credit growth, as any additional funding costs would impact other sectors of the economy that might not be contributing to stresses in the cycle. 12. Two submitters noted that the SCR would have proportionately more impact on banks whose business is focused on particular sectors (e.g. housing). It was argued that this could lead to reduced competitiveness of the banking activities in that particular sector. One submitter expressed a preference for SCRs to be imposed via an overlay rather than by adjusting sectoral risk weights used to calculate capital requirements. 13. While many submitters saw value in including LVR restrictions in the toolkit, the need to balance distributional and efficiency considerations was often raised (refer section on LVR restrictions for further details). A few submitters argued that LVR restrictions could not be justified on a cost-benefit basis. 14. Three submitters noted their preference for LVR ‘speed limits’ (which constrain the share of high-LVR lending) to outright prohibitions. It was argued that speed limits should apply to new lending only, as applying them to the banks’ existing mortgage portfolios could result in significant gaming should LVR definitions not be consistently understood or applied across banks. 15. A number of submissions suggested that the Reserve Bank should consider including a debt service coverage measure in the macro-prudential toolkit, either in place of, or in combination with, an LVR restriction. Reserve Bank’s response 16. The systemic nature of the risks that macro-prudential tools are designed to address means that their implementation needs to be system-wide. That being said, if risky behaviour was limited to a minority of banks, this could well be expected to attract a supervisory

4 response (i.e. measures focused on the bank(s) in question) rather than a macro-prudential one. An assessment of the impact of macro-prudential measures on various sectors will be part of the cost-benefit process. 17. The Reserve Bank recognises that banks that are highly concentrated in particular lending sectors are likely to be more affected by measures that target lending to the sector in question than banks with a more diversified lending portfolio. However, this is not a compelling argument for regulatory forbearance in the face of rising financial system stability risks. 18. Sectoral capital requirements would typically be applied through a capital overlay that is calibrated as a proportion of banks’ risk-weighted exposures to the sector, such as for housing lending or agricultural lending. In some cases however, where the aim was to target a lending segment such as high-LVR housing loans, it may be more appropriate to apply them via temporary adjustments to risk weights. 19. The Reserve Bank concurs with the view that LVR speed limits are a potentially less costly way of implementing LVR restrictions, as they would allow banks to continue some high-LVR lending to creditworthy borrowers. If we were to implement LVR restrictions, we would favour this approach in the first instance, but we would not rule out the use of outright limits should this prove ineffective. In terms of the gaming risks around the use of LVR restrictions, the Reserve Bank acknowledges the need for clear definitions of LVRs, robust statistical collections of LVR stocks and flows, and careful specification of any restrictions that are applied. A process is underway to improve the quality and range of LVR data collected from the banking sector. Our intention is that LVR speed limits would only be applied to new high-LVR residential mortgage lending. 20. The Reserve Bank agrees that debt servicing is an important factor in assessing borrower vulnerability and we would not rule out the possibility of incorporating a debt servicing ratio restriction as part of the macro-prudential policy toolkit in the future. 4.0 The impact of macro-prudential requirements on banks’ funding, pricing and balance sheet management Submissions received Adjustments to core funding ratio 21. Some submissions raised the risk that the measure would have limited impact because of banks’ existing voluntary funding buffers. It was suggested that banks might look to absorb some or all of a temporary minimum CFR increase by reducing their existing voluntary buffers. Overall, however, responses suggested that banks would generally choose not to absorb a temporary increase in the CFR by reducing buffers. 22. Submissions were mixed on whether banks would meet a requirement for additional core funding by increasing deposits or increasing wholesale term funding. Those banks that were reliant on deposits suggested this would likely lead to increased competition for deposits and higher deposit rates. 23. The banks’ submissions suggested that an increase in the CFR could have some impact on the cost of funding, but at times it might be very small. One bank said a five percentage point increase in the CFR would result in an overall cost of funding impact of less

5 than three basis points. Some other banks’ responses implied the effect could be bigger, through a rising retail deposit spread. 24. Banks noted that the impact on loan pricing would depend on a number of factors. Return on Equity (ROE) and cash earnings targets could mean that higher funding costs are passed on, but the competitive environment would be critical in deciding this. 25. One bank questioned the mechanism by which a reduction in the CFR might be achieved and the timing around which such a reduction would actually come into effect in the market. The underlying concern appeared to be that it might be difficult to achieve a swift reduction in the CFR given the sticky nature of term funding. The CFR was characterised as an asymmetric tool in this regard. 26. A number of potential unintended consequences of higher CFR requirements were raised, including: • Unintentional limits on the deposits of large fund managers (such as Kiwisaver pools), should banks decide to switch out of large deposit funding – due to its lower CFR weighting – towards term funding; • A further reduction in the use of short-term wholesale funding, with the risk that some banks might close their US Commercial Paper programmes. This could put access to this source of funding at risk in a stress scenario; • Banks reaching their covered bond limits sooner than would otherwise be the case. Countercyclical capital buffer 27. Again, submitters commented on the possibility that the imposition of a CCB might be partly or fully absorbed by reduced voluntary buffers. However, some banks noted that internal policies on minimum capital ratio settings would rule out such a response. Overall, it appears that banks would look to raise additional capital in response to use of the CCB. 28. The submissions suggested that the impact of the CCB on pricing would again be uncertain: • Submitters noted that ROE and cash earnings targets could drive increases in the price of lending, as could capital market conditions (constrained access and/or expensive pricing). • It was noted that pricing decisions would partly depend on whether capital is allocated on an economic capital basis or regulatory capital basis. • Submitters noted that the competitive environment would remain a key consideration behind pricing, with banks keenly aware of protecting their market share. • Two submitters stated that should a CCB be imposed, the price increase would likely be achieved via increased low equity fees rather than through higher interest rates. Sectoral capital requirements 29. Submissions suggested that an increase in funding costs due to SCRs would be more likely to be passed on than the CCB, reflecting the role of the banks’ internal pricing models. Banks noted however, that this might not necessarily be passed on to the lending sector in question but could take the form of an across the board increase in interest rates, or higher pricing in another lending sector. Alternatively, it was suggested a bank might choose to absorb the cost of the extra capital to protect market share. Again, two banks said the cost would probably be passed on via low equity fees rather than through interest rate rises.

6 30. One bank stated that a change in SCRs would advantage standardised banks, as they would not be directly affected. This likely reflects a misunderstanding, as SCRs would apply to all banks, not just the internal model banks. General comments 31. Overall, submissions suggested that use of the CFR, CCB or SCR could result in higher lending rates but the linkages are unpredictable and can be highly context￾dependent: e.g. abundant wholesale funding supply could negate CFR impact and when competition is strong there might be little pass-through to prices. Reserve Bank’s response 32. The submissions provided useful insights on the determinants of banks’ pricing and lending policies, and their potential influence on the impact of a macro-prudential policy intervention. The Reserve Bank expects to use this information in its instrument design and modelling work. 33. The range of levers that can affect banks’ asset pricing and allocation decisions creates substantial uncertainty around the likely impact of the tools at any time. This underscores the importance of detailed cost-benefit analysis at the time that an intervention is being considered, which can incorporate current market conditions, competitive considerations, the state of banks’ balance sheets and other relevant information. 34. Overall, it appears that macro-prudential tools – apart from LVR restrictions – may only have a weak impact on dampening the credit cycle. However, it bears emphasising that the primary purpose of such tools is expected to be to build additional resilience in the banking system rather than exerting a strong dampening effect on credit supply and demand. This is in line with International Monetary Fund (IMF) findings on the effectiveness of such instruments. We would also note that the work of the Basel Committee on Banking Supervision points to the role of the CCB as being primarily for the purpose of building loss absorbency. 34. Regarding release of the CFR, banks would immediately be free to transition to the new CFR minimum – how quickly this occurred would depend on the structure of their funding and balance sheet growth. We would also point out that, in proposing a CFR as a macro-prudential tool, our focus has been on scenarios that might warrant temporary reductions in the CFR as well as those that might warrant temporary increases. A temporary reduction might be necessary in response to severe stresses in global funding markets. In that instance, lowering the CFR would provide a safety valve that would allow banks to access short-term funding in the event of disruptions to term funding markets. 5.0 Operational considerations relating to the imposition of sectoral capital requirements Submissions received 35. One bank raised the possibility of targeting SCRs to particular regions; e.g., high-LVR Auckland lending. The bank described this approach as feasible but noted that it would require some changes to its internal systems.

7 36. A bank noted that, where a SCR results in an internal decision to modify lending practices, this would require a lead time in order to adjust internal pricing models, update staff training and procedure manuals and prepare suitable communications for customers. Reserve Bank’s response 37. Sectoral capital requirements can be targeted to particular exposures: e.g. housing lending, agricultural lending, or to a particular lending segment within a sector. The latter could be high-LVR lending, or as suggested by one submitter, high-LVR Auckland lending. However, while theoretically possible, the Reserve Bank does not currently envisage the use of SCRs at a regional level as part of its base framework. Such a requirement would raise many of the same issues around regional targeting of LVR restrictions that are discussed below. 6.0 The impact of restrictions on high loan-to-value ratio mortgage lending Submissions received 38. The potential impact on first-home buyers (FHBs) was a frequent theme, with a number of submitters favouring targeted implementation (e.g. FHB exemptions, targeting of housing investors). Some submitters cited examples from other jurisdictions that have used targeted LVR caps. A few submissions also called for high-LVR lending that is backed by lenders’ mortgage insurance to be exempt from any restrictions. 39. Some submitters linked LVRs to current housing market risks, and argued that these are highly localised – particularly in Auckland and Christchurch. It was argued that broad￾based LVR restrictions could unfairly penalise buyers outside of these areas, with one submitter suggesting that it could spark price falls in regions where house prices are flat. The possibility of applying LVR caps on a regional basis was also raised by some submitters. 40. Other ideas for targeting included targeting LVR restrictions by property value (which is not uncommon in Asia) and exempting residential loans taken out to fund small businesses. 41. One submitter argued that “LVR caps may result in banks placing disproportionate focus on LVR at the expense of other credit risk management principles including affordability and willingness to pay”. 42. A couple of submitters argued that the drivers of house prices are much broader than credit availability, and that macro-prudential tools may not be able to effectively lean against these factors. 43. A number of submitters referred to the Supplementary Stabilisation Instruments (SSI) Report prepared jointly by the Reserve Bank and the Treasury in 2006, and the adverse findings at that time regarding LVR restrictions as a policy tool, including problems of targeting, timing and disintermediation.

8 Reserve Bank’s response 44. The Reserve Bank expects that LVR restrictions would typically be applied in a broad-based fashion. Setting exemptions would not necessarily be appropriate and could significantly dilute the effectiveness of the instrument. Although LVR restrictions overseas have sometimes exempted first-home buyers, this might only be desirable if first-home buyers were not driving a substantial part of the high-LVR borrowing. 45. The Reserve Bank, however, does not rule out the possibility of targeting LVR restrictions if risks are found to be significantly concentrated in particular segments of the housing market. However, the potential costs of a more targeted approach would include higher administrative costs and a higher probability of circumvention. In addition, targeted application of LVRs would require a more granular data on the pattern of LVRs than is currently available. The Reserve Bank is currently requesting a breakdown of high-LVR housing lending from the banks by borrower type that may be used to help assess the pattern of high risk borrowing in the housing sector, and the regulatory impact of high-LVR lending restrictions. 46. Targeting LVR restrictions to particular regions may be feasible but would entail significant practical difficulties and could create other distortions (e.g. shifting high-LVR lending to the boundary around which the restrictions were drawn). Implementing LVRs at a regional level would require a detailed analysis of the economic case, and collection of new data at a regional level. 47. The Reserve Bank proposes to exempt high-LVR lending under Housing New Zealand’s Mortgage Insurance Scheme (MIS), including the Welcome Home Loan scheme and Kainga Whenua programme. MIS loans serve clear government housing policy objectives and present minimal risks to financial stability, as any bank losses are either underwritten or guaranteed by Housing New Zealand. 48. At this stage, the Reserve Bank does not intend to grant exemptions to LVR restrictions for high-LVR loans that are insured using lenders mortgage insurance (LMI) provided from private insurers. LVR restrictions are designed to be used in times of rising systemic risk. In the event of a severe systemic shock, the resilience of private insurance can be severely tested, as was illustrated in the United States, with mortgage insurers not having adequate capital to meet claims. 49. The Reserve Bank agrees that LVRs are only one dimension of sound residential mortgage underwriting practices. We would expect banks to continue to assess borrowers across the range of dimensions, in particular the ability of the borrower to repay the loan. 50. The Reserve Bank notes that macro-prudential policy is not intended to supplant measures designed to address fundamental balances between housing demand and supply. Freeing up land for subdivisions, reducing the time and cost of housing developments and improving productivity in the building sector are likely to be key in reducing house price pressures emanating from excess demand for housing. However, in the interim, macro￾prudential policy may offer scope to make the financial system more resilient and help mitigate excesses in credit and asset price cycles that may flow from such imbalances. 51. The analysis in the 2006 SSI Report has been helpful in identifying some of the costs and risks of LVR restrictions. However, the report was written prior to the financial crisis, and with quite different objectives. Whereas the SSI Report was looking for tools to support monetary policy and reduce pressure on the exchange rate, the macro-prudential focus is on supporting financial stability. The Global Financial Crisis has also prompted greater acceptance of the wider use of these tools, particularly in advanced economies. There is a

9 growing body of post-crisis research by the IMF, the Bank for International Settlements and other international organisations which has informed our analysis and findings, and which has found that these tools can assist in promoting financial system resilience and, in some cases, in dampening the credit and asset price cycle. 7.0 The proposed decision-making framework for macro￾prudential policy Submissions received 52. Submissions that commented on the decision-making framework tended to describe it as appropriate at a high level. There were requests for further detail around the risk assessment process, and the calibration and timing of policy interventions. 53. One submitter suggested augmenting the risk assessment with a standard stress test exercise, while another suggested defining ex-ante criteria and trigger points for macro￾prudential intervention, and including them in an agreement similar to the Policy Targets Agreement used for monetary policy. 54. A number of submitters commented on the importance of a clear, well communicated framework. One bank noted that public communication early in the process could help to change behaviour without actual recourse to policy measures. Two submitters stressed the importance of accountability mechanisms, such as clear explanations of the reasons for macro-prudential intervention at the time of intervention, and regular reviews of the framework. 55. There was some comment on the roles of the baseline micro-prudential regime, macro-prudential policy and monetary policy, and how the Reserve Bank would co-ordinate decisions across these functions. One submitter suggested that “it would be beneficial for the RBNZ to elaborate on how it sees the proposed macro-prudential instruments interacting with monetary policy, micro-prudential policy and liquidity policy. It is important everyone understands these potential interactions, commonalities and displacements”. 56. Banks stressed the need for macro-prudential policy to be applied in a uniform manner. Reserve Bank’s response 57. The Reserve Bank agrees on the importance of clear communications and a transparent and well understood decision making process. The Reserve Bank intends to publish guidance on the indicators and judgements that underlie its macro-prudential policy decisions in its regular Financial Stability Reports (FSRs). Given the broad range of factors shaping financial system risk, the Reserve Bank does not believe it will be able to publish simple thresholds or trigger points for decisions. We are also aware that our capacity to pre￾specify these matters is constrained by the dynamic and innovative nature of the financial system, and the limited state of knowledge on macro-prudential policy (both with respect to risk assessment and tool effectiveness). This means there is a strong role for policymaker judgement in the process, as emphasised in the consultation documents. 58. The Reserve Bank, in conjunction with the banks, undertakes stress testing exercises to better understand the nature of emerging financial system risks. The Reserve Bank is continuing to develop its capacity in this area and that of the banks undertaking these stress tests. Stress testing forms an integral part of our risk assessment framework,

10 with results from the latest rounds of stress tests having been published (see “Box D Recent stress tests of the major banks”, November 2012 Financial Stability Report). 59. The Reserve Bank agrees that greater clarity around the co-ordination of the Reserve Bank’s various policy functions is important for the transparency of the decision￾making framework. The Memorandum of Understanding that has been signed between the Governor and the Minister of Finance sets out the accountability mechanisms. While the FSR will be a key channel for macro-prudential policy communications, other channels such as media releases may also be used. As agreed in the Memorandum of Understanding, there will be a review of the macro-prudential policy framework after five years. 8.0 Indicative notice periods for deployment of tools Proposed notice periods 60. The following table contains the proposed notice periods for imposition of a macro￾prudential requirement via banks’ Conditions of Registration (COR). Submissions received 61. There were mixed responses on the proposed notice periods – some submitters thought them appropriate, others too long, others too short. A number of submitters emphasised the need for timely responses, and the danger of borrowing and lending activity being brought forward to avoid macro-prudential restrictions. 62. Some banks argued that it would be difficult to raise capital within the prescribed notice period. One bank submitted that the lead time for instrument deployment meant they would not be able to be implemented in time to be effective. i.e. the length of time needed to raise capital would mean that by the time the capital requirement came into effect, the crisis might already have hit. 63. Submitters wanted clarification over the notice period for LVR restrictions. Several banks submitted that they would be unable to implement LVR restrictions within two weeks. They stated that bank-wide changes would be required to policies, processes, IT systems and resourcing and that these would not be possible within a two week timeframe. Reserve Bank’s response 64. The Reserve Bank’s proposed notice periods for each instrument are aimed at striking a balance between the need to act pre-emptively, and the need to provide a reasonable time for banks to meet the macro-prudential requirement. The mixed responses confirm the Reserve Bank’s view that this is an issue on which it is difficult to be definitive. The Reserve Bank intends to be flexible in its approach but confirms that the notice periods set out in paragraph 60 will apply. For the avoidance of doubt, it should be noted that the proposed Instrument Notice period Countercyclical capital buffers Up to twelve months Sectoral capital requirements Up to three months Adjustments to core funding ratio Up to six months Restrictions on high-LVR housing lending At least two weeks

11 notice period for LVR restrictions provides the Reserve Bank with scope to provide a notice period greater than two weeks should this appear necessary. 65. The need to provide a reasonable time for banks to respond will include consideration of the time needed for banks to raise extra capital. However, CCBs and SCRs that affect the conservation buffer can be implemented relatively quickly, since failure to fully meet these requirements will not constitute a breach of regulatory minima (rather it would result in restrictions on banks in making earnings distributions). The Reserve Bank acknowledges the lead times for formal imposition of the requirement, but expects that there will also be an announcement effect, which will see banks respond prior to the deadline. 66. The operational details of macro-prudential instruments will be included in the Banking Supervision Handbook (BSH) and, when individual tools are used, in the banks’ COR, with the normal consultation processes applying. It is expected that the technical implementation details for each of the instruments will be incorporated into the BSH over time. These changes will set out the base framework for each tool, which should assist banks to pre-position their procedures, policies and processes for tool deployment, and to train their staff. The lead time for such changes will in part reflect the time required for pre￾positioning. 67. Whereas the BSH will describe the base framework, the COR will specify the calibrations. This could be the size of a CFR increase or capital requirement, or high-LVR thresholds and quantitative limits. 9.0 Costs, benefits and risks Submissions received 68. A number of submitters noted the need for careful cost-benefit analysis, with some requesting further detailed work and a regulatory impact analysis. 69. Discussion of risks and costs was almost entirely focused on LVRs, and revolved around issues identified in the consultation paper and background paper such as avoidance, disintermediation and the potential for unintended consequences. Submitters specifically mentioned the risk of disintermediation to higher-risk financial intermediaries, offshore financial institutions and foreign bank branches. 70. Avoidance risks included the use of funding from family, friends or loan sharks, greater use of second mortgages and unsecured lending. Submitters raised the example of Sweden, where unsecured lending has risen since the introduction of LVR limits. One bank raised the risk that valuations might be gamed, arguing that “the New Zealand valuation system is not currently robust enough to support any form of LVR cap”. Banks stressed the importance of the Reserve Bank enforcing compliance, to ensure that all banks were equally affected. 71. A number of unintended consequences of LVR restrictions were cited. Submitters argued there would be a disproportionate impact on such groups as FHBs, low-income buyers, small business, immigrants and people re-entering the property market after a relationship breakdown. One submitter argued that LVR restrictions could lead to reduced supply of land and housing developments aimed at new entrants to the housing market. Another argued that banks already typically limit LVRs to 80 percent for investors so LVR restrictions would not be binding for investors. The potential adverse impact of LVR restrictions on the Christchurch rebuild was also raised. The potential for ‘trapped borrowers’ was raised by some banks, whereby high-LVR customers might not be able to switch banks.

12 72. On the CFR, one submitter argued that introducing the CFR has had the unintended consequence of disadvantaging those who raise funds from retail investors in favour of the banks, with the virtual disappearance of mortgage funds, cash and bond funds struggling to compete, and the demise of the retail corporate bond market. This was in contrast to another submission, which stated the proposed measures will support “the development of capital markets and direct investment into the productive / capital intensive sector”. 73. One submitter stressed that there are broader costs to financial system instability, in the form of a heightened risk of market misconduct, such as poor or inappropriate advice in the marketing of property, or misconduct on the part of directors of stressed financial institutions. The submitter noted that such instability also places stress on investors who rely on property and/or equities for their retirement saving. Reserve Bank’s response 74. The Reserve Bank recognises the importance of carefully assessing and weighing the costs and benefits of macro-prudential intervention. The consultation document and background paper provided a preliminary, high-level exploration of the potential transmission channels of macro-prudential policy, and associated costs and benefits. The consultation process has been very helpful in providing feedback from the banks on their likely responses, which we will look to incorporate into our existing analytical framework. 75. Where the Reserve Bank decides to actively investigate the possibility of macro￾prudential intervention, a case-specific assessment of the expected costs, benefits and risks will be made. Should a decision to intervene be made, the Reserve Bank would expect to publish a regulatory impact statement that would set out its assessment of the expected costs and benefits. Macro-prudential policy is a new field and we would expect international techniques and perspectives on the measurement of costs and benefits to develop over time as countries gain more experience in the use of such instruments. We expect to assimilate and learn from such practices in terms of our own framework. 76. The Reserve Bank acknowledges the risk of disintermediation, and would expect to closely monitor this risk when implementing macro-prudential tools. The risk of disintermediation may be mitigated to a degree by the use of macro-prudential tools only when they are required rather than as a permanent change to prudential settings. In addition, the current dominance of the banking sector in financial intermediation than has been the case historically may help to reduce scope for opportunistic lending by domestic non-bank lenders. Nevertheless, it will be important to carefully monitor and report developments in the non-regulated finance sector, including the activity of overseas lenders. The Reserve Bank intends to consider the case for extending the framework to non-bank lenders in due course. 77. It is recognised that borrowers may be able to avoid LVR restrictions by borrowing from unofficial or unregulated sources. Such borrowing would reduce the effectiveness of LVR restrictions in addressing excessive credit growth, but would present less risk to the resilience of the financial system, given that these lenders sit outside the ‘core’ system. 78. The Reserve Bank acknowledges the need for clear boundaries, so as to reduce the risk of restrictions being avoided through reclassification, gaming of valuations and other means. In designing the instruments, the Bank will carefully consider the definitions of key residential lending concepts such as ‘residential mortgage’, ‘origination event’, ‘valuation’, so as to reduce the potential for such gaming.

13 79. We also acknowledge that borrowers could resort to unsecured lending to make up any shortfall in deposit created by LVR restrictions. This could reduce the extent to which LVR restrictions act to curtail credit growth. However, unsecured lending would likely only be provided at a high price (reflecting the additional risk) which would help to dampen demand. Moreover, to the extent unsecured lending would allow banks to continue lending to borrowers with high cash flow but limited wealth, this would help to reduce the negative impact of LVR restrictions on otherwise creditworthy borrowers. 80. The Reserve Bank agrees the potential for some borrowers to become ‘trapped’ unless restrictions are designed in a way that prevents this. Ideally we would not want to unduly constrain such borrowers, and are exploring whether it would be possible to operate LVR restrictions in a way that allows refinancing of pre-existing high LVR loans, on the condition that the loan amount and LVR are not increased. The Bank is also investigating the feasibility of allowing borrowers to ‘port’ their LVR when selling and buying a home, again on the condition that the loan amount and LVR are not increased. 81. The Reserve Bank recognises the potential for unintended distributive and efficiency effects when imposing restrictions on high-LVR lending, and would carefully consider these issues in any decisions on using LVR restrictions. As discussed, targeted LVRs could help to mitigate these costs but would be practically difficult to implement. On balance, we believe the use of the ‘speed limit’ approach rather than outright caps on high-LVR lending would help to mitigate these costs by enabling banks to continue to undertake some high LVR lending to creditworthy borrowers. While the operational costs of such an approach would need to be considered, banks have expressed a preference for this approach. 10.0 Other Submissions received Other instruments 82. A few submitters suggested alternative tools that could be considered by policymakers to help support financial stability and address emerging risks. For example, one submission raises measures such as taxation, rules on New Zealand property ownership or regional stamp duty, while another submission revisits the possibility of a mortgage interest levy. Process issues 83. Some submitters considered the consultation period of six weeks was too short, partly because other Reserve Bank consultations were taking place at the same time. Some submitters also felt that insufficient detail was provided on the proposed framework and instruments and noted that they faced difficulties in analysing the impacts as a consequence. 84. Some banks noted it would be difficult to operationalise the proposed macro-prudential tools within the indicative notice periods, and requested that sufficient time be allowed for any such changes to be made.

14 Reserve Bank’s response Other instruments 85. The non-prudential measures noted above are outside the scope of this consultation, and the Reserve Bank’s powers. The Reserve Bank will, however, bring these suggestions to attention of the Treasury and other authorities. Process issues 86. The Reserve Bank acknowledges comments about the timeframe available for submissions and the view of some submitters that further details on the framework would have been desirable in the consultation papers. The consultation was deliberately kept high￾level in nature. We expect to undertake further engagement and consultation with the banks shortly around the operational design of the instruments. 87. Some preliminary workshops with the banking industry provided early opportunities for input, which have helped inform the consultation paper. Some bank submissions have provided very useful information on the likely transmission channels of macro-prudential policy, and we are now incorporating that into our framework design and modelling. 88. The next stage of the process will be consultation with industry on the proposed changes to the BSH, for each of the tools (excluding the CCB which has already been consulted on). 89. The Reserve Bank will also shortly be holding workshop consultations with the banks around new data collections on high LVR residential mortgage lending, which will provide a much more granular view of activity in the housing sector. The Reserve Bank looks forward to sharing this industry-level data once consistent, standardised data are available. 90. Should macro-prudential intervention be judged necessary at some point, the Reserve Bank expects to do this through banks’ COR. As part of this process, the Reserve Bank would publish the ‘settings’ for the particular intervention. This could be the size of a CFR increase or capital requirement, or high-LVR thresholds and quantitative limits for LVRs. For example, calibrations of any speed limit that might apply, such as a LVR threshold of 90 percent and maximum high-LVR lending share of 5 percent of new housing lending. 91. The Reserve Bank appreciates the banking industry’s co-operation to date, and looks forward to further productive engagement in operationalising the macro-prudential policy framework.

15 Glossary BSH – Banking Supervision Handbook CCB – Countercyclical Capital Buffer CFR – Core Funding Ratio COR – Conditions of Registration FSR – Financial Stability Report IMF – International Monetary Fund LMI – Lenders mortgage insurance LVR – Loan-to-value ratio RBNZ – Reserve Bank of New Zealand ROE – Return on Equity SCO – Sectoral capital overlay SCR – Sectoral capital requirement SRW – Sectoral risk weights