Regulatory Impact Assessment – Removal of Loan-to-Value Restrictions following the onset of COVID-19

Removal of Loan-to-Value Restrictions following the onset of COVID-19 REGULATORY IMPACT ASSESSMENT

1 Regulatory Impact Assessment – Removal of Loan-to-Value Restrictions following the onset of COVID-19 30 April 2020 Adequacy Assessment This Regulatory Impact Assessment (RIA) provides an analysis of the options by the Reserve Bank for Loan-to-Value Ratio (LVR) restrictions following the onset of the COVID￾19 pandemic in New Zealand and internationally. The RIA has been prepared by the Reserve Bank in accordance with the requirements of section 162AB of the Reserve Bank of New Zealand Act 1989 (the Act). This Assessment provides a qualitative assessment of all decisions, and a quantitative assessment where this is possible. This analysis has a number of limitations, which are particularly acute following the onset of COVID-19. The timing and duration of the health and economic consequences of COVID-19 are highly uncertain, as are the medium and long term responses of households and businesses. While this introduces significant uncertainty into the analysis in this RIA, the document identifies the key factors and assesses the likely impacts at an aggregate level, and uses this to identify the most suitable options. Consultation A short, seven day, consultation period was completed by the Reserve Bank following the announcement on 21 April of the proposal to temporarily remove LVR restrictions. Banks and the general public were provided the opportunity to respond to the consultation. A short consultation period was necessary in order to respond in a prompt way following an unprecedented set of economic events. The Reserve Bank also directly approached a number of NGOs, including Māori and Pasifika groups to make them aware of the proposed changes, and to provide them a chance to provide on the proposals. The Reserve Bank consulted Treasury and the Ministry of Business, Innovation and Employment (MBIE) during the preparation of this RIA. Quality Assurance The RIA was reviewed by a number of Reserve Bank staff.

2 Decision Summary

1. The Loan-to-Value Ratio (LVR) policy is aimed at reducing the risks to financial stability from a severe correction in house prices. The Reserve Bank introduced LVR restrictions in October 2013 in response to high levels of high-LVR mortgage lending, and has adjusted policy settings as risks have evolved.
2. LVR restrictions have been effective in supporting financial system stability during the upswing in the credit and housing cycles since the introduction of the restrictions in
3. By significantly improving the equity positions of mortgage borrowers, it is likely that a significantly smaller number of borrowers will have to sell their house or default on their mortgage as a result of the current economic shock. This reduces the likelihood of a large and disorderly fall in house prices as a result of large numbers of forced house sales.
4. The current economic conditions resulting from the COVID-19 crisis have led to the Reserve Bank considering the removal of LVR restrictions, given the counter-cyclical nature of the LVR policy. This action will also avoid any uncertainty around the implications of LVR limits from the mortgage deferral scheme. It is important that banks continue to provide support to borrowers during these extraordinary times and that they continue to provide access to credit for credit-worthy borrowers.
5. The Reserve Bank intends to remove the LVR restrictions for a period of one year, until 1 May 2021. The Reserve Bank has not yet determined what, if any, LVR limits will be needed in the future. This will be further considered ahead of the end of the 12 months, and will be consulted on as necessary. The decision to remove LVRs should not be seen as a longer-term assessment of their role, but as a consequence of the current set of conditions faced by the New Zealand economy.
6. LVRs operate in a broadly counter-cyclical fashion. The LVR restrictions have done their job during the upswing in the market and have added resilience to the financial system over the course of the last seven years, since their introduction in 2013. 1 It is now appropriate to remove the restrictions. The requirements guard against credit growth, asset price inflation, and the risk of a correction in the credit cycle. These are all intended to mitigate future risks to financial system resilience due to a weakening of bank lending standards. Current economic conditions greatly reduce the need for such requirements.
7. Removing LVRs now does not weaken the resilience of the system. Rather, removing LVR restrictions now supports financial stability by removing one potential obstacle to the flow of credit in the economy. Removing the LVRs, for an initial period of one year, provides a signal that banks should continue lending during the COVID￾19 crisis and during the recovery. Nevertheless, the impact in this regard of removing LVR restrictions is likely to be relatively small, as banks are likely to maintain high credit standards.
8. The removal of the LVR restrictions also removes the risk that an increase in LVRs might discourage banks from extending mortgage deferrals to households with

1 For analysis of the revisions to LVR policy settings see Lu, B. (2019) “Review of the Reserve Bank’s Loan-to-Value Ratio Policy.” Reserve Bank of New Zealand Bulletin, 82(6).

3 temporary liquidity problems, if banks perceive the LVR restrictions to be an impediment. 8. In the current circumstances, banks are likely to tighten lending standards. LVR restrictions are unlikely to provide benefits to financial stability in these circumstances. However, continuing with LVR restrictions at such a time could introduce some unintended consequences. These include risks that banks may be reluctant to provide mortgage deferrals, or may reduce funding available to the non￾bank sector. 9. There are risks associated with removing the LVRs, including increases in household debt and the possible distributional impacts. The Reserve Bank considers these risks to be low in the current environment. Banks are unlikely to have a large appetite for providing loans to borrowers at a high risk of being unable to service the debt, especially if the borrower’s income is at risk of falling during the expected period of weakness in economic activity. 10. There are measures in place that will help mitigate the risks of removing the LVRs. Any increase in high-LVR lending will still lead to additional capital requirements through the prudential framework, which sees high-LVR loans receive a higher risk weighting in the calculation of risk weighted assets and capital ratios. The Reserve Bank will also maintain a close supervisory relationship with banks to quickly assess and respond to emerging risks. 11. While LVR restrictions are being removed, the surveys of banks’ new mortgage lending commitments and outstandings by LVR remain in place for statistical and continuity purposes. The impacts of the removal of LVRs will be carefully monitored and the impacts assessed in the Reserve Bank’s Financial Stability Report, which is published in May and November each year

4 Background 12. The purpose of macro-prudential policy is to reduce the risk that the financial system amplifies a severe downturn in the real economy. An unsustainable boom in credit and asset prices can result in a bust that creates losses for banks, businesses and households, and hampers the ability of banks to continue lending to the economy. Macro-prudential policy aims to reduce the likelihood and severity of a bust for the economy and for society. 13. The Memorandum of Understanding (MoU) on macro-prudential policy between the Minister of Finance and the Reserve Bank sets out a range of macro-prudential tools, including loan-to-value ratios (LVRs), which can potentially be used to mitigate rising systemic risks. 14. Macro-prudential tools transmit to financial stability and sustainable economic growth by directly improving the resilience of the financial system, mitigating the amplitude of the financial cycle, and reducing the potential decline in economic activity and asset price in a downturn. 15. LVRs place limits on the amount of lending banks can do above specified LVR ratios. This can limit the level of household debt – and therefore the potential losses that banks have to absorb in a downturn. This can curb the extremes of the boom-bust cycle, making the financial system less vulnerable to a financial crisis. 16. In the 2019 Macroprudential Policy Framework published by the Reserve Bank, the framework emphasised the role of LVR policy as moderating the amplitude of the financial cycle during the upturn and mitigating the feedback effects on the economy during the downturn.2 This represented a shift since when LVRs were introduced in 2013, when house price inflation was emphasised more as a policy goal. LVR restrictions are now considered to be most effective, and efficient, when there are signs of negative feedback effects between bank lending standards, household indebtedness and consumption, and the economy. 17. The onset of the COVID-19 pandemic, initially in China, then spreading across the world, including to New Zealand, has sharply curtailed economic activity. Treasury have produced scenarios showing falls in annual average GDP growth in New Zealand of between 0.5 percent and as much as 23.5 percent in the fiscal year ending 30 June 2021, and their estimates of the unemployment rate in the June 2021 quarter range from 5.5 percent to 22 percent.3 18. The Government and Reserve Bank have implemented a number of policies to help cushion households and businesses against these impacts. For households with mortgages, one of the key interventions has been the mortgage deferral programme. This covers a deferral of up to six months of mortgage payments for mortgage holders and Small-to-Medium Enterprise (SME) borrowers whose incomes have been affected by the economic disruption from COVID-19.

2 https://www.rbnz.govt.nz/- /media/ReserveBank/Files/Publications/Background%20papers/Macroprudential-policy￾framework.pdf?revision=02bcd033-70a7-4d28-829f-ee28d71fa7cb&la=en 3 https://treasury.govt.nz/news-and-events/news/covid-19-economic-scenarios-and-weekly-data￾published

5 19. Other interventions to support households and businesses include wage subsidies, and the Business Finance Guarantee Scheme. 4 The Reserve Bank has also implemented a number of policies to support the stability and efficiency of the financial system.5 20. In light of these unprecedented events the Reserve Bank has considered whether or not LVR restrictions remain appropriate for New Zealand. In particular, the Reserve Bank has considered whether retaining LVRs at the present time helps to achieve the objectives of macro-prudential policy. Status Quo 21. A mortgage loan-to-value ratio (LVR) is a measure of the size of a borrower’s loan, relative to the value of the property collateralised against the loan, often expressed as a percentage. LVR restrictions operate by requiring banks to limit their lending to borrowers with a high LVR, as defined by the policy, to below a specified proportion of their new lending (the ‘speed limit’). The Reserve Bank introduced the LVR policy in October 2013 in response to rising financial stability risks. The restrictions have been revised over time.6 22. LVR lending restrictions differ depending on whether a loan is secured by residential investment property or owner-occupied residential property. The current limits that apply to bank lending to investors and owner-occupiers are described below. Investor loans – 30% deposit / 5% of investor lending 23. LVR lending restrictions are tighter for investor loans due to the higher risks associated with this type of loan. The current policy classifies investor loans as high￾LVR if they are more than 70% of the property’s value, and restricts high-LVR lending to no more than 5% of a bank’s total new investor lending. Owner occupier loans – 20% deposit / 20% of owner occupier lending 24. This class of loan is for borrowing secured against owner occupied property. The current policy classifies owner occupier loans as high-LVR if they are more than 80% of the property’s value, and restricts high-LVR lending to no more than 20% of a bank’s total new owner-occupier lending. Application of LVRs 25. The BS19 document sets out the framework for the LVR policy. It describes how banks should calculate LVRs and exemptions. Annex 1 provides a summary.

4 Details of the Government response package are available here: https://treasury.govt.nz/information-and-services/new-zealand-economy/covid-19-economic￾response/package 5 Details of responses implemented by the Reserve Bank are available here: https://www.rbnz.govt.nz/covid-19 6 A summary is available in this paper: Lu, B. (2019) “Review of the Reserve Bank’s Loan-to-Value Ratio Policy.” Reserve Bank of New Zealand Bulletin, 82(6).

6 26. It is important to note that while falls in property values would, for a given loan size, increase LVRs for existing loans, such changes are not captured in the speed limits for LVR restrictions. The LVR restrictions only apply to new lending, whether that is for top-ups of existing loans, or completely new loans. 27. In cases where borrowers seek a mortgage deferral, the interest will continue to accrue on the loan. This will increase the total size of the loan and increase the LVR for that loan. However, for the purposes of calculating the LVR restrictions, only the new portion of lending is included in the calculation of each bank’s LVR ‘speed limit’. This process is described in more detail in Annex 1. Connection between LVRs and prudential policy 28. A bank’s capital ratio is calculated as capital divided by risk weighted assets. Higher risk weights mean banks must have more capital per dollar of lending for the type of exposure in question. High-LVR lending is therefore subject to higher risk weights, with risk weights increasing sharply as LVRs get higher. This is shown in Figure 1. Figure 1: Relationship between risk weights and LVRs Source: Bloor and Lu, 2019. Note that IRB is the Internal Ratings Based approach to estimating credit risk used by approved banks. Effectiveness of LVR restrictions since 2013 29. In 2019 the Reserve Bank reviewed the effectiveness of LVR policy.7 The Review found that the LVR restrictions have significantly improved the resilience of the banking system. 30. The Review found that the LVR policy has reduced the scale of mortgage defaults and credit losses that would occur in a housing downturn, due to a reduction in risky loans on bank balance sheets and the mitigation of a potential house price decline.

7 Bloor and Lu 2019: https://www.rbnz.govt.nz/research-and-publications/analytical￾notes/2019/an2019-07

7 This resilience benefit has been partly offset by a fall in capital requirements that results from lower credit risk, reducing the banks’ buffer for absorbing credit losses. Nevertheless, the LVR policy is estimated to have reduced mortgage losses – as a share of the capital banks hold against their housing loans – by 12 percentage points. The policy was found to have mitigated about half of the deterioration in bank resilience from 2013 that would have occurred in the absence of the policy. 31. In the November 2019 Financial Stability Report the Reserve Bank concluded that: “Restrictions on low deposit (or high loan-to-value ratio) loans have improved the resilience of banks and households by reducing the volume of loans with a higher risk of defaulting in a downturn. Current policy settings remain appropriate and we will be watching how risk evolves over the coming months.” 32. The onset of the COVID-19 pandemic and the expected sharp fall in economic activity has prompted the Reserve Bank to now review whether LVR restrictions remain appropriate. Summary of key statistics 33. The figures below summarise some of the key LVR statistics. Figures 2 and 3 show that nearly half of the stock of loans are at LVRs of less than 60 percent. Figure 4 shows that the proportion of new loans with high LVRs has been increasing in recent years. Figure 2: Percentage of loans by LVR Figure 3: LVRs over time Source: RBNZ calculations Source: Bloor and Lu, 2019 Figure 4: Proportion of new loans with high LVRs Figure 5: Credit, house prices and income Source: RBNZ. Note that high LVRs are defined here as new loans to owner-occupiers with LVRs above 80 percent and new loans to investors with LVRs above 70 percent Source: RBNZ, November 2019 Financial Stability Report 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% 0 2 4 6 8 10 12 14 16 18 20 Jan-16 Jan-17 Jan-18 Jan-19

8 Objectives of macro-prudential policy 34. The Reserve Bank’s objective when deploying macro-prudential policy instruments, such as LVR restrictions, is to promote financial stability by building the resilience of the financial system and dampening excessive growth in credit and asset prices. This helps the Reserve Bank to meet its statutory purpose of promoting the maintenance of a sound and efficient financial system. These objectives were set out when the LVRs were first implemented in 2013.8 35. Under section 1A of the Act, the Reserve Bank is responsible for, among other things, promoting the maintenance of a sound and efficient financial system. The framework for the Reserve Bank’s regulation and supervision of registered banks in New Zealand is governed by Part 5 of the Act. Section 68 requires the Reserve Bank to exercise the powers conferred on it for the purposes of:  Promoting the maintenance of a sound and efficient financial system.  Avoiding significant damage to the financial system that could result from the failure of a registered bank. 36. In 2019 the Reserve Bank published its updated framework for macro-prudential policy (RBNZ, 2013c).9 This sets out the objective of macro-prudential policy as: “The purpose of macroprudential policy is to reduce the risk that the financial system amplifies a severe downturn in the real economy. An unsustainable boom in credit and asset prices can result in a bust that creates losses for banks, businesses and households, and hampers the ability of banks to continue lending to the economy. This is important because financial instability – a disruption to the supply of essential services provided by the financial system – can have significant and lasting economic and social costs. Macroprudential policy aims to reduce the likelihood and severity of these costs.” 37. The framework notes that damaging boom-bust cycles can develop due to the pro￾cyclical interaction of the financial system with the real economy. The limited incentive for banks to recognise and internalise the costs of a future financial crises – that costs to the financial system and to the economy far exceed the private costs to banks and households – tends to amplify the boom-bust cycle. A tightening of macroprudential policy becomes necessary when these systemic risks are excessively heightened. 38. These objectives link back to the purposes in sections 1A and 68(a) of the Act, namely promoting the maintenance of a sound and efficient financial system, in that:  Macro-prudential policy complements baseline prudential requirements in maintaining the soundness of the financial system by lowering the likelihood and severity of a financial crisis on the economy and society.  Excessive growth in credit and unsustainable asset price developments are likely to impinge adversely on financial system efficiency due to allocative inefficiency

8 Regulatory impact assessment: Restrictions on high‐LVR residential mortgage lending, August 2013 9 https://www.rbnz.govt.nz/- /media/ReserveBank/Files/Publications/Background%20papers/Macroprudential-policy￾framework.pdf?revision=02bcd033-70a7-4d28-829f-ee28d71fa7cb&la=en

9 arising from asset prices diverting significantly from their fundamentals, and the tendency for the flow of credit to tighten sharply and disruptively in a downturn. 39. These two objectives also have the positive externality of promoting the purposes in section 68(b) as well, as they will help to avoid the significant damage to the financial system that could result from the failure of a registered bank. 40. The framework includes the a three step decision making framework for LVR policy decisions:  Assess credit growth/asset price growth.  Assess system resilience  Assess the risk of feedback effects between the financial system and the economy (whether banks are maintaining lending standards by reference mainly to LVRs and DTIs). 41. The “Comparison of options with objectives” section of this RIA uses the framework above to consider whether it is appropriate to remove LVR restrictions at the current time. Problem definition 42. LVRs were introduced to help support financial stability by building resilience in the system and dampening excessive credit growth. LVRs operate in a broadly counter￾cyclical fashion, reducing the ‘boom and bust’ cycles of the economy that affect the credit cycle and capital market flows. 43. The financial system is inherently pro-cyclical.10 Booms in the financial cycle tend to be exacerbated due to banks, investors and households underestimating the level of risk (Borio, 2010). Investor and borrower behaviour can be shaped by an over￾reliance on recent news and asset performance. As asset prices rise, provisions for non-performing loans fall and credit spreads narrow, banks and households assume that risks are lower than they are. These benign indicators, together with low volatility in markets, not only lower risk perceptions but also encourage banks and households to become more leveraged. 44. This process can continue until a macroeconomic downturn occurs, whether due to a specific shock or a general economic slowdown that causes investors and borrowers to reconsider the serviceability of their debt. The process then works in reverse: a slowdown in the economy or an external shock undermines households’ ability to service their debt and/or to buy houses. Heightened risk perceptions will develop, particularly if borrowers have leveraged their positions in the expectation of capital gains. As defaults rise and collateral asset values fall, banks are incentivised to raise the cost of lending and/or deleverage – reduce their lending to the economy. This results in a contraction in the availability of credit, which causes a further contraction in lending and asset prices. In other words, the downturn in the financial cycle and the slowdown in the economy become mutually reinforcing.

10 This section draws from the Reserve Bank’s 2019 Macroprudential Policy Framework.

10 45. If left unchecked, the pro-cyclicality of the financial system can amplify the scale of the boom-bust cycle and cause significant damage both to the financial system and to the real economy. The downturn can be severe and prolonged as banks and households take time to repair their balance sheets, particularly if they have become excessively leveraged during the upturn in the expectation that relatively cheap funding will continue to be available and that asset prices will continue to rise. 46. As noted in the previous section, the LVR restrictions have been effective at increasing the resilience of the financial system, and addressing the problems described above. However, with circumstances changing following COVID-19, the nature of the problem has also changed. 47. The onset of COVID-19 has seen the economy enter into a sharp contraction of economic activity. The exact shape of the slowdown and the speed of the recovery is highly uncertain. Even with the unprecedented levels of government intervention to support the economy, there is likely to be sharp drop in activity. Treasury have produced scenarios showing falls in annual average GDP growth by between 0.5 percent or as much as 23.5 percent in the fiscal year ending 30 June 2021, and estimates of the unemployment rate in the June 2021 quarter range from 5.5 percent to 22 percent.11 48. The central problem considered in the assessment of LVR restrictions is whether or not the LVR restrictions will continue to support financial stability during the COVID￾19 induced drop in economic activity. Another way of looking at this is whether the LVR restrictions - that have been effective in enhancing stability during an upswing in the credit cycle - remain appropriate during the sort of sharp fall in economic activity that New Zealand is now experiencing. 49. In particular, the Reserve Bank has been conscious of the following risks:  With uncertainty over the outlook, and the possibility of losses from credit defaults, banks may become overly cautious in the provision of credit, by cutting back lending, even to creditworthy borrowers. This could prevent the flow of credit within the economy, adding further to the slowing in economic activity. In an extreme example this could damage financial stability.  The existence of LVR restrictions could discourage banks from allowing borrowers to defer mortgage payments. The accumulation of interest during a deferral ultimately leads to a higher loan balance, and therefore a higher loan-to￾value ratio. If the LVR restrictions led banks to be reluctant to provide deferrals it would undermine the effectiveness of mortgage deferrals to manage financial stresses on temporarily liquidity-constrained households. If this led to a large run up in defaults it could damage financial stability. 50. There are a range of complex factors that will determine the extent to which these factors are a problem, including the size of the economic shock, the speed of the recovery and the impact on household incomes, unemployment and house prices. These outcomes are all highly uncertain.

11 Treasury, (2020), COVID-19 Economic scenarios

11 51. In addition, as shown in Annex 1, the technical definitions of how LVR restrictions are calculated each month mean that new lending through mortgage deferrals are unlikely to push existing loans into high levels of the LVR restrictions. Nevertheless, it is possible that the existence of the restrictions could affect banks’ appetite for such lending. The restrictions could also affect their appetite for provision of lending to non-banks, such as Non-Bank Deposit Takers, if they expect such lending to breache the “spirit” of the LVR restrictions. 52. If these factors were to combine in a way that saw a contraction in the availability of credit to creditworthy borrowers, this could see the economy face a ‘credit crunch’. Such an event would add to the already significant drop in output and activity that is likely following the COVID-19 pandemic. There is a risk that this could affect financial stability. The options considered by the Reserve Bank for the treatment of LVRs are designed to support financial stability and ensure that access to credit remains available. 53. This RIA also considers the risks that removing the LVR restrictions could bring. In particular, if loosening the LVR restrictions leads to sharp increases in new high-risk lending to borrowers with little capacity to service the debt, this could damage financial stability, if there was a subsequent sharp fall in housing prices. This could see households’ housing equity fall, which, if combined with falling incomes, could lead to rising defaults and risks to financial stability, undermining the objectives addressed in this RIA. 54. This risk is particularly relevant during a period where house prices may come under downward pressure – a sharp rise in household debt to high LVR borrowers could leave households exposed if house pries subsequently fell. This RIA considers all of these conflicting potential impacts on stability. 55. Removing LVR restrictions could also alter the balance of owner-occupiers and investor activity. While the removal of LVRs should make it easier for some first home buyers to buy a house, it may also make it easier for investors to grow their property holdings. The RIA considers these factors. Comparison of options with objectives 56. The Objectives section described the Reserve Bank’s three step decision making framework for LVR policy decisions:  Assess credit growth/asset price growth.  Assess system resilience  Assess the risk of feedback effects between the financial system and the economy (whether banks are maintaining lending standards by reference mainly to LVRs and DTIs). 57. Each of these steps are considered in Table 1.

12 Table 1: Three step decision making framework Step Assessment Assess credit growth/asset price growth The Reserve Bank considers that following the onset of Covid-19 the New Zealand economy will experience a sharp drop in economic activity. The recovery from this is likely to take a prolonged period of time. As a consequence, credit growth and asset prices are likely to be subdued Assess system resilience The financial system is in good shape. Trading banks have capital above minimum levels and have solid levels of liquidity. Assess the risk of feedback effects between the financial system and the economy Current drop in activity, and uncertainty about the future path of labour market developments and household incomes, means banks are likely to be conservative with their lending standards. Key risk is that banks are overly cautious, adding to the risk that credit does not flow to credit-worthy borrowers. 58. The last of the three steps above suggests it is appropriate to consider LVR changes as a way to remove a potential barrier to the flow of credit in the economy to credit￾worthy borrowers. The Reserve Bank considered three options for setting LVRs following the onset of the COVID-19 pandemic, alongside the status quo of no change:  Option 1: Include an exemption in the LVR policy (BS19) for customers in hardship. This would require the Reserve Bank to provide guidance to banks that any high-LVR increase in loan value that arose from helping a borrower manage financial hardship could be excluded from the total of high-LVR lending for the purpose of the speed limit.  Option 2: Ease LVR rules. This would see LVR restrictions retained, but set at a less restrictive level.  Option 3: Remove LVR restrictions for a period of one year. 59. Table 2 considers each of these options against the objective described in the Objectives section of this RIA, relative to the status quo (Option 1). Table 2: Assessment of options against LVR objective Reduce the risk that the financial system amplifies a severe downturn in the real economy Option 1: Exemption in BS19 for hardship ▬ Option 2: Ease the LVR restrictions ↑ Option 3: Remove LVR restrictions ↑↑ 60. The existence of LVRs since 2013 have met this objective, during a period of expansions in credit and rising house prices.

13 61. Given that the economy is entering a downturn phase, removing or easing the LVRs now (Option 2 or 3), would be consistent with the countercyclical nature of the LVR policy. 62. Options 2 and 3 are ranked ahead of Option 1 against the objective Table 1 and are both ranked as superior to the status quo. Option 1 is likely to help manage consumer hardship, but due to its narrow focus it is unlikely to affect stability or extremes in the cycle in any way. 63. Options 2 and 3 are also ranked ahead of the status quo because they both provide clear signals that banks should continue lending to credit-worthy borrowers to help support the flow of credit. They also avoid the risk that LVR restrictions might discourage banks from offering mortgage deferrals. 64. The assessment above is not clear-cut and reflects a judgement about the following two competing factors:  Supporting lending helps maintain the flow of credit.  Looser LVR restrictions might encourage banks to undertake more LVR lending, if this resulted in more lending to high risk borrowers, this could reduce financial stability. 65. The Reserve Bank’s assessment is that the first of the factors above will outweigh the second factor in each of Options 2 and 3, given the current state of the economy, and the possibility that banks might be overly cautious and stop lending to credit￾worthy customers. The issue is considered in more depth in the Costs and Benefits section of this RIA. 66. In principle, Option 2 should be able to deliver similar payoffs as Option 3, if calibrated to the right level. However, given the uncertainty around the outlook following COVID-19, calibrating the LVRs to an appropriate level is complex, and subject to a wide margin of error. As a result, Option 3 is ranked as superior to Option 2, when assessing against the objective of macro-prudential policy. 67. A number of the factors above are also considered in more depth in the next section of this RIA. In addition, there are a range of other factors, considered in further depth in the next section, that sit outside the objectives described above. They are consequences that might arise from the implementation of any of these options and were points raised in submissions from the public, so are considered as part of the wider assessment of costs and benefits. Costs and benefits of the options 68. This section of the RIA assesses the costs and benefits of the options above. A full quantification of the costs and benefits would require a detailed model that incorporated the expected economic scenarios following COVID-19, as well as household, business and bank responses, to compare a counterfactual of the status quo against each option. Such an exercise would be complex and subject to wide ranges of margins of error. Instead, Table 3 summarises the range of possible costs and benefits, while Table 4 compares the costs and benefits of each option against

14 the status quo. This is similar to the approach taken in the 2013 Regulatory Impact Assessment that accompanied the original introduction of the LVRs.12 Table 3: Possible costs and benefits of removing LVRs Possible Cost Commentary Risk of sharp increase in lending to high risk borrowers Banks and borrowers may underestimate the size of a future house price correction and extend high-LVR loans at the top of the market. This could create future risks to stability. Impact on first home buyers Some options could make it harder for first home buyers to buy a house. Less provision of rental accommodation and/or higher price of rent If the options result in fewer investors buying residential property, this could reduce the availability of accommodation, or increase rents. Possible Benefit Commentary Increased resilience, supporting financial stability Some of the options may help support financial stability, by supporting ongoing access to credit and dampening asset price cycles. This helps reduce the risk that banks restrict credit to credit-worthy borrowers, driving a ‘credit crunch’ and damaging financial stability. Less risk of financial hardship for mortgage deferral customers The downturn resulting from the COVID-19 pandemic will likely impact many New Zealanders financially. Options may support additional provision of credit to borrowers experiencing temporary liquidity constraints during COVID-19. More residential construction activity More access to credit may help stimulate residential construction activity. Impact on first home buyers Some options could make it easier for first home buyers to buy a house. 69. Table 4 compares the likely costs and benefits in Table 3 for each of the options considered by the Reserve Bank compared with the status quo of retaining the existing LVR restrictions.

12 Reserve Bank of New Zealand, (2013), Regulatory impact assessment: Restrictions on high‐LVR residential mortgage lending

15 Table 4: Assessment of costs and benefits of each option compared with the status quo Option 1: Exemption in BS19 for hardship Option 2: Ease the LVR restrictions Option 3: Remove the LVR restrictions Costs Risk of sharp increase in lending to high risk borrowers ▬ ▬ ▬ Risk of sharp drop in availability of credit, reducing financial stability and adding to asset price volatility ▬ ↑ ↑↑ Impact on first home buyers ▬ Uncertain Uncertain Less provision of rental accommodation and/or higher price of rent ▬ ↑ ↑ Benefits Increased resilience, supporting financial stability ▬ ↑ ↑ Less risk of financial hardship for mortgage deferral customers ↑ ↑ ↑↑ More residential construction activity ▬ Uncertain Uncertain Impact on first home buyers ▬ Uncertain Uncertain ↑ means better than the status quo, ▬ means no different, ↓ means worse than the status quo

16 70. Table 4 suggests that the largest benefits are likely to be associated with Option 3, removing the LVR restrictions, and reassessing in one year. As such, Option 3 is the preferred option identified in this RIA. 71. Option 3 goes the furthest towards encouraging banks to continue to supply credit to credit-worthy borrowers. This is the option most closely aligned with avoiding instability in the domestic financial system arising from credit, asset price, or liquidity shocks. It also helps dampen the credit cycle that could emerge if banks were too cautious in lending to credit-worthy borrowers, driving an even larger adjustment in asset prices than would be consistent with falls in income and rises in unemployment. Box 1 on the following page considers resilience impacts further. 72. Removing the LVR restrictions will also limit the extent to which the restrictions could have led to a sharp adjustment in the availability of credit to the Non-Bank Deposit Taking (NBDT) Sector. A number of NBDTs access credit from registered banks to fund their own mortgage lending. To the extent that higher LVR lending may be likely in the NBDT sector, including through mortgage deferrals, removing the LVR restrictions should reduce the risk that NBDTs will lose access to credit. 73. There are risks if borrowing with high LVR loans was to expand quickly, following the removal of the restrictions, driving an upward credit cycle, a lift in risky lending to highly leveraged borrowers, and raising the possibility of an even bigger eventual fall in house prices. This is the main area of risk and uncertainty identified in this RIA. 74. The Reserve Bank assesses the likelihood of this scenario to be small. Banks’ appetite for such lending in the current environment is likely to be low. Banks will be wary of lending to borrowers at an elevated risk of defaulting, owing to the general lift in default risk at the present time. In addition, there are safeguards to help minimise the risk of a sharp increase in lending fuelling an upward shift in the credit cycle: 13  The risk weighting system in the prudential framework discourages such an outcome by imposing rising capital requirements through higher risk weights on high LVR loans.  Supervision: in its supervisory role the Reserve Bank will carefully monitor household and bank responses, and has scope to intervene if bank lending is not prudent. 75. This RIA does not attempt to quantify the extent to which removing LVR restrictions will increase financial stability, or minimise credit and asset price cycles during the recovery from COVID-19 economic impacts. There are no historical incidents of a similar size or shape as the present circumstance that can provide a benchmark. 76. Option 3 also goes the furthest to support households facing temporary liquidity constraints following COVID-19. However, as noted in the problem definition, the final impact of LVR restrictions on responses to mortgage deferrals may be relatively small.

13 There are other regulations that will influence lender decision making, such as the Credit Contracts and Consumer Finance Act (CCCFA) and the responsible lending code for consumer lending, all of which are administered by the Ministry of Business, Innovation and Employment.

17 77. The impacts on first home buyers are uncertain. While the absence of LVR restrictions may reduce the size of deposit required for a first home buyer, it may also increase the ability of investors to borrow. It is unclear exactly how these dynamics will play out in the future, particularly during a period where house prices are likely to be falling. As a result, the impact on first home buyers is ambiguous. 78. Likewise, the impact on rents is unclear, given the incentives investors will face, and uncertainty around the future path of house prices, however, the impact of removing LVRs is unlikely to see rents increase. There are a range of possible responses by investors that have their borrowing capacity increased by the removal of LVRs:  Some may not respond at all, if they consider housing to be a riskier investment in the future, however this is likely to be determined by the fundamental drivers of house prices and rents, not the removal of the LVR restrictions.  Some may seek to buy more properties if LVR restrictions have previously prevented this response, and they remain confident about the long term outlook for housing.  Some may seek to sell properties if they consider the outlook for housing investment to have deteriorated, however this is likely to be determined by the fundamental drivers of house prices and rents, not the removal of the LVR restrictions. 79. Removing LVR restrictions for only owner-occupied housing is another option that the Reserve Bank could have considered. This option would clearly provide first home buyers an advantage relative to investors, as high LVRs would provide less regulatory impediment to first home buyers. However, this objective sits outside of the Reserve Bank’s mandate and it is not clear that such an approach would increase financial stability relative to Option 3. Indeed, a situation where banks curtailed lending to investors due to concerns about LVR restrictions might add to financial instability. Box 1: A focus on resilience The impact of removing LVRs on financial system resilience is a key judgement in this RIA. Removing LVR restrictions will make it easier for banks to lend to borrowers with smaller deposits. This is likely to have two direct effects on the risks banks face, which are inputs into the approach to modelling credit risk for the large banks:  The probability of default (PD) will increase as borrowers have smaller buffers in the event of a fall in income.  The loss given default (LGD) will increase, as with a smaller deposit there is a larger loan remaining in the event of a default. These factors will increase the possible risks that banks face and will increase their risk weighted assets if they increase their high LVR lending. Moreover, it is possible that they will face this regardless, if house prices fall following the COVID-19 economic shock. These factors could reduce financial stability. Balanced against these factors is the risk that with the deteriorating economic environment banks will curtail lending, driving a credit crunch, with clear risks for financial stability. The Reserve Bank’s assessment is that the signalling impacts of removing LVR restrictions will help support lending during the recovery from COVID-19 and that this will offset the possible negative impacts described above. This is a key judgement underpinning the RIA.

18 Consultation 80. A short, seven day, consultation period was completed by the Reserve Bank following the announcement on 21 April of the proposal to temporarily remove LVR restrictions. Banks and the general public were provided the opportunity to respond to the consultation. A short consultation period was necessary in order to respond in a prompt way following an unprecedented set of economic events. The Reserve Bank also directly approached a number of NGOs, including Māori and Pasifika groups to make them aware of the proposed changes, and to provide them a chance to provide feedback on the proposals. 81. Their views are summarised below. These views were considered by the Reserve Bank throughout the consultation period, and some are more thoroughly considered in the Regulatory Impact Assessment (RIA). The Reserve Bank has also published a summary of the submissions as part of the announcement of the LVR decision. The summary includes a brief response to some of the issues raised by submitters. Locally incorporated banks were in favour of the proposals 82. All locally incorporated banks who responded were in favour of removing LVR restrictions for one year, noting that this will allow banks to support customers through the impact of COVID-19. Submitters from the public in favour of the proposal noted the benefits to first home buyers and stabilising house prices 83. Many first home buyers were in favour of the proposal, noting that this would get them into the property market sooner. Some of them noted that their Kiwisaver balances had declined, and so they had lower deposit amounts with which to enter the market. Some submitters noted that it may make it easier for low income households to save a deposit, including for Māori and Pasifika households. A few submitters also noted that this will also allow investors to come back into the market, who may have been disincentivised due to other government policy, noting the need for more rental properties. Issues around first home buyers are outside of the direct financial stability mandate of the Reserve Bank. 84. Those submitters in favour of the proposal noted that it would be beneficial for stabilising house prices, given a likely house price decline and some forced property sales for those having trouble meeting debt obligations in the future. Some submitters noted that banks’ response to the removal of LVR requirements would be important, noting that banks should still adopt sensible deposit/equity requirements for investors and first home buyers. Submitters against the proposal expressed a broad range of views 85. Of those submitters that were against the removal of the LVR policy, many were concerned about potential adverse impacts on financial stability, such as the risk of bank failure. They noted that the current economic circumstances are highly uncertain, and they expressed concern that removing restrictions would increase the level of household and personal indebtedness at a time when taking on new debt should be discouraged. These submitters noted that the removal of the policy could encourage banks to undertake significant new lending at high LVRs against

19 potentially declining house prices, and that those who borrowed with no restrictions now would risk falling into negative equity. They also noted that the economy has weakened and job security has reduced, and the ability of people to service a mortgage will likely decline in the coming months. Some submitters believed that this, in conjunction with the delay in implementing bank capital requirements, would lead banks to lend imprudently. 86. Many of the submitters arguing against the removal of LVR restrictions cited its success at achieving financial stability, as indicated by stress test results and 2019 review of the LVR policy. They noted that the policy has ensured banks are available to withstand a significant downturn, and questioned why the Reserve Bank would be considering removing LVRs given that they have been an effective tool for keeping the banking system secure. A couple of submitters noted that under the current LVR speed limit rules, most banks could still have enough room to lend to high LVR loans. 87. As noted elsewhere in this RIA, given the current uncertainty around the economic outlook, the Reserve Bank considers it to be unlikely that banks will weaken lending standards to high risk borrowers. The more likely risk is that banks are overly cautious with their ending to credit-worthy borrowers. 88. A number of submitters also proposed alternative approaches to continue a differentiation between owner-occupier and investors. These submitters focused on ways to reduce the competition that first home buyers face from investors in the housing market. These considerations are outside of the Reserve Bank’s financial stability mandate. Implementation 89. The removal of LVR restrictions will be implemented through banks’ conditions of registration. The Reserve Bank has consulted banks about the proposed changes about their conditions of registration. 90. The central change to the conditions is to remove the references to the LVRs, which are defined in BS19. 91. While LVR restrictions are being removed, the surveys of banks’ new commitments by LVR remain in place for statistical and continuity purposes. Monitoring, Evaluation and Review 92. The Reserve Bank currently monitors macro-prudential indicators, including household indebtedness and debt servicing burdens, the level of banks’ high-LVR lending, and a range of metrics on house prices, sales, and other housing market activity. 93. Following implementation of the removal of LVR restrictions, the Reserve Bank will monitor the effect of the removal. 94. To monitor the effect of removing the restrictions, the Reserve Bank will assess the flow of residential mortgage lending by banks, and changes to aggregate household credit growth and house price growth. The Reserve Bank will report on these developments in future Financial Stability Reports.

20 95. Part of the monitoring of the removal of LVR restrictions will include close assessment of whether risks to financial stability are rising. For example, sharp increases in high LVR lending may be evidence that risks are building. The Reserve Bank will monitor such developments carefully, including whether other policy responses need to be considered, if such patterns emerge.

21 Annex 1: Summary of BS19 BS19 sets out the Reserve Bank’s framework for imposing quantitative restrictions on the share of high loan-to-valuation ratio (LVR) loans by registered banks to the residential property sector. Loans to the residential property sector include loans secured by owner￾occupied residential property and loans secured by residential investment property. BS19 provides the formulas banks should follow, guidelines about that is covered and a list of exemptions. Exemptions include Welcome Home Loans and bridging finance. The framework set out in BS19 follows this approach quoted below, with variants for owner￾occupiers and investors: “qualifying new mortgage lending amounts must not— (a) for residential properties with a loan-to-valuation ratio of more than x%, exceed x% of the total of the qualifying new mortgage lending amounts arising in the loan-to￾valuation measurement period.” Further to this, Section 10 of BS19 defines “qualifying new mortgage lending” as: “ “qualifying new mortgage lending amount”, is the sum of — (i) the total of any loan values associated with a qualifying mortgage loan made by the registered bank in a loan-to-valuation measurement period in accordance with a qualifying mortgage loan commitment; and (ii) the total of any increases in loan value that are a qualifying increase in mortgage loan value made by the registered bank in a loan-to-valuation measurement period.” Section 11 clarifies that: “If the qualifying new mortgage lending amount does not arise from a commitment for a new mortgage, the amount is measured as the increase in the loan value associated with the mortgage. Qualifying new mortgage lending amounts do not include commitments to provide a new residential mortgage loan, or increases in the loan value of an existing loan, that the bank treats as exempt under one of the exemption categories.”

22 References Bloor, C. and Lu, B. (2019), Have the LVR restrictions improved the resilience of the banking system? Reserve Bank of New Zealand Analytical Note Series Borio, (2010), ‘Implementing a macroprudential framework: Blending boldness and realism’, Bank for International Settlements, July 2010. Lu, B. (2019) “Review of the Reserve Bank’s Loan-to-Value Ratio Policy.” Reserve Bank of New Zealand Bulletin, 82(6) Ovenden, P. 2019, Reserve Bank of New Zealand Macroprudential Policy Framework. Reserve Bank of New Zealand, (2019), Financial Stability Report, November 2019 Reserve Bank of New Zealand, (2016), Regulatory Impact Assessment of revised LVR restriction proposals Reserve Bank of New Zealand, (2013), Regulatory impact assessment: Restrictions on high‐ LVR residential mortgage lending Reserve Bank of New Zealand, (2013), BS19, Framework for Restriction for High Loan-to￾Value Residential Mortgage Lending Reserve Bank of New Zealand, (2013), Memorandum of Understanding between the Minister of Finance and the Governor of the Reserve Bank of New Zealand Treasury, (2020), COVID-19 Economic scenarios