Consultation Paper on Adjustments to Restrictions on High-LVR Residential Mortgage Lending

Consultation Paper:
Adjustments to restrictions on high-LVR residential mortgage lending The Reserve Bank invites submissions on this Consultation Paper by 13 July 2015. Please ensure that responses are sent in before the closing date. Submissions received after this date cannot be considered. Submissions and enquiries about the consultation should be addressed to:
Attention:
Head, Macro Financial Department
Reserve Bank of New Zealand
PO Box 2498
Wellington 6140
Email: macroprudential@rbnz.govt.nz
When responding, please state whether you are doing so as an individual or on behalf of an organisation. Please note that a summary of submissions may be published. If you think any part of your submission should properly be withheld on the grounds of commercial sensitivity or for any other reason, you should indicate this clearly. Corrections after publication:
In point 83, “should be treated as a non-property investor” has been corrected to read “should be treated as a property investor” June 2015

Introduction

1. The Reserve Bank introduced speed limits on high loan-to-value ratio (LVR)
   mortgage lending in October 2013 in response to rising financial stability risks associated with strong and rising house price inflation at that time. This speed limit required that banks limited mortgage lending at greater than 80% LVR to no more than 10% of new commitments.
2. The policy helped to slow the rate of house price growth through 2014. It has also
   had an ongoing effect on the resilience of the New Zealand banking system by reducing the share of high-LVR loans on bank balance sheets. However, there has been a significant resurgence in house price pressures in the Auckland market since October 2014, and this is leading to a re-emergence of financial stability risks.
3. This consultation paper sets out proposed changes to restrictions on high-LVR
   mortgage lending in order to address this build-up of risks. The broad shape of these policy changes was announced at the release of the Financial Stability Report on May 13.
4. The specific proposals are to:
    Restrict property investment residential mortgage loans in the Auckland region at LVRs of greater than 70 percent to 2 percent of total property investment residential mortgage commitments in Auckland.  Retain the existing speed limit of 10 percent for other residential mortgage lending, as a proportion of total non-property investment residential mortgage commitments, in the Auckland region at LVRs above 80 percent.  Increase the speed limit on residential mortgage lending at LVRs above 80 percent outside of Auckland to 15 percent of residential mortgage commitments outside Auckland.
5. A number of loan categories are exempted from LVR speed limits, and these
   exemptions will be retained under the proposed policy changes. Specifically, loans that are made as part of Housing New Zealand’s Welcome Home Loan scheme, and loans that are made for the purpose of refinancing an existing mortgage loan, moving house (without increasing borrowing amount), bridging finance or constructing a new dwelling will continue to be exempt from the policy. Problem definition
6. New Zealand experienced one of the strongest increases in house prices in the
   OECD in the period prior to the global financial crisis (GFC), with house prices more than doubling between 2002 and 2007. Over the same period, household debt increased from 114 percent to 160 percent of household disposable incomes (including debt held against rental properties). In contrast to many other countries, New Zealand did not experience a significant housing market downturn following the GFC, leaving house prices at relatively stretched levels relative to both incomes and rents.
7. The current high-LVR speed limits were introduced from 1 October 2013 in response
   to renewed rapid house price inflation, which was exacerbating stretched valuations.

The Reserve Bank’s concern was that the risk of a subsequent sharp housing market correction was rising, and could pose risks to the New Zealand financial system.
8. The LVR policy had a significant impact on housing market activity and growth in
house prices in the year to September 2014. National house price inflation slowed from 9.3 percent to 4.9 percent between September 2013 and September 2014. The slowdown in house price inflation was more dramatic in the Auckland region, falling from 16.5 percent to 8.5 percent over the same period.
9. Since September 2014, there has been a significant increase in both housing market
activity and house price inflation in the Auckland region. Auckland house prices have increased by 12 percent in the past four months, taking annual growth to 17 percent in April. Increased housing demand has been driven by a combination of record net immigration into the region and a reduction in fixed mortgage rates, particularly at longer terms.
10. Net housing credit growth has been fairly subdued, at 5.2 percent in the year to April.
However, households are already heavily indebted, with household debt at 160 percent of disposable income. Net credit growth has also been held down by relatively strong repayment of existing balances recently, which is disguising strong flows of new lending. Gross lending flows have been running at much stronger rates (see figure 1).
11. Gross lending flows are particularly important for financial stability, as it is new
lending (net of refinancing) that drives movements in asset markets. Also, loans are most prone to default in the years immediately following origination, creating a risk should the property market turn down in the next few years.
Figure 1: Mortgage approvals, housing credit and house sales
(3-monthly seasonally-adjusted totals, percent of quarterly nominal GDP) Source: RBNZ, REINZ.
12. Increased demand to purchase property, accompanied by existing shortages of
housing in Auckland and significant supply constraints, is resulting in upward pressure on house prices.

13. At the same time that pressures in the Auckland housing market have increased, the
    Canterbury housing market has eased, and price pressures in the rest of the country have remained more subdued (figure 2).
    Figure 2: House price to income and current house price growth
    (as at March 2015)
    Source: Interest.co.nz, QV.

14. In addition to divergence in house price growth, Auckland house prices are
    significantly more elevated relative to incomes and rents. The Auckland housing market is now one of the least affordable housing markets in the world, with a house price to household income ratio of around 8. By some estimates, this surpasses ratios seen in London and Melbourne, and approaches Sydney multiples.

15. There are logical reasons why house prices are higher in Auckland than in other
    parts of New Zealand, even relative to income. Population inflows have created a shortage of housing, and inward migration to Auckland is expected to remain persistent. This creates a prospect that rents may rise in the future, and the current low-interest rate environment increases the sensitivity of house prices to that expected future rental growth. However, the levels and continuing momentum in Auckland house prices appear to go beyond what is justifiable on the basis of these fundamental factors.

16. Auckland is the largest city in New Zealand, but is much smaller than most cities that
    have sustained high house price to income ratios for prolonged periods (e.g. London, Sydney, Hong Kong, New York). New Zealand also has substantially higher mortgage rates than most of the countries or cities noted above.

17. Historically, the house price to income ratio tends to drift down after periods when it
    has risen strongly. Even real house prices show no clear trend over long periods in many countries, despite the tendency for incomes to rise over the last few centuries (see e.g. Shiller 2006). This partly reflects that fundamentals can change. In the case of Auckland today, migration trends can reverse, interest rates may rise in the future, and the Government and Council are working to address the supply imbalances.

18. As a result, the case for specifically targeting policy measures to the Auckland
    market is significantly stronger than it was at the time when the LVR policy was introduced. While there is a risk that there could be a re-emergence of house price pressures outside of Auckland, it would take a sustained period of strong growth before valuations became as stretched as in Auckland.

19. There is evidence that investors are playing a growing role in the Auckland market,
    and an increase in investor demand is likely to be one factor contributing to stronger market conditions. The investor share of transactions in Auckland increased from around 36 percent prior to the introduction of LVR restrictions to over 40 percent in the first three months of 2015. Some of this increase in the investor share is due to falling participation by first home buyers following the introduction of LVR restrictions. Nevertheless, the level of sales to investors is around 12 percent higher now than immediately prior to the introduction of LVR speed limits and has been strong recently.
    Figure 3: House sales to multiple property owners (% of total sales)

Source: CoreLogic.
20. Overall, it appears unlikely that supply constraints will be materially alleviated in the
near term, and there is no obvious immediate trigger for a softening in housing demand. Therefore, there is a significant risk that housing market pressures will persist for a number of years, exacerbating existing imbalances.
21. Currently Auckland accounts for around a half of outstanding mortgage lending, and
most likely more than half of new lending flows. This concentration of lending in a single geographic market means that developments in the Auckland housing market are of systemic importance to the New Zealand financial system.
22. The Reserve Bank’s concern is that the risk of a substantial correction in the
Auckland property market is rising, and will continue to rise the more that prices

depart from fundamental values. A significant correction in house prices could be triggered by a range of factors, including a sharp economic downturn leading to a marked deterioration in the labour market and a turnaround in migration flows, or a material increase in mortgage interest rates. Such a correction could threaten the stability of the banking system if the resulting increase in mortgage default rates and credit losses were sufficiently large. A large correction could also generate a significant period of macroeconomic weakness, particularly if a large number of households ended up in a position of debt overhang following a market correction. Such a scenario would further exacerbate stress on bank balance sheets.
23. The Reserve Bank, in conjunction with the Australian Prudential Regulation Authority,
ran stress tests of the New Zealand banking system during 2014. These stress tests featured a significant housing market downturn, concentrated in the Auckland region, as well as a generalised economic downturn. While banks reported generally robust results in these tests, capital ratios fell to within 1 percent of minimum requirements for the system as a whole. Since the scenarios for this test were finalised in early 2014, Auckland house prices have increased by a further 18 percent. Further, the share of lending going to Auckland is increasing, and a greater share of this lending is going to investors. The Reserve Bank’s assessment is that stress test results would be worse if the exercise was repeated now. Loans to investors carry greater risk
24. Residential property investment loans appear to have relatively low default rates
during normal economic circumstances. However, the Reserve Bank has looked at evidence from extreme housing downturns during the GFC, and this clearly indicates that default rates can be higher for investor loans than for owner occupiers in severe downturns. For example, as shown in table 1, forecast loss rates on Irish mortgages were nearly twice as high for investors as for owner-occupiers. Similarly, actual arrears rates were about twice as high for investor loans (29.4 percent) than for owner occupied loans (14.8 percent) as at December 2014.1

Table 1: Ireland residential loans - forecast loss estimates, 2011-2013
Owner
Occupier
Investor Total
Central Bank of
Ireland
5.9% 10.7% 7.0%
Financial Measures
Programme
(BlackRock Solutions)
7.6% 14.3% 9.2%
Source: Kelly (2011).
25. Furthermore, studies which have separately estimated default rates by LVR for
investor loans and owner occupier loans suggest that investor loans are substantially riskier at any given LVR. The figure below is from Kelly (2012) and shows an estimate of default rate based on current LVR. For example, if a loan was initially written at a 70 percent LVR and then prices fell 30 percent, the loan would appear in the chart below as LTV=100. This would have a mildly increased rate of default compared to a low-LVR loan for an owner occupier. But for an investor, the rate of default would be higher, and would have increased more sharply as a result of a given decline in house prices.

1 Central Bank of Ireland Residential Property Arrears and Repossession Statistics.

Figure 4: Default rates by current LTV in Ireland
Source: Kelly (2012).
Note: PDH is principal dwelling house, BTL is buy to let. LTV (loan to value ratio) is conceptually the same as LVR, but this dataset uses the current LTV (after the sharp falls in house prices) rather than origination LTV.
26. Evidence on investors from the US GFC experience is complicated to interpret as it
appears statistics distinguishing investors from owner occupiers may not always have been reliable. Most studies of the US have data on declared intentions rather than actual occupancy status. However, a New York Fed study which defined investors as owners of multiple properties (rather than using declared intentions) found that investors were an important driver of downturn defaults (Haughwout et al, 2014). In contrast, investor’s role was much harder to see if investors were identified using the borrower’s declared intentions. Palmer (2014) reports that default rates increased in a multivariate regression with loan to value ratio and for loans that were to declared non-owner occupiers.
27. Evidence from New Zealand and Australia is significantly more limited, as there has
not been a severe downturn, and investor loans are not always separated from owner occupied loans in default data. Fitch Ratings (2012) has reported on empirical work using data from securitised mortgages in Australia, which suggests that investor loans performed similarly to owner occupier loans in normal times but significantly worse in downturns. Rating agency models of residential mortgage default also tend to treat investor loans as riskier at any given LVR.

28. There are several structural factors which appear likely to make investor lending
    riskier at any given LVR. First, for a typical investor who owns their own home and several others at high LVRs, gearing relative to income (whether including or excluding rental income) will be substantially higher than for a typical owner occupier at the same LVR. This means that a substantial fall in house prices would leave the investor much more heavily underwater relative to their labour income. This diminishes their incentive to continue to service the mortgage (relative to alternatives such as entering bankruptcy).
29. Second, some investors may not own their own home directly (it may be in a trust
    and not used as security, or they may rent the home they live in). Again, this is likely to increase the incentive to stop servicing debt if it exceeds the value of their investment property portfolio. The Reserve Bank considers ‘strategic default’ to be unlikely for NZ owner occupiers in most circumstances, but it is a more realistic prospect for investors in severe downturns.
30. Finally, investors may face additional income volatility related to the possibility that
    the rental market they are operating in weakens in a severe recession (if tenants are in arrears or are hard to replace when they leave, for example). That income volatility is more directly correlated with the valuation of the underlying asset, since it is harder to sell an investment property that can’t find a tenant.
31. Investor lending can also be a strong driver of speculative rises in property markets,
    as the US and Irish experience indicates. Coates et al (2015) document a strong rise in investor activity in Ireland during the period of strong house price appreciation up to 2007. Q1: Do you have any comments on this analysis or the Reserve Bank’s rationale? Policy options
32. The Reserve Bank considered a range of options to address the build-up of risks
    identified in the previous section. These options were adjustments to LVR restrictions, Loan or total debt to income restrictions (LTIs or DTIs), or a temporary increase in capital requirements for housing lending.
33. Restrictions related to the debt servicing capacity of borrowers, such as LTIs or
    TDTIs, would also reduce banks’ exposure to high-risk loans and would also lower demand pressure in the property market.2 LTI restrictions have recently been announced by the Bank of England and the Central Bank of Ireland.
34. The Reserve Bank is currently gathering data on the exposure of banks to high-LTI
    and high-TDTI loans. Policy initiatives related to debt servicing capacity could be considered should our investigations conclude that loans are being originated at income multiples that pose a prudential risk. However, because LTI and TDTI are not defined as formal regulatory concepts in the same way as LVRs, policy initiatives in the area are relatively complex.

2 Loan to income multiples relate the size of a single loan facility to the borrowers income, whereas a total debt to income ratio attempts to capture all debts of the borrower. This distinction is particularly important for property investors whose mortgage debt may be spread across multiple facilities at multiple institutions.

35. Capital-based measures, such as temporary higher capital requirements for housing
    lending, were also considered in order to bolster the resilience of the banking system. The Reserve Bank has already consulted on a new asset class for property investment loans, which is expected to result in higher capital requirements for this class of lending. The Reserve Bank also plans to undertake a review of current bank capital requirements, including risk weights for standardised banks and IRB banks and capital ratio requirements. The Reserve Bank considers it undesirable to apply temporary macro-prudential capital overlays when permanent capital requirements are under review. Furthermore, a capital overlay is unlikely to have a material effect in dampening momentum in the Auckland housing market.
36. The Reserve Bank believes that tighter LVR restrictions on Auckland property
    investors is the most effective currently available tool to address the specific risks that are arising at the moment. Policy objectives and expected effectiveness
37. There are two primary objectives of the proposed changes to the LVR policy. The
    first is to reduce the rate of growth in Auckland house prices in order to limit the probability and magnitude of a subsequent correction. However, macro-prudential policy responses can only have a moderate effect on house price growth, and this impact tends to wane over time. Given the size of the imbalances in the Auckland housing market, house price growth is likely to persist in the near term.
38. The second objective is to improve the resilience of bank balance sheets to a
    housing market correction, by reducing the sector’s exposure to riskier loans and by reducing the magnitude and probability of a housing correction.
39. Bloor and McDonald (2013) set out a framework for analysing the impact of LVR
    restrictions on housing market activity. This paper estimated that the initial policy was likely to reduce house price inflation by 1-4 percentage points over the first year that the policy was in place. Subsequent work by Price (2014) conducted a counterfactual analysis of the initial impact of the policy, and concluded that the impact was at the upper end of that distribution.
40. Using a similar methodology to Bloor and McDonald (2013), but updated to reflect
    the actual experience of LVR restrictions, the Reserve Bank estimates that tighter LVR restrictions on Auckland investors will reduce Auckland house price growth by 2- 4 percentage points over the first 12 months following the adjustment.
41. The Reserve Bank currently collects data on lending to investors at a national level,
    which shows that around 30 percent of new mortgage lending flows are for investor property. Around half of this lending is at LVRs of above 70 percent. There is currently no lending data available specifically for Auckland, but based on national lending data and sales transaction data from Corelogic, the Reserve Bank estimates that around 17 percent of all mortgage lending in Auckland is to investors with LVRs of over 70 percent. This lending finances about 13 percent of Auckland property transactions.

Figure 5: New mortgage commitments by LVR
Source: RBNZ.
42. Ultimately, the proposed Auckland investor LVR restriction is expected to dampen
housing transaction volumes by rather less than 13 percent, allowing for some policy leakage and substitution to less leveraged property investors. Policy leakage is likely to occur as some property investors restructure their finances in order to meet higher deposit requirements. A reasonable estimate based on previous experience with the LVR speed limit suggests that around a quarter of investors may be able to do this. We further assume that the policy changes will encourage some less-leveraged investors to enter the market, offsetting around a quarter of the decline in investor purchases.
43. After taking into account these offsetting effects, the Reserve Bank estimates that
tighter restrictions on investors will reduce Auckland housing market transactions by around 7 percent over the first year following implementation. This would represent around a 20 percent decline in property purchases by investors in Auckland.
44. The Bank expects that there may be a small additional dampening effect on property
transactions via a marginal reduction in owner occupier activity in Auckland. Presently banks are known to preference lending to owner occupiers, especially first home buyers, in allocating high-LVR lending within the existing national speed limit. With the proposed policy change, banks will be required to restrict high-LVR lending to Auckland owner occupiers to no more than 10 percent of Auckland owner occupied lending. This is expected to reduce the share of high-LVR lending to owner occupiers by around 2 percentage points, resulting in a further decline in housing market transactions of around 1 percent.
45. The aggregate expected impact is an 8 percent reduction in housing market turnover
in the Auckland property market. Based on the model described in Bloor and McDonald (2013), a fall in house sale volumes of this magnitude would normally be

associated with a 2.5 percent reduction in house price inflation over the following 12 months.
46. There is a range of uncertainty around both the expected impact on transaction
volumes, and how this will transmit through to house prices. Based on the model estimates, and previous experience of LVR restrictions, we believe the ultimate effect on Auckland house prices is likely to be in the 2-4 percent range. It is likely that most of this impact would occur in the first six months following policy implementation.
47. There is some evidence of rising speculative activity in the Auckland property market,
which may, in part, be based on unrealistic expectations of future capital gains. Where unrealistic expectations exist, a policy response that reduces market momentum for a period of time could help to reset these expectations. As such, the proposed policy could act as a circuit breaker that results in a larger impact on house price growth over time than the Reserve Bank’s modelled estimates suggest.
48. Outside of Auckland, the policy package would allow the high-LVR share to rise by
around 5 percentage points, possibly by more if existing high-LVR lending has been weighted more heavily towards Auckland. We estimate that this could increase ex￾Auckland house price growth by around 1 percent. In addition, there could be an additional stimulatory effect if Auckland-based property investors start looking further afield to continue building property portfolios. The Reserve Bank has not attempted to quantify this effect, which may be particularly concentrated in close proximity to Auckland – for example Hamilton and Tauranga. Outside of Auckland, house sales are expected to increase by around 4 percent.
49. Given offsetting effects inside and outside Auckland, we do not expect a dramatic
effect on credit growth. We expect new commitments to fall by 3-4 percent, resulting in a 1 percent fall in net credit growth. The overall expected impact on macroeconomic aggregates is summarised in table 2 below.
Table 2: Expected policy impact over first year
Auckland Ex-Auckland New Zealand
House sales -8% +4% unchanged
House price growth -2-4% +1% -0.5-1%
New housing loan commitments -3-4%
Housing credit growth -1%
50. The proposed policy also reduces average customer LVRs, particularly with respect
to Auckland investor lending. Lower average LVRs would be expected to make the banks more resilient to a downturn in borrower repayment capacity for any given level of capital.
51. Although not a direct aim of macro-prudential policy, which focuses on promoting the
soundness of the financial system, the proposed policy package would also tend to promote the resilience of household balance sheets, thereby helping to minimise damage to the wider macroeconomy should the Auckland housing market fall materially. Weak household balance sheets cause two significant problems for the financial system in a severe downturn. First, it would tend to exacerbate the extent of the downturn, as firesales can arise when a significant proportion of borrowers are pushed into a forced sale situation. Second, the ability of the economy to recover from the initial shock can be significantly constrained if a high proportion of households are in a position of negative equity and simultaneously act to reduce debt

levels. Reducing the risk of firesales and the risk of household sector distress also enhances bank resilience.
52. Over time, the proposed policy would reduce the number of investors with debt in
excess of 70 percent secured on Auckland property. The Reserve Bank estimates the share of investor loans on bank balance sheets with an LVR of greater than 70 percent will fall from around a third currently to more like a quarter one year after the implementation of the new restriction. This would materially reduce the proportion of loans in a position of negative equity in the event of a 30-40 percent decline in Auckland house prices.
53. Banks have always been prepared to lend at somewhat higher LVRs to owner
occupiers than to investors, reflecting the issues discussed above. The Reserve Bank considers that it is reasonable to regard 70 percent as a high LVR for investor lending in an exuberant market, and that a typical investor loan with an LVR of 70 percent is riskier than a typical 80 percent owner occupier loan in a downturn. Kelly (2014)’s results discussed above are consistent with this.
54. The Central Bank of Ireland has used similar analysis of the Irish GFC experience to
conclude that 70 percent is an appropriate level for calibrating their investor LVR speed limit. In contrast, LVR restrictions implemented in Hong Kong, Israel, Malaysia and Singapore have featured tighter restrictions for non-owner occupied property, with some of these limits as low as 50-60 percent. Q2: Do you have any comments on the objectives or expected impact of the policy? Unintended consequences
55. There is some risk that the proposed changes to the policy could put pressure on
rents or impede the supply of new housing in the Auckland region.
56. The Reserve Bank believes that the effect on rental inflation will be limited. Over time
there is likely to be some reduction in the supply of rental property, in line with a relative shift from investor to owner occupier purchases. However, this transition will also result in a reduction in demand for rental properties. There could be some upward pressure on rents if this transition results in fewer people occupying each dwelling. However, any effect on occupancy is not expected to be large.
57. The LVR policy already includes an exemption for mortgage lending to fund the
construction of new dwellings. This exemption will continue to apply under the new policy settings, and is available for both investors and owner occupiers.
58. A condition of the construction exemption is that lending must be committed prior to,
or at an early stage of, construction. This condition was applied to ensure that exempt lending funds true additions to the supply of housing, rather than distorting demand for newly built houses.
59. The Reserve Bank has not seen any evidence that restricting the exemption in this
way is unnecessarily impeding the supply of new housing. However, this could be revisited if there was evidence to suggest that expanding the scope of the exemption would make a material difference to housing supply.
60. One consequence of a more targeted approach to policy is that it will distort investor
decisions to some extent, which will entail some efficiency costs. For example, it is

likely to encourage some Auckland investors to switch to purchasing property outside of the Auckland region to take advantage of lower deposit requirements.
61. There are likely to be some costs incurred by the banking system to enact the
necessary system changes in order to meet the new speed limits. Most of the costs are likely to relate to identifying lending to property investors, changes that are already required in relation to the new property investor asset class. There may be further costs incurred to ensure that bank systems are properly configured to identify lending in the Auckland region, and to meet new data reporting requirements.
62. While this policy is likely to create some distortions, the Reserve Bank judges that
these are not overly large when compared to the benefits of a material reduction in the probability of financial stress arising as a result of a severe housing downturn in the Auckland region. Q3: Do you think that the framing of the construction exemption is still suitable with the proposed policy changes? Q4: Please quantify costs in relation to implementing this proposed policy change, excluding changes required to identify property investor lending for BS2A/BS2B purposes. Specific policy details
63. This proposed policy change would be enacted through changes to the Banking
Supervision Handbook document “Framework for Restrictions on High-LVR Mortgage Lending” (BS19) and changes to bank conditions of registration. This consultation document has been released alongside a proposed redraft of BS19. The proposed conditions of registration are in the appendix to the proposed BS19. Calculation of speed limit
64. The proposed policy change splits the existing speed limit on high-LVR mortgage
lending into three separate speed limits. These speed limits would apply to the following categories of residential mortgage loans:
 Auckland property investment loans at LVRs of greater than 70 percent divided by all non-exempt Auckland property investment loans.  Auckland non-property investment loans at LVRs of greater than 80 percent divided by all non-exempt Auckland non-property investment loans.  Non-Auckland residential mortgage loans at LVRs of greater than 80 percent divided by all non-exempt non-Auckland residential mortgage loans.
65. For each of these speed limits, current exemptions from the policy would continue to
apply.
66. Thought was given to expressing speed limits with respect to a common base, for
example all non-exempt mortgage lending. However, it was felt that doing so would provide less exact targeting and would unfairly disadvantage banks with a larger share of lending in Auckland. Q5: Do you have any comments on the proposed speed limit definitions?

Calibration of speed limits
67. The introduction of LVR restrictions via speed limits recognises that in some cases it
is appropriate for banks to provide loans at higher LVRs. Allowing banks to provide a small proportion of high-LVR loans allows banks to fund purchases for more creditworthy first home buyers and allows for special circumstances, such as extra financing required for property remediation for earthquake damage or leaky buildings.
68. The Reserve Bank sees less need to allow a flow of lending at higher LVRs to
property investors, and the intent of the policy is to essentially curtail this form of lending. However, the Reserve Bank sees some benefit in retaining a small speed limit. Under a zero limit, even a single loan to a residential property investor at an LVR of greater than 70 percent would represent a breach of a bank’s conditions of registration. Such a breach could occur if an administrative error inadvertently resulted in a loan being granted.
69. Further issues could arise with loans that have been approved prior to the
introduction of the proposed policy, but have not yet reached the commitment stage. Our expectation is that banks would actively manage their pipeline of lending to ensure the vast bulk reaches the commitment stage prior to the implementation date, but there is a possibility that certain transactions involving long property settlement periods may not be processed in time.
70. There may also be special circumstances surrounding property remediation and loan
restructures under borrower hardship that warrant loans to be granted outside of normal limits. A small speed limit is likely to be more effective in allowing for these special circumstances than providing extra exemption categories.
71. The Reserve Bank proposes a 2 percent speed limit in order to allow for these
special cases.
72. For other lending in the Auckland region we continue to view 10 percent as an
appropriate speed limit.
73. Outside of Auckland there is little evidence of exuberance in property markets, which
suggests an easing of LVR policy restrictions is appropriate. However, there is a possibility that there could be a resurgence in activity as restrictions are eased. This risk is accentuated by continued low interest rates and high rates of immigration. For these reasons, it is proposed that restrictions be lifted gradually outside of Auckland. An initial increase in the speed limit to 15 percent will allow the market impact to be assessed, and further easing in policy is envisaged if house price pressures remain subdued. Q6: Is the proposed speed limit for Auckland property investment loans suitable to achieve desired objectives? Q7: Do you have any comments on the calibration of other speed limits?

Definitions
74. The proposed changes require new definitions in BS19 to define ‘Auckland’,
‘Auckland loan’ and to align with the changes made to BS2A and BS2B to subcategorise residential mortgage loans as loans to property investors or loans to owner-occupiers (non property-investment residential mortgage loans).
75. For the purposes of this policy, an Auckland loan is a residential mortgage loan (or
portion of a loan, see below) that is secured on residential property that falls within the boundary of the Auckland council.
76. It is possible that there may be some instances where missing information within
bank systems create ambiguity as to whether the underlying collateral is in the Auckland region.3 Where such ambiguity exists and cannot be rectified, the collateral should be treated as being within Auckland.
77. Property investment residential mortgage loans and non-property investment
residential mortgage loans will shortly be defined in BS2B for internal ratings based banks, and in BS2A for other banks. This definition was released for consultation on Friday 29 May. These definitions will be imported into BS19, and for new loans the amounts in each category will be the same as under the Capital Adequacy Framework.
78. An owner occupied residential property is a property that meets the following criteria:
 a natural person or related party of a natural person owns the property and is the obligor under the loan;  that natural person or their spouse intends to occupy the property either as their principal or secondary residence (a secondary residence includes holiday homes or a second home where the natural person or related party spends significant time); and  in the case of a secondary residence, no rental income is derived from that property, except to the extent that the rental income is minimal (e.g. a bach that is rented for 6 weeks a year).
79. This definition of owner occupied property is intended to treat situations where a
property owner earns some rental income from boarders or flatmates on the owner’s primary residence as owner occupied lending.
80. A related party includes:
 A trustee of a trust of which the person is the beneficiary;  A company the person owns or controls;  Spouses (including civil unions and de facto relationships) (and the estate of spouses)

3 For example if bank systems only capture postcode, rather than region of lending, given that there isn’t a strict concordance between postcodes and administrative boundaries.

81. Appendix A provides a set of example scenarios with guidance on how they would be
    classified for the purposes of BS19.
    Q8: Do you have any comments on these proposed definitions?
    Mixed loans and calculation of new commitment values

82. The new loan value, or change in loan value is the new commitment that should be
    recorded for LVR measurement purposes. This is straightforward where the entire loan belongs in one of the three speed limit classes (Auckland non property￾investment loan, Auckland property-investment loan, or non Auckland loan). This would be the case if the collateral was (for example) multiple Auckland rental properties.

83. Where there is mixed collateral (between rental and owner occupied properties) BS2
    allows the loan to be split between the property investor and non-property investor asset classes. The same loan allocation (for LVR calculation) is applied in BS19. BS19 also allows for loans to be split in a similar way between Auckland and non￾Auckland collateral. In each case, the requirement is that the loan is attributed to each property, proportionately to the relative value of the properties. This effectively means that the LVR is calculated at the portfolio level and is hence the same for each portion of the loan. While it is possible to split loans, it is not necessary (if they are not split, a loan secured over both an Auckland and non-Auckland property should be entirely treated as an Auckland loan, or a loan secured over an owner occupied property and investor property should be treated as a property investor loan). Providing the ability to split introduces some complexity, but seems important to avoid providing an incentive to divide collateral (for example by splitting banking arrangements).

84. Where there is mixed collateral and banks are choosing to split the total loan into two
    or more of the speed limit classes, the total new commitment must also be split. For a new loan (a loan that is not secured over existing collateral), the loan is split based on the relative values of the properties. Where the loan value of the loan increases, only the increase in the loan value is a new commitment, so only the increase in the loan is relevant for the LVR speed limits. The Reserve Bank proposes to define the method for splitting increases in loan values as follows:
     Determine the LVR of the portfolio as the total loan divided by the total value of property, after any collateral revaluation and removal or additions of collateral.  Calculate the total increase in loan value (the total new commitment).  Apportion the total new commitment to any new collateral up to the point where the new commitment attributed to the new collateral is equivalent to the new collateral value multiplied by the LVR (or the entire new commitment has been attributed).  Apportion any remaining new commitment to existing collateral proportionately to the values of the existing collateral.

85. While there are a few steps in this process, it does not appear to be particularly
    complex. Furthermore, banks will be able to avoid the need to do any apportionment by simply placing the total loan into a single speed limit class, and the disadvantages of doing so are ameliorated by the combined collateral exemption discussed below.

86. However, the Reserve Bank has also considered a simpler option. This would be:
     Calculate the total increase in loan value (the new commitment)  Apportion the new commitment across the loan collateral proportionately to the values of that collateral.

87. Where a loan has a mixture of collateral including an Auckland investment property
    and the portfolio LVR is above 70 percent, the portion of the loan treated as an Auckland investment property loan will be high-LVR. However, if the customer took out separate loans separately collateralised against each property, the customer would be able to borrow up to 70 percent on the investment property and 80 percent on the other collateral (e.g. by using multiple banks or formally splitting collateral). This may result in a higher portfolio LVR. To avoid creating incentives to split collateral, the combined collateral exemption (in section 13 of BS19) is designed to allow loans to be exempted in these cases. In effect the exemption will apply where the total borrowing is less than 70% of the value of the Auckland rental property collateral plus 80% of the value of other residential mortgage collateral. Q9: Do you agree that banks should be able to split loans with combined collateral across speed limit classes? Q10: What is your view on the proposed approach to splitting new commitments (and the simpler alternative also suggested)? Q11: Does the combined collateral exemption appear appropriate? Do you have any comment on how it has been designed? Measurement periods and transition arrangements

88. The Reserve Bank proposes that the policy changes take effect from 1 October

89. This relatively long notice period is to allow banks to make the necessary
    systems changes in order to properly classify new lending.

90. There is a risk that a notice period of this length could lead to some Auckland
    property investors rushing in to beat the policy changes. However, our expectation is that banks will observe the spirit of the proposed restrictions, and will act to curtail lending at LVRs of above 70 percent to Auckland property investors well in advance of 1 October.

91. Currently, compliance with the LVR policy is measured over a three-month rolling
    window for banks with monthly lending of over $100m, and over a six-month rolling window for banks with monthly lending of less than $100m.

92. At the time that LVR restrictions were first introduced, all banks were provided with
    an initial six-month measurement period. This was done to accommodate outstanding pre-approved loans, and recognised the relatively short notice period provided. A longer first measurement period does not appear to be warranted for this change to the restriction, given more than four months’ notice of an intention to change the restriction. Further, the low speed limit for Auckland property investment mortgage lending does not provide much scope to smooth lending over a longer measurement period.

93. These rolling windows present complications at the point where the calibration of
    speed limits changes. This is mostly a problem in cases where policy is tightened, as immediately assessing compliance with the new calibration in the following month would essentially bind on lending prior to the new restriction taking effect.

94. We propose that the existing LVR restriction apply to lending committed until 30
    September 2015, with compliance assessed based on the measurement period ending on this date. The new speed limits will take effect from 1 October, but compliance with these new limits will not be assessed until the end of the first measurement period – either 31 December 2015 for larger banks, or 31 March 2016 for smaller banks.

95. This switch over means that there will be a short period in which compliance with the
    policy will not be assessed. This does not appear to be a material issue, as any lending that takes place over this period will be captured in the first measurement period. Q12: Will the proposed implementation timeframes and transition arrangements create any significant difficulty? Q13: Is the $100m boundary for monthly lending for having shorter (3 month) speed limit periods still appropriate? Residential mortgage lending that is not in the residential mortgage asset class

96. The boundary of BS19 is lending that is classified in BS2A and BS2B as a residential
    mortgage loan. This definition excludes some lending that is secured by residential property but is classified in another (often corporate) asset class. This lending tends to take two forms. The first is lending that is predominantly for business purposes, but may be partially secured by residential mortgage collateral. The second is loans to large scale property investors, who banks manage on an individual basis as business customers.

97. When LVR restrictions were initially implemented, this second class of lending wasn’t
    considered to be of concern, as banks reportedly rarely offered LVRs of over 80 percent to this class of borrower. Anecdotal evidence suggests banks tend to prefer lending to these borrowers at an LVR of below 70 percent, although this doesn’t appear to be a hard limit.

98. Once the proposed restriction on Auckland property investors is in place, we expect
    that banks will discontinue lending at LVRs of above 70 percent to property investors with Auckland collateral regardless of whether the loan is classified as a residential mortgage loan or a corporate loan. There are a number of ways that this could be achieved in BS19.

99. Option 1 is to redraw the boundary of BS19 to capture any lending that is secured on
    residential property. However to do so is likely to entail significant compliance costs and may also affect business lending which is partially secured on residential property. This lending is outside the intended scope of the restriction.

100. Option 2, our preferred option, is to treat this as an avoidance issue to be addressed
     through the ‘anti-avoidance’ provisions of BS19. This is likely to be supported by data requirements in respect of this form of lending. Should evidence emerge that the

corporate asset class was being used as an avoidance channel, we would revisit the case for redrawing the BS19 boundary. Q14: How much lending has been originated in the past six months to residential property investors that is primarily secured on residential property and is in the corporate or some other non-residential mortgage asset class? What proportion of this lending is at LVRs of above 70 percent? Q15: Is it possible to provide regular reporting on this form of lending? Q16: Do you have a preference over the two proposed options, or do you consider that another approach would adequately address the risk identified? Anti-avoidance
100. The Reserve Bank is aware that customers who buy a property with the
declared intention of owner occupation may not always end up owner occupying the property. Similarly, they may buy a new home intending to sell their existing one, but ultimately end up renting it out. This may mean a customer obtains an owner occupation new commitment but ultimately turns out to be an investor. The Reserve Bank wishes to avoid imposing excessive costs on banks, but has a concern that misrepresented intentions could be used by borrowers to avoid LVR restrictions.
101. Banks should shift customers from the owner occupier to investor asset class
for capital purposes if they become aware that a declared owner occupier is no longer an owner occupier. We anticipate that banks would inquire as to the owner occupancy of the property during credit events. We are also interested in whether loan documentation typically requires borrowers to tell banks if the owner occupancy status changes, and whether it would be appropriate for this to become standard practice.
102. It would be possible for BS19 to treat loans transitioning to the investor asset
class as representing a new commitment in the period they are shifted (they would then be included in the speed limit calculation in that period). However, this could create some difficulties for banks in managing the speed limit, as they may not be able to effectively control the proportion of their customers that transition categories in each period.
103. An alternative definition of new commitment would only include customers
that are transitioning between categories if their loan met certain initial conditions that made it appear relatively likely they could be seeking to misrepresent their intentions. For example, customers that meet criteria such as the following could be flagged as ‘questionable’ and the loans would then count as new commitments if reclassified:
 The loan is at an LVR of above 70 percent, and the declared intention is to owner occupy  The bank was aware that the borrower owned two or more properties at the time that the new mortgage was committed, whether or not the bank holds mortgages against those properties.
104. Another policy option would be to allow reclassification without requiring
reclassified loans to count as new commitments. The Reserve Bank would be

uncomfortable doing this if reclassifications from owner occupier to investor occurred frequently. If the policy initially operates this way (as currently drafted in the attached BS19), the Reserve Bank considers it would need to monitor reclassification through regular reporting by banks. Conditions relating to broking loans and second mortgages
105. The existing standard conditions of registration applying LVR restrictions also
impose requirements on banks to not allow second mortgages (or provide first mortgages where there is a second or subsequent charge) in most circumstances. The Reserve Bank’s initial intention is to retain these conditions in the revision of the LVR policy. However, the Reserve Bank notes that second mortgage activity is rare within the banking sector in New Zealand and that isolated errors could cause banks to inadvertently breach these conditions without having much effect on the economic impact of LVR restrictions. For these reasons, the Reserve Bank may in the future consider shifting these obligations into the anti-avoidance section of BS19. This may be considered, for example, as part of the more general regulatory stocktake. Q17. Should BS19 require that multiple properties that a customer has provided as collateral be grouped as a single collateral set, or take some other approach to customers with mixed collateral? Q18: How should the risks of the ‘intention to owner occupy’ being misrepresented by customers be handled? How often do banks typically become aware that customers who initially said they would owner occupy are not actually owner occupying? Would it be reasonable to expect banks to count loans that change category as new commitment in some circumstances? Is there another way of dealing with this avoidance risk? Q19: Do loan contracts typically require borrowers to inform banks if their owner occupancy status changes? Would it be appropriate for this to become standard practice (if it is not already?) Q20: Would there be a material risk of second mortgage activity if the relevant conditions of registration were replaced by an anti-avoidance clause? Q21: Are there other material avoidance risks that need to be addressed? Data requirements
106. The proposed changes will necessitate some adjustments to the new
commitment survey to monitor compliance. We will liase with banks separately on the necessary changes. Timeline and next steps
107. The consultation period for these proposals will run until 13 July. Following
that, the Reserve Bank expects to release a summary of submissions and final policy position in early August, with the policy taking effect from 1 October 2015.

References
Bloor C and C McDonald (2013), “Estimating the effects of restrictions on high LVR lending”, Reserve Bank of New Zealand Analytical Note, AN2013/05. Coates D, R Lydon and Y McCarthy (2015), “House price volatility: the role of different buyer types”, Central Bank of Ireland Economic Letter 2015/2. Fitch ratings (2012) “Investment Versus Owner Occupied: Divergence in Mortgage Performance across Australia”. Haughwout A, D Lee, J Tract and W van der Klaauw (2014), “Real estate investors, the leverage cycle, and the housing crisis”, Federal Reserve Bank of Ney York Staff Report, 514. Kelly R (2011), “The Good, The Bad and The Impaired: A Credit Risk Model of the Irish Mortgage Market”, Central Bank of Ireland Research Technical Paper, November. Kelly R (2012), “House prices, Unemployment and Irish Mortgage Losses”, Paper presented to De Nederlandsche Bank Conference “Household finances and behavior in times of crisis”, October. Palmer C (2014), “Why did so many subprime borrowers default during the crisis: Loose credit of plummeting prices?”, Working paper, http://faculty.haas.berkeley.edu/palmer/cpalmer-subprime.pdf Price G, “How has the LVR restriction affected the housing market: a counterfactual analysis”, Reserve Bank of New Zealand Analytical Note, AN2014/03. Shiller R (2006) "Long-term perspectives on the current boom in home prices." The Economists' Voice 3.4

Appendix A: Potential Scenarios for Homebuyers
Scenario How will LVR restrictions apply?
1 Buyer owns and occupies a house in Auckland (with LVR at 80%) and wishes to buy a rental property in Auckland Buyer will probably need at least a 30% deposit for the purchase (possibly more depending on bank policies). 2 Same as scenario 1 (but LVR on residence is only 30%) Buyer can probably obtain equity by increasing mortgage on primary residence and then purchase rental property with a low LVR loan. 3 Buyer owns and occupies a house in Auckland (with LVR at 80%) and wishes to buy a bach in the Coromandel, which won’t be rented out. Buyer may be able to obtain 80% or even higher LVR loan on the Coromandel property. 4 Buyer owns and occupies a house in Auckland (with LVR at 30%) and wishes to buy a rental property in Auckland (fully secured over the OO home, not the rental property). The amount borrowed would be the equivalent of 110% of the rental property’s value but would take the LVR on the house to just 70%. If loan is genuinely unsecured over the rental property, this is a simple low-LVR Auckland non￾investor loan and should be available. 5 An owner-occupier (with LVR of 80%) is posted overseas and now wishes to rent out the property Bank will need to reclassify loan for capital purposes if it is aware property is now rented out. But this will not be a new commitment. 6 Buyer lives in Auckland city Monday to Friday and wants to buy a property in Pukekohe to live in at weekends Both properties will be owner-occupied - investor LVR restriction does not apply 7 Investor has portfolio of 4 rental properties in Auckland with an existing pooled LVR of 75%. They wish to buy another rental property in Auckland. How much can they borrow? At existing bank, will probably need to keep LVR across all properties after the purchase below 70%. 8 Same as scenario 9 but one of the rental properties is part of a mixed purpose building (e.g. industrial building) which This mixed purpose building is unlikely to meet the criteria for a residential mortgage loan. It will probably be treated separately rather than being grouped with the other loans for LVR purposes.

the owner will be using as a small business.
9 Purchaser buying first home with a 2 bedroom flat underneath (intends to rent flat to help pay the mortgage). This meets the definition of owner occupation, so would be classified as a non-property investor loan. 10 Buyer owns a portfolio of rental properties (value $1m) outside of Auckland with a combined LVR of 60%. What is the maximum LVR that applies if we also wish to purchase a $1m rental property in Auckland? Bank may be prepared to lend up to a combined LVR of 75% (using the combined collateral exemption).