Housing risks require a broad policy response
A speech delivered to New Zealand Institute of Valuers in Wellington
New Zealand is experiencing a housing market boom. House prices are increasing at 13 percent per annum nationally, and at 15-20 percent in Auckland and close-by regions. Evidence from housing cycles in several advanced economies suggests that the longer this continues, the more likely there will be a severe correction.
The Reserve Bank is mandated to promote the soundness and efficiency of the financial system. Our concern is that a severe housing correction would pose real risks for financial system stability and the broader economy. The banks are heavily exposed to housing with mortgages making up around 55 percent of total assets. Household debt, at 163 percent of household income, is at a record level.
Many domestic and international factors are contributing to the strength of the market. The current record low interest rates are a world-wide phenomenon linked to post-GFC caution and very low inflation in the global economy. Also driving local housing demand has been an unprecedented net migration inflow over recent years reflecting New Zealand’s stronger economic performance relative to many other advanced economies.
While strong demand has been underpinned by low interest rates, rising credit growth and population increases, the housing supply response has been constrained by planning and consenting processes, community preferences in respect of housing density, inefficiencies in the building industry, and infrastructure development constraints. The resulting housing market imbalance has been exacerbated by New Zealanders’ on-going preference for investment in bricks and mortar over financial investments, due in part to the ready availability of credit and a tax system that favours debt funded capital gains.
Given the complexity of factors underlying the housing situation, there is no simple policy solution. We need to tackle housing on many fronts. The key challenge in the long run is to expand housing supply to meet the growing demand. The Reserve Bank has no direct influence over supply, but can influence housing demand through the credit channel. The Bank’s interest rate policy is targeted primarily at keeping future CPI inflation between 1-and-3 percent on average over the medium term, although it must also have regard to the soundness and efficiency of the financial system. The Bank’s other relevant instrument is macro-prudential policy. This can improve the resilience of banks’ balance sheets on a lasting basis and help restrain credit and housing demand, at least for a period.
Today I will describe recent developments in the housing market and the imbalances that appear to be widening across the country. I will touch on the range of policies that could help to address these imbalances and focus on Reserve Bank policies that might assist. Specifically, I will review the macro-prudential policy options that the Bank is considering, with a view to possible implementation over the months ahead.
Initial loan-to-value (LVR) restrictions
The Reserve Bank first introduced a broad 80 percent LVR restriction in October 2013, in response to growing housing market pressures and an increasing proportion of high LVR lending from 2011. Then, in November 2015, a resurgence of pressures in Auckland led the Reserve Bank to tighten the LVR restriction to 70 percent for Auckland investors. The LVR restriction was eased somewhat outside of Auckland as housing risks in the regions had not increased significantly since the implementation of the initial LVR policy.
The Reserve Bank also established a new residential property investor class for banks. Loans in that class now attract a higher risk weight than for owner-occupier mortgages, requiring banks to fund such loans with a higher proportion of equity.
The LVR restrictions have strengthened bank balance sheets against housing market shocks. The share of banks’ exposures to riskier mortgages has fallen across a variety of borrower types. Nationally, new investor lending at LVRs above 70 percent has fallen by around one third following the 2015 policy changes. It is now largely impossible to borrow more than 70 percent against an Auckland rental property. More broadly, the share of high LVR (+80 percent) mortgages on bank balance sheets has continued to trend downwards. High LVR loans now account for 12 percent of banks’ residential mortgage exposures, compared to around 21 percent just prior to the initial introduction of LVR restrictions in 2013 (Figure 2.1). This amounts to a reduction in high LVR (+80 percent) lending of around $20bn. Over time, these balance sheet trends will help to reduce banks’ losses on riskier loans in the event of a downturn in the New Zealand housing market. In so doing, it would facilitate a continued flow of credit through a downturn.
Figure 2.1: Stock of high-LVR mortgages (% of total bank mortgage lending)
The LVR restrictions were expected to have a temporary impact on house price inflation, house sales and housing credit. Auckland house price inflation and house sales fell more than expected in the months immediately following the introduction of the new restrictions in November 2015. However, the housing-related government tax changes, together with reduced offshore interest due to new requirements for IRD numbers, probably also restrained housing demand over this period. Our assessment is that the 2015 LVR policy changes brought about a 2-4 percentage point reduction in Auckland house price inflation over a six month period. This price impact is similar to that assessed for the initial LVR restrictions in 2013.
Recent worsening of housing imbalances
New Zealand house price inflation began to accelerate again from around March 2016 as demand pressures intensified in Auckland. In the meantime, other regions were contributing to higher national house price inflation from mid-2015, particularly those areas adjacent to Auckland. Most regional centres are now experiencing annual house price inflation in excess of 8 percent (Figures 3.1, 3.2). Similarly, sales activity increased across the country in the first half of 2016. Reflecting the underlying housing shortage, new listings have remained flat. Listings as a proportion of sales are now 40 percent below the previous low seen at the height of the pre-GFC boom in 2007.
Figure 3.1: Annual house price inflation (3 month moving average, s.a.)
Auckland house price inflation in excess of income growth has seen the median house price to income ratio grow to 9.7 times in the most recent Demographia survey, making Auckland the fourth most expensive city relative to income out of 367 cities worldwide. This ratio is at an historical peak, having doubled since the early 2000s. Outside Auckland, house price to income ratios in most centres are around 4 to 6 (Figure 3.2). However, if house prices in the regions continue to grow at current rates, those ratios will worsen. Overall, New Zealand house prices relative to incomes are 32 percent above their long run average, and the second highest in the OECD. The IMF estimates that New Zealand house prices are around 20-40 percent overvalued based on long run affordability metrics.
Figure 3.2: House price inflation and price-to income ratios by urban area
Increased housing demand has been driven by record net immigration, low mortgage interest rates and increasing investor participation. Net migration flows continue to hit new records, with annual net PLT migration now approaching 70,000 persons. Migration flows are also becoming more dispersed, with non-Auckland numbers recently rivalling the flow into Auckland. Mortgage rates declined over 2015 as the Reserve Bank eased interest rates in light of on-going low CPI inflation. One and two year rates are now in the low 4 percent range, which New Zealand has not seen since the 1950s.
A dominant feature of the housing market resurgence has been an increase in investor activity (Figure 3.3). In recent months, investors have accounted for around 43 percent of sales in Auckland and 38 percent in other regions. Investor borrowing has maintained much of its momentum even though it is taking place at somewhat lower average LVRs. The prospect of capital gains appears to remain a key driver for investors in the face of declining rental yields.
Figure 3.3: New Zealand house sales by buyer type (number of dwellings, s.a.)
The declining affordability of New Zealand housing and increasing investor presence have seen a downward trend in the share of households owning their own home. This ratio has fallen steadily since the early 1990s, reaching 64.8 percent at the 2013 Census. The recent increase in investor housing activity suggests that the home-ownership rate may have declined further since 2013.
The Reserve Bank considers that rising investor participation tends to increase the financial stability risks relating to the household sector in severe downturn conditions. Evidence from the UK and Ireland shows mortgage default rates significantly higher for investors (at any given LVR). There are likely to be a variety of reasons for this, but an important one is that owner-occupier households need to move out of their own home if they default, giving a powerful incentive to continue servicing their mortgages if at all possible. Investors do not face the same incentive for their rental properties and are also more likely to face income shocks (like rental vacancy) at the same time that house prices fall.
Our view on the riskiness of investor lending is shared by other banking regulators. In recent years concern about investor lending has led the Australian Prudential Regulation Authority to limit growth in Australian bank lending to residential investors, the Basel Committee on banking supervision has recently recommended significant increases in the amount of capital held against investor mortgage lending, and the Bank of England has proposed rules that effectively tighten lending criteria for investor mortgages.
Despite high rates of debt repayment, housing credit increased by 8 percent in the year to March 2016, its highest growth rate since 2008. New mortgage commitments are also elevated, running at an annual rate of 35 percent of outstanding housing debt. With high rates of churn in mortgage books, recent increases in interest-only lending and high debt-to-income lending could rapidly affect the quality of banks’ overall mortgage portfolios, tending to dilute the improvement generated by the LVR restrictions. Interest-only and high debt-to-income (DTI) lending to housing investors has increased considerably more than for owner occupiers.
Persistent housing credit growth in excess of income growth has caused the household debt-to-income ratio to grow steadily since 2012. At 163 percent this ratio now exceeds the previous peak reached during the GFC. There is a clear risk that on-going high house price inflation could lead to a further deterioration in the household debt-to-income ratio. While low interest rates have helped to contain debt-servicing ratios (DSRs) for New Zealand as a whole, high and rising debt levels leave households very vulnerable to any future increases in interest rates or deterioration in economic conditions. While a large increase in mortgage rates is unlikely in the current global environment, at today’s high debt-to-income ratios, a relatively small increase in interest rates could put significant pressure on some borrowers. This is particularly the case in Auckland, where DSRs for new buyers are close to 50 percent (Figure 3.4). A 1 percentage point rise in interest rates for these new buyers would boost the proportion of income devoted to mortgage servicing by around 5 percentage points.
Figure 3.4: Representative Buyer DSRs (annual, % of average gross household income)
On the other side of the housing market is the inability of new supply to meet the growing demand, particularly in Auckland. Although Auckland annual housing consents recently reached an 11 year high and are some 20 percent above last year in value, the number of residential building consents per capita in Auckland is currently around 40 percent of the peak level achieved in the mid-2000s. We estimate the shortage of houses in Auckland has increased over the past year and may now be in the order of 20,000-30,000 houses. Furthermore, the overall housing shortfall is expected to increase further as supply is growing more slowly than demand. More encouraging is the higher-intensity building development that is now contributing a greater share of new dwelling consents. Apartments, townhouses and other attached dwelling types accounted for 42 percent of new residential building consents in Auckland over the past two years, compared to 28 percent nationally.
Figure 3.5: New Zealand housing demand and supply (annual)
Note: this is based on an assumed average of 2.6 persons per household across New Zealand.
Outside Auckland, housing supply has been more responsive to population growth. Consent issuance has been relatively strong in regions that have experienced the fastest rates of population growth. Supply in the Bay of Plenty has been particularly responsive, with an elasticity of consents to population growth that is twice the size of Auckland’s (Figure 3.6). International evidence from the GFC suggests that elastic housing supply responses can dampen house price cycles by reducing the degree of housing scarcity, and price growth, in the up phase of a housing cycle.
Figure 3.6: Regional building consent elasticities
A broad policy response is needed
On-going supply-side constraints combined with rapid growth in the demand for housing continue to create significant imbalances in the housing market, particularly in Auckland. In order to relieve those imbalances, a range of initiatives is needed to increase the long-term housing supply response and moderate housing demand. The relevant policy areas extend well beyond financial policy and the responsibilities of the Reserve Bank. In this regard, we see the Reserve Bank as part of a team effort.
The Productivity Commission’s Report, Using land for housing, released in October 2015, notes that an insufficient supply of land ready for new housing remains at the heart of the rise in house prices seen in Auckland and other high growth areas of the country. Costly and restrictive planning rules and regulations, insufficient provision of infrastructure required to make land ready for development, and a planning system unable to respond adequately to changing demand patterns were among the factors the Commission cited as contributing to the shortfall in housing-ready land.
The Commission proposed a range of measures for the Government and local councils. The proposals include setting expectations for councils about the availability of land for development, creating a system to measure development capacity, providing options to intervene when development capacity falls short of what is required and improvements around the financing and provision of infrastructure. In respect of the latter, the recent Government initiative to establish a Housing infrastructure Fund will help to relieve an important constraint.
The Government has also recently released its proposed National Policy Statement (NPS) on urban development capacity which requires councils to ensure sufficient land supply to meet projected housing and business development needs. Councils will be required to monitor developments around housing affordability, building and resource consents and the value of land on urban boundaries. They will be required to ensure price competition, coordinate infrastructure development and streamline consenting processes. The NPS appears consistent with recommendations made by the Productivity Commission and should facilitate a greater supply of housing-ready land over time.
A major supply-side milestone will be the outcome of the Auckland Unitary Plan (AUP) Independent Hearings Panel, which is expected to make final recommendations to the Auckland Council by 22 July. The Panel’s recommendations and the Auckland Council’s response to those recommendations will be crucial in setting the future path towards reducing the housing market imbalance.
As a means of gaining further traction on the supply side, the Reserve Bank supports the idea of a framework for establishing Urban Development Authorities (UDAs). These would be Crown or local body entities, with powers to assemble and acquire land, accelerate planning and consent processes and oversee coordination of all the parties involved in housing development. Such entities could also potentially undertake the funding and development of infrastructure. UDAs have been used in various forms in a number of countries including Australia, the UK and the US to facilitate major housing developments in designated areas. However, by their nature, the role and powers of UDAs would be contentious and their creation would have to be carefully managed.
While boosting the capacity for development and housing supply is paramount, it is also important to explore policies that will keep the demand for housing more in line with supply capacity. Two areas for on-going consideration include tax and migration policy. On the tax front, the implementation of the bright line test for housing investors introduced in October last year has helped curb short-term speculative activity in the housing market. Consideration might be given to further reducing the tax advantage of investing in residential housing.
Like taxation of investor-owned housing, migration policy is a complex and controversial issue. However, we cannot ignore that the 160,000 net inflow of permanent and long-term migrants over the last 3 years has generated an unprecedented increase in the population and a significant boost to housing demand. Given the strong influence of departing and returning New Zealanders in the total numbers, it will never be possible to fine-tune the overall level of migration or smooth out the migration cycle. However, there may be merit in reviewing whether migration policy is securing the number and composition of skills intended. While any adjustments would operate at the margin, they could over time help to moderate the housing market imbalance.
What role can the Reserve Bank play?
Low domestic interest rates have contributed to the increasing housing demand. Under the Policy Targets Agreement (PTA), the 1-3 percent inflation target range is the central focus of the monetary policy framework. However, the Bank must consider whether its monetary policy choices could undermine the efficiency and stability of the domestic financial system. In current circumstances, the PTA rules out actively leaning against the housing cycle using monetary policy. Doing so would risk driving CPI inflation below the target range over the medium term. Conversely, further reductions in the OCR could pose a risk to financial stability through their effect on credit growth and house prices. While the outlook for CPI inflation will ultimately determine the future path of monetary policy, the trade-off against financial stability risk needs to be carefully considered.
The Reserve Bank’s primary policy instrument for promoting financial stability is prudential policy. The Reserve Bank’s baseline prudential policies require banks to hold permanent capital and liquidity buffers against adverse shocks that might occur in a severe business cycle downturn. We also have macro-prudential policies that impose additional safety requirements in periods when there is heightened risk of an extreme housing cycle. Such policies are relevant for the New Zealand banking system where residential mortgages make up around 55 percent of banking system assets and where a large share of total housing credit is to residential investors who have a relatively high risk profile.
The main macro-prudential tools include loan to value ratios (LVRs); debt-to-income ratios (DTIs) and higher capital requirements for housing loans (capital overlays). These three instruments tackle housing cycle risk from different perspectives and can be seen as complements. The LVRs, which are already in place, help to reduce the impact of mortgage defaults on bank earnings by increasing the security coverage on housing loans. LVRs also tighten up banks’ lending conditions, potentially leading to a slow-down in credit growth and house price inflation for a period.
Debt-to-income ratios have been used internationally but not yet in New Zealand. DTIs aim to improve the safety of borrowers’ balance sheets, thereby reducing the likelihood of mortgage defaults in a downturn. In particular, debt-to-income limits are intended to better equip borrowers to continue servicing mortgages in the face of income losses and/or increases in interest rates. DTIs, like LVRs, tighten credit conditions, resulting in some brake on credit and house price inflation.
Capital overlays, like LVRs, improve the capacity of banks to absorb losses from mortgage defaults in a downturn. Additional capital requirements might also slow credit growth as banks adjust to higher equity funding.
How might these macro-prudential instruments assist in the context of the current housing market imbalances?
The original 2013 LVR restrictions (requiring a maximum 80 percent LVR for most mortgage borrowers) and the Auckland investor LVRs brought in last year (reducing Auckland investor LVRs to a maximum of 70 percent), have had the intended effect of making bank balance sheets more resilient to a potential housing downturn. Together with the Government’s October 2015 tax measures, the Auckland investor LVRs also helped to reduce Auckland house price inflation from its peak of 27 percent pa in September 2015 to 12 percent in May 2016.
As discussed earlier, we have recently seen spill-over effects in the regions close to Auckland and a more general resurgence in housing market pressures across the country. The proportion of sales to investors nationally has grown from 34 percent in January, to 39 percent in May this year. LVRs on new investor lending have reduced significantly, but new credit commitments to investors have recently been growing about twice as fast as for the overall market. Investors have effectively used equity generated by increased valuations on their existing portfolios to raise the larger deposits needed for new acquisitions.
The Reserve Bank has a range of policy options available. One is tighter LVRs to counter the growing influence of investor demand in Auckland and other regions, and to further bolster bank balance sheets against a housing market downturn. Given the growing housing market pressures across the country, one approach would be to adopt a single national LVR limit for investors. Given that the banks have much of the relevant systems work in place, we expect that such a measure could potentially be introduced by the end of the year.
Another option is a new debt-to-income (DTI) speed limit to complement the LVR requirements by improving the resilience of household balance sheets to income or interest rate shocks. A DTI limit would make defaults less likely in a downturn. Furthermore, a DTI and LVR in combination would constrain credit growth and house price pressures on a more sustainable basis than would LVRs alone. A DTI would be a new instrument that would need to be agreed with the Minister of Finance under the Memorandum of Understanding on Macro-prudential Policy. Adoption would require more analysis and systems preparation than an extended LVR. We intend to consult with the banks on the viability of a DTI policy and data issues before making a decision on implementation.
A third option is a housing capital overlay. The Reserve Bank has already indicated that it will be conducting a full review of bank capital requirements over the coming year. We will consider whether macro-prudential overlays have a role to play as part of that process.
In conclusion, the Reserve Bank is concerned about the risks to financial and economic stability inherent in the growing housing market imbalances. Auckland pressures are re-emerging following an easing in the market from late 2015, and house price inflation has accelerated in a number of regions.
The causes of the imbalances are complex with a number of important drivers on both the demand and supply side. Addressing these imbalances will require policy action by a variety of agencies on a number of fronts. The underlying housing shortage needs to be urgently addressed, particularly in Auckland where population growth continues to outstrip housing construction. A step up in supply is required and finalisation of the Auckland Unitary Plan will be a key opportunity to facilitate such a step.
On the demand side, the key drivers are population growth and easy credit. The low cost of credit is making higher debt levels affordable, particularly for investors who can deduct interest costs from taxable income. Residential investors are accounting for an increasing share of house sales and new mortgage credit.
The Bank’s interest rate policy must have regard to financial stability concerns, but the global environment is likely to keep interest rates low for some time yet. Macro-prudential policy can assist in containing the growing risk to financial stability as the current housing market reaches new extremes. In light of the growing risk, the Reserve Bank is closely considering measures that could be progressed in the coming months.
Cox, W and Pavletich, H. ‘12th Annual Demographia International Housing Affordability Survey. Rating Middle-Income Housing Affordability’ Demographia International Housing Affordability Survey (2016 Edition: Data from 3rd Quarter 2015)
Mohommad, A., Nyberg, D., and Pitt, A. ‘New Zealand: selected issues’. IMF Country Report 16/40, February 2016
Andrews, D., Sanchez, A.C. and Johansson, A. ‘Housing Markets and Structural Policies in OECD Countries’. OECD Economics Department Working Paper No. 836 January 2011
Spencer, G. ‘Action needed to reduce housing imbalances’. Speech delivered to the Chamber of Commerce in Rotorua, 15 April 2015.
1 Cox and Pavletich (2016)
3 Mohommad, Nyberg, and Pitt (2016)
4 For clarity, the estimate of Christchurch’s elasticity (ratio = 2) has been omitted from the chart as the estimate is affected by the post-earthquake spike in building activity.