A speech delivered to the Chamber of Commerce in Rotorua
Introduction
The housing market is a key driver of economic and financial developments in New Zealand. It is watched closely by the Reserve Bank as a barometer of domestic demand and an indicator of financial stability.
Over the past 12-18 months, Reserve Bank policy has helped to moderate pressures in the housing market. Increases in the OCR totalling 1 percent between March and July 2014 were intended to reduce the degree of monetary policy stimulus to domestic demand while the loan-to-value ratio restrictions (LVRs), introduced in October 2013, were aimed at reducing risk in the financial system. These policies were mutually supportive of each other and slowed the rate of growth of house prices during the first half of 2014, despite strong net inward migration and the ongoing shortage of housing.
Since late 2014, housing market imbalances have become more accentuated, particularly in Auckland where the supply shortage is greatest. This is of concern to the Reserve Bank as it poses a threat to financial and economic stability. House prices as multiples of income and rents are already near record levels and further increases will accentuate the risk of a sharp price correction once demand and supply come in to closer balance.
Today I will discuss these housing market imbalances: their causes and the risks they present. I will also touch on policies that could assist in reducing the current imbalances. But first, let's recap housing market developments over the past year.
Housing pressures moderated in 2014
House price inflation nationwide accelerated from mid-2012 and peaked in the third quarter of 2013, supported by rapid growth in high-LVR lending by the banks and record low mortgage interest rates. The Reserve Bank became concerned that strong house price inflation and rapid growth in high-risk lending was increasing the risk of a sharp housing market correction at the same time that bank balance sheets were becoming more vulnerable. In response, the Reserve Bank imposed a limit on high-LVR lending effective from 1 October 2013. These restrictions, the subsequent tightening of monetary policy and election uncertainty helped to slow housing demand and ease house price inflation through to September 2014.
House sales slowed markedly in the immediate aftermath of LVR restrictions, with all regions trending down except Christchurch. The slowing was evident for all buyer types, although there were shifts in the composition of sales. There was generally a larger fall in sales to first-home buyers and existing owner occupiers. Sales to investors remained relatively stable while ‘new to market' and ‘returning to market' sales increased – reflecting purchases by new migrants and returning New Zealanders.
Figure 2.1: Monthly house sales by buyer type
House price inflation slowed from an annual rate of 9.3 percent in September 2013 to 4.9 percent in September 2014. The slowing in house price inflation was most evident in Auckland, with a more muted effect elsewhere. Reserve Bank research suggests that, nationally, house price inflation would have been 3-5 percent higher without the LVR policy.
Figure 2.2: Annual house price inflation (three-month moving average)
Actual and anticipated increases in the OCR from March to July 2014 increased the cost of mortgage credit and contributed to a slowing in mortgage commitments from $4.5b per month at their peak to $4.2b in September 2014.[1] The share of lending at high LVRs fell from over 30 percent in mid-2013, to a low point of 3.6 percent in March 2014.
From late 2014, however, there has been a pick-up in new mortgage lending with around half of new mortgage commitments at LVRs between 70-80 percent. The increase in new lending has not led to significant growth in bank balance sheets, due to many households taking advantage of low interest rates to accelerate their repayment of principal.
Figure 2.4: New housing loan approvals, new commitments and credit growth
New housing market pressures emerging
Since September, coincident with the upturn in new mortgage lending, housing market turnover has increased sharply (figure 3.1). Some of this reflects the normal spring pick-up, and possibly a bounce back from pre-election uncertainty. Nonetheless, the data point to a strong resurgence in housing demand, particularly in Auckland.
Figure 3.1: Monthly house sales
(Seasonally adjusted)
New listings onto the market have not kept pace with the increase in house sales, causing inventory levels to decline from already low levels and adding to house price pressures. This is true for most regions, but is most pronounced in Auckland where current inventory would be exhausted in only 12 weeks at the current level of sales. Consistent with this, annual house price inflation in Auckland has been increasing, reaching nearly 17 percent in March. House price inflation has remained more moderate elsewhere, although price levels remain stretched in most regions.
What has caused this resurgence in the housing market? The answer is not straightforward. Several demand and supply factors are at play.
Population growth over the past year has contributed to increased demand for housing. Net permanent and long-term (PLT) migration inflows reached a record rate of over 55,000 in the year to February 2015 – boosting the population by over 1 percent. While strong growth in migrant arrivals has pushed up the net immigration numbers, there has also been a significant slowdown in the number of New Zealanders leaving the country, currently at its lowest level for over a decade.
It is estimated that over half of the net migrant inflow have settled in Auckland. This is providing a major boost to Auckland's population, which grew by nearly 34,000 people in the year to June 2014 (figure 3.2). Although consent issuance of new dwellings in Auckland has continued to increase, reaching 7,700 in the year to February 2015, this has been insufficient to match the surge in population, let alone cut into Auckland's housing shortage.
Figure 3.2: Auckland annual net population change
(rolling quarterly totals)
People-per-dwelling is a useful metric for looking at the physical balance between housing supply and demand. Over time, people-per-dwelling should slowly trend down as the population ages, families shrink and people become wealthier. This means that the number of dwellings required will tend to increase at a faster rate than the population. If residential building is insufficient to keep up with population growth, then the number of people-per-dwelling will trend up rather than down.
In Auckland, people-per-dwelling has been increasing since 2008 while it has continued to trend down elsewhere in New Zealand (figure 3.3). The sharp rise in people-per-dwelling in Auckland in 2014 highlights that home building remains well below the level required to accommodate the rising population. We estimate the shortage of houses in Auckland has increased over the past year to between 15,000 and 20,000 houses.[2]
Figure 3.3 People per dwelling
An increase in investor demand in Auckland can be inferred from various indicators: we have seen a continued decline in the rate of home ownership, which reached a record low of 61.5 percent in 2013; there has been an increase in the share of house sales going to investors since October 2013 – from 33.8 percent in September 2013 to 37.4 percent in February 2015; and rental inflation has remained considerably below house price inflation. Average rental yields in Auckland have fallen from 4.6 percent in 2010 to 3.7 percent in 2014. Residential investors have continued to expand their interest in housing as rental returns have declined. This points to an excess demand for home ownership over and above the simple excess demand for accommodation.
New housing supply in Auckland is falling short of expectations
Turning to the other side of the equation, it is clear that housing supply constraints are persisting. National residential construction activity has been increasing steadily for the past few years and, in constant prices, is now in line with the previous peak of 2004. Much of the increase has been due to earthquake repairs and rebuilds in Canterbury, which are now beginning to ease the accommodation shortage that has existed since the earthquakes. We have seen a substantial easing in Christchurch rental inflation over the past year. About 10,000 rebuilds or major repairs are yet to be completed[3] and residential building in Canterbury is expected to remain around its current high levels for the next couple of years. After that, however, we are likely to see a gradual tailing off of residential construction in Canterbury.
Figure 4.1: Annual new dwelling consent issuance by region
Auckland's outlook is very different. Residential construction will need to continue at a high pace for many years. At current rates of building, it appears the shortage of housing is getting worse, not better. Auckland Council estimate that over 10,000 new homes each year will be needed to satisfy future population growth.[4] In essence, it could take many years to work off the current housing shortage.
The Auckland Accord is a step in the right direction, and improves the outlook for further growth in housing supply. However, while the target for the first year (9,000 dwelling and section consents) was comfortably met, targets are more challenging in the next two years. The target for the final year to September 2016 is an ambitious 17,000 new dwelling and section consents.[5]
To help facilitate faster growth in home building, 80 Special Housing Areas (SHAs) have been established. These include brownfield and greenfield areas with an estimated capacity for 43,000 dwellings over a period of 10 years. So far, SHAs have yielded few completed dwellings or "shovel-ready" sections. However, 11,000 new dwelling or section consents in SHAs are expected by September 2016.
While the Accord is helping, significant factors continue to hold back new housing supply in Auckland. These include a limited supply of land for housing, a restrictive urban planning regime, public concerns over densification, uncertainty around future infrastructure development, and a fragmented and inefficient residential building industry. These factors were discussed in the Productivity Commission's 2012 inquiry into housing affordability. Although many of the constraints apply nationally, they are more binding in Auckland due to the stronger pressures from population growth.
An improved supply of useable land is essential. The land component of housing costs in New Zealand has grown considerably over the past 20 years with land now accounting for more than 60 percent of the total cost of a new dwelling in Auckland, and around 50 percent in the rest of the country.[6] The Auckland Plan establishes a Rural Urban Boundary that provides scope for urban growth outside of the old Municipal Urban Limit. However, land blocks with disparate ownership and speculative investment in ‘land-banking' continue to tie up large areas of buildable land.
Urban planning rules are complex and often restrictive. Planning must take account of the Resource Management Act, Local Government Act and Land Transport Act. Planning requirements can add significantly to the cost of a new dwelling and/or restrict the volume of new supply. A recent study[7] estimates the cost of planning regulatory requirements to be between $32,500 and $60,000 per dwelling in a subdivision and between $65,000 and $110,000 for an apartment.
The relatively high cost impost for apartments is due to a raft of height, area and other planning restrictions that appear to have contributed to the relatively low number of apartment consents in Auckland over recent years (fig.4.2). Auckland also has a relatively low proportion of apartments compared to Sydney: 25 percent versus 39 percent of total dwellings. While apartment construction has picked up recently, the evidence suggests that more active encouragement of apartment development could be an effective means of easing the housing supply constraint in Auckland.
Figure 4.2: Annual consent issuance in Auckland
Providing the infrastructure for new developments is a slow and complex process, in part due to multiple service providers with often conflicting incentives and funding arrangements.[8] Better collaboration and incentives across infrastructure providers is required for more efficient infrastructure planning and a better delivery of services.
Overall delays and uncertainties throughout the development process can have significant implications for the total cost of development and ultimately whether proposed developments proceed. The planning approval process to convert designated SHA land to ‘shovel-ready' sections requires multiple approvals[9] and is estimated by the Council to typically take over two years to complete in greenfield situations. For existing residential zoned (brownfield) areas, the time to achieve all approvals for multi-unit dwellings is in excess of a year.[10]
Further, capacity constraints in the construction sector could constrain further increases in housing construction in Auckland over the next couple of years. The needed pick-up in Auckland housing supply comes on top of the significant ongoing demands of the Canterbury rebuild, which are not expected to ease for at least two years. Also, employment in the construction sector now accounts for 6 percent of the working age population, similar to that seen in the mid-2000s property boom. Over the past year, vacancies in the construction industry have increased and construction firms have been finding it increasingly difficult to find skilled labour. Construction cost inflation in Auckland increased to 7 percent in the year to December 2014 (figure 4.3), highlighting the increasing resource pressure.
Figure 4.3: Annual construction cost inflation by region
Housing pressures are a threat to stability
Rising house price inflation, particularly in Auckland, represents a risk to financial and economic stability. Irrespective of the mix of demand and supply-based factors, the longer excess demand persists, the further prices will depart from their underlying fundamental determinants and the greater the potential for a disruptive correction. New Zealand's house prices, when compared to incomes or rents, are high on an international basis and very high when compared to New Zealand's historic trend. Figure 5.1 shows that New Zealand's house price-to-income ratio in 2013 was 30 percent above its historical average. Only Norway had a higher historical deviation.
Figure 5.1: International house price-to-income ratios
(Deviation from historical average, 2013)
Auckland house prices are particularly stretched, having increased by three times since the start of 2002. This stretch is emphasised by the recent Demographia affordability survey that shows Auckland to be the ninth least affordable city out of 86 major metropolitan areas (population over 1 million)[11]. In September 2014, the median Auckland house sale price was 8.2 times the median household income, compared to a survey median ratio of 3.8. Nationally, New Zealand's median house price-to-income ratio is about 5.0.
New Zealand is one of the few advanced economies that has not had a major house price correction in the past 45 years. During this period, most of the advanced economies have had at least one significant house price correction that has created difficult challenges for financial and economic stability. Several countries saw 20 percent plus house price declines in the GFC, including the US, Ireland, Spain, Greece and the Netherlands.
A correction in house prices in New Zealand could be prompted by a range of potential shocks to the economy or financial system. Global interest rates are at record low levels and large mortgages have been ‘affordable' at these rates. As global interest rates return to more normal levels, many mortgage borrowers could come under pressure as they are required to refinance at higher rates. Alternatively, a downturn in the global economy and financial markets could lead to a drop in national income and rising unemployment, at the same time as foreign creditors are requiring an increase in the interest rate premium charged to New Zealand borrowers. In such circumstances, we could see the cost of credit rising at the same time that incomes and employment were under pressure. Such stresses would be expected to cause housing demand to ease. If this occurred when new supply was coming through in volume, then prices would begin to fall. A withdrawal of speculative interest in residential property or decline in migration inflows would accentuate any such fall.
With 60 percent of its lending in residential mortgages, the New Zealand banking system could be put under severe pressure in such scenarios. The resulting contraction in credit would amplify the impact of an adverse external shock to the domestic economy and financial system, making it more difficult to avoid a severe downturn.
What can policy do?
Given the large physical shortage of housing in Auckland, and the prospect that population growth will outstrip new supply for some time yet, a primary issue to address is the slow supply response in Auckland. While progress has been made expanding the supply of land available for residential development and improving the resource consent process via the Auckland Accord, considerable scope exists to streamline the multiple approval processes required to complete a residential development. This could include: introducing greater flexibility and efficiency into urban planning processes; a requirement for transparent cost-benefit analyses of urban planning regulations; and a more integrated approach to the planning and funding of infrastructure development.
The Government's review of the Resource Management Act (RMA) has the potential to substantially improve the planning and resource consenting processes. Also, the Productivity Commission is currently exploring ways to improve the policies, strategies and processes that underpin New Zealand's housing development capacity, including the provision and funding of infrastructure.[12]
Other measures to facilitate residential development might include a review of taxes and other incentives affecting land banking. Measures to increase productivity in the construction sector would also help facilitate larger-scale and more affordable housing development.
While the housing supply constraint must be the main policy priority, it is clear that, under any scenario, it will take a number of years to eliminate the housing shortage. In the meantime, it is important that population-based demand pressures are not exacerbated by expanding investor and credit-based demand that simply amplifies the price effects of the physical shortage. Increases in house prices above their already high levels are unlikely to bring forth further supply; rather they make houses even less affordable and add to financial system risk.
An important policy question on the demand side is whether measures could be taken to stem the high net inflow of migrants. However, there are real risks in trying to manage the migration cycle. Any measures taken will have long lags and therefore run the risk of reducing inflows just when the cycle is already turning down. Further, any such measures can only operate on gross inflows of non-residents. Policy cannot influence the volume of gross outflows or returning New Zealanders.
The tax treatment of housing is a major factor with potential to influence the demand/supply imbalance in the housing market. As reflected in our submission to the Productivity Commission's inquiry on housing affordability, housing is the most tax-preferred form of investment, particularly when it is highly leveraged. Investors are often setting the marginal market prices that are then applied to the full housing stock within a regional market. Indicators point to an increasing presence of investors in the Auckland market and this trend is no doubt being reinforced by the expectation of high rates of return based on untaxed capital gains. While there are difficult issues and trade-offs to consider in this area, the Reserve Bank would like to see fresh consideration of possible policy measures to address the tax-preferred status of housing, especially investor related housing.
Mortgage credit growth is facilitating the renewed expansion in housing demand. While net growth in mortgage credit remains moderate, the flow of new mortgage lending has increased sharply since October 2014. Banks are competing aggressively by reducing margins on fixed mortgages and new mortgage commitments are growing at a 20 percent annual rate. We estimate that at least half of this new credit is going to the Auckland market.
The Bank considers using macro-prudential policy in the housing area on two potential grounds:
1) to counter new credit-based demand and price pressures that are increasing financial system risks; and/or
2) to increase the resilience of the banking system to a potential housing downturn.
The existing LVR restrictions on mortgage lending helped to moderate the market for a year and they continue to constrain risk in the banks' balance sheets. With the current overheated housing market in Auckland, it would not make sense to remove these restrictions in the near term. To do so would invite a further surge in credit-based demand for housing. When conditions warrant, however, it will be appropriate to ease or modify the incidence of the LVR policy.
The Bank is currently consulting on a new asset class for residential investment mortgages that will attract a higher risk weighting than owner-occupier mortgages. This is consistent with the international evidence showing larger loan losses for investor loans than for loans to owner occupiers during periods of housing market stress. A new asset class will see a greater consistency of treatment for investor mortgages across the banks and ensure that investor loans have capital backing consistent with their higher risk. A well-defined asset class would also facilitate the introduction of macro-prudential policy and the Bank is currently reviewing some options.
Of course the Reserve Bank's most powerful policy tool is the OCR. However, while the pressures in the housing market would suggest a tightening of interest rates, the primary objective of monetary policy, CPI inflation, is projected to remain below its target range for some time. Thus it would be inappropriate for monetary policy to lend assistance to the Bank's financial stability objective at this time.
This presents something of a policy conundrum, resulting from the current very unusual international environment. Persistently low global inflation and recent declines in commodity prices are acting to soften CPI inflation in New Zealand. At the same time, the low global inflation is causing central banks around the world to hold interest rates at historically low levels and, in the case of Europe and Japan, to adopt large quantitative easing programmes. The easy monetary policies are in turn contributing to upward pressure on global asset prices, including real estate, and increasing risks to financial stability.
Conclusion
After moderating in the period following the introduction of the LVR restrictions, housing market pressures re-emerged in the fourth quarter of 2014. The market imbalance is considerably greater in Auckland than in other parts of the country, with house price inflation now at 17 percent per annum and median price-to-income multiples over 7. The increasing degree of stretch in prices means that an eventual market correction is increasingly likely to be disruptive to financial stability and the economy.
The causes of the housing market imbalance are primarily linked to the rapidly expanding migrant population and a relatively slow housing supply response. The latter has been a longer-term issue and is due to a complex range of factors including: a scarcity of suitable land; restrictive planning regulations; a lack of coordination in infrastructure planning; and a fragmented and inefficient building industry. There have also been bottlenecks in the construction sector with both Christchurch and Auckland competing for resources. An acceleration in new credit commitments since September has facilitated the growth of house sales and prices over this period.
Supply-side policy solutions are unlikely to yield quick results. The proposed RMA reforms could significantly improve the planning process but could take years to be felt in new supply. Greenfield developments will inherently be slow to deliver new stock. A release of resources from the Christchurch rebuild is likely to be a couple of years away. The best prospect for substantially increasing the supply of dwellings over the next one to two years appears to be in apartment development. The Government and the Auckland Council might therefore consider focussing their efforts on streamlining the approvals process and increasing the designated areas for high-density residential development.
On the demand side, we consider that greater attention needs to be given to issues relating to the tax treatment of investor housing. There are practical difficulties in using migration policies to influence the housing cycle. Monetary policy, which in past cycles has been supportive of financial stability objectives, cannot be used currently to dampen housing demand. This cycle is unusual in that CPI inflation is staying very low - requiring interest rates to also stay low - even though rising asset prices are raising financial stability concerns.
Macro-prudential policy is a potential instrument to help restrain credit-based demand pressures and/or improve the resilience of bank balance sheets in the face of a potential housing downturn. The LVR restrictions had a significant moderating effect in 2014 and are still acting to improve the resilience of bank balance sheets. They will be removed or modified as market conditions allow. Other macro-prudential options are being assessed, including in relation to investor lending. However, such tools are not a panacea – their impact is inevitably smaller than the main drivers of the current housing market imbalance.
References
Auckland Council (2013) ‘Shaping a business friendly city. The Proposed Auckland Unitary Plan' December 2013.
Auckland Council (2014) ‘Housing Delivery Cycle' prepared by the Housing Project Office, December 2014.
Grimes, A and I Mitchell (2015) ‘Impacts of Planning Rules, Regulations, Uncertainty and Delay on Residential Property Development' Motu Working Paper 15-02, Motu Economic and Public Policy Research, Wellington.
Ministry of Business, Innovation and Employment (2014) ‘Auckland Housing Accord, Monitoring Report #4'
Ministry of Business, Innovation and Employment (2015) ‘Auckland Housing Accord, Monitoring Report #5'
Productivity Commission (2012), ‘The housing affordability inquiry' New Zealand Productivity Commission Final Report, March 2012
Productivity Commission (2014), ‘Using land for housing' New Zealand Productivity Commission Issues paper, November 2014
Watson, E (2013) ‘A closer look at some of the supply and demand factors influencing residential property markets' Reserve Bank of New Zealand Analytical Note, 2013/11
[1] Seasonally adjusted, and measured on a three-month moving average basis.
[2] This is based on an assumed trend of 2.9 people-per-dwelling. If the assumed trend for people-per-dwelling in Auckland was downward sloping, as we have seen in the rest of New Zealand, then the estimated shortfall of houses would be larger.
[3] Based on data from the Insurance Council of New Zealand
[4] Auckland Council (2013).
[5] The split between the dwelling and section consents is expected to be around 10,000 new dwelling consents and 7,000 new section consents.
[6] Productivity Commission (2014).
[7] Grimes and Mitchell (2015)
[8] In Auckland for example, the Auckland City Council, Watercare, Vector, NZTA and developers themselves are all involved in the provision of basic infrastructure.
[9] Including resource consent for a development master plan, applications for plan variations, engineering approvals for civil works and associated resource consents, survey plan approvals, and the approval of individual land titles.
[10] Auckland Council (2014).
[11] 11th Annual Demographia International Housing Affordability Survey, Jan 2015
[12] The inquiry is also looking into ways to improve governance, transparency and accountability of the planning system, and the involvement and engagement with the community. A draft report is due in June 2015.