Update on the New Zealand housing market

Release date
09/05/2014
Speaker
Grant Spencer
Main file

A speech delivered to Admirals' Breakfast Club

Introduction

Housing is a key sector of the New Zealand economy, and an important factor influencing the Reserve Bank's monetary and financial policies. And of course housing is never far from the front pages.

Since 2012 we have seen rapid rates of house price inflation, reflecting a demand – supply imbalance. There has been a physical shortage of houses, particularly in Christchurch and Auckland. On the demand side there has been strengthening economic growth and inward migration as well as easy credit conditions, in terms of both the price and availability of mortgage finance. His combination of supply and demand factors has contributed to the price pressures we have seen.

The Reserve Bank introduced Loan to Value Ratio restrictions (LVRs) in October with the aim of reducing the systemic risk arising from increasingly overvalued house prices. More recently, the Official Cash Rate (OCR) has been increased from 2.5 percent to 3 percent with the aim of forestalling general inflation pressures in the broader economy. Housing supply conditions have also started to improve with a recovery in residential construction, centred in Christchurch and Auckland. These factors are all working to reduce the housing market imbalance. The one factor that is not helping at present is the strong inward migration flow.

Our overall view is that housing market pressures are easing gradually. Today I will review the progress being made towards a better balance in the housing market and discuss the potential implications for Reserve Bank policies.

Housing Market activity is weakening

After gathering momentum from mid-2012, the housing market looks to have slowed in the six months since LVR restrictions were introduced. The volume of nationwide house sales has dropped considerably (figure 2.1), with the number of sales in March 11 percent lower than in September last year.

Figure 2.1: Monthly house sales (seasonally adjusted)

5724016_files/gs-speech-9-may-201400.jpg

Source: REINZ.

The decline in sales volumes has been reasonably widespread across the country, with the exception of Canterbury where sales were down only 1 percent over the same period. The recent uptrend in Auckland's share of national house sales appears to be flattening out. Comparing market segments, we are seeing a smaller proportion of sales below $400,000. Some of this represents the reduced presence of first home buyers. BNZ-REINZ survey data (Figure 2.2) indicate that first home buyers accounted for 16 percent of sales in Q1 2014, compared to 24 percent a year earlier. The share of existing owner-occupiers has gone up by a similar amount, while the share of sales to investors has remained stable at 19 percent.

Figure 2.2 House sales by purchaser segment

(BNZ-REINZ Residential Market Survey)

5724016_files/gs-speech-9-may-201401.jpg

Source: BNZ, REINZ.

The overall slowdown in housing market activity has also been reflected in house prices. After increasing steadily over 2012 and 2013, house price inflation has slowed since the introduction of LVR restrictions, despite the large increase in inward migration (figure 2.3).

According to REINZ data, New Zealand house prices increased at an annual rate of 8.4 percent in the three months to March 2014, down from a peak of 9.8 percent in the three months to November 2013[1]. The slowing in national house price inflation is largely due to easing house price inflation in Auckland, although this remains at a high annual rate of 15 percent (figure 2.3). Annual house price inflation is at 9 percent for Canterbury, and around 4 percent for the rest of New Zealand. House price inflation is not confined to Auckland and Christchurch. Outside the main centres, some of the stronger regions include Waikato/Bay of Plenty, Nelson/Marlborough and Otago.

Figure 2.3: Annual house price inflation by region*

5724016_files/gs-speech-9-may-201402.jpg

*Three-month moving average

Source: REINZ, RBNZ estimates.

Slower credit has contributed to the easing

Data indicates that all banks have complied with the requirement to keep high LVR lending at no more than 10 percent of total lending over the six month period from October. For the system overall, the share of high LVR lending over the six months was 5.6 percent.

This compares with a 30 percent high LVR share in early 2013. There has been some compensating increase in sub-80 percent LVR lending over the six months as the banks have intensified competition in this segment. However, the overall effect has been a reduction in the rate of growth in new mortgages. (figure 3.1).

Figure 3.1 Housing new loan approvals, new commitments and credit growth (quarterly)

5724016_files/gs-speech-9-may-201403.jpg

* Rolling three month change in housing loans, seasonally adjusted

** Estimates based on private bank reporting from February 2012 to July 2013, LVR monthly survey data from August 2013

Source: RBNZ.

LVR impact has been close to expectations

The moderation in housing market indicators is broadly in line with the expected impact of the LVR restrictions. Table 4.1 shows actual reductions in national sales, price inflation and credit relative to our initial expectations. On a counterfactual basis, modelling work undertaken by the Bank suggests that, had banks continued to maintain 30 percent of their lending at high LVRs, annual house price inflation might be running some 2.5 percent higher than at present.

We are comfortable that the restrictions are so far meeting their objective of helping to restrain the demand for housing while supply gradually catches up. In so doing, we believe the restrictions are mitigating the systemic risk of a housing market downturn that becomes more likely as house prices and debt levels become more stretched. We also believe that the LVR restrictions have helped to make banks' balance sheets more resilient to any adverse housing market shock. I will come back to the LVRs shortly.

Table 4.1 Housing and credit developments since the introduction of LVR speed limits


Expected reduction over first year

Reduction since peak

House sales (monthly)

3-8%

11.5%

House price inflation

(annual percent change*)

1-4%

1.5%

Housing credit growth

(annual percent change**)

1-3%

2.5%

* 3 month moving average, stratified REINZ house price index.

** Annualised three-month percent change.

All figures rounded to the nearest half percentage point, based on seasonally- adjusted monthly REINZ data unless otherwise stated.

The supply gap remains large

Demand is only one side of the story. Supply conditions are a very important influence on housing market outcomes. In its report on housing affordability, the Productivity Commission (2012) identified a number of supply factors – including restrictive urban planning, and the time and costs associated with development and construction – as factors constraining new housing supply in New Zealand. An important constraining factor has been the availability of land, particularly in Auckland. Section prices more than doubled in Auckland between 2000 and 2012 and, for sections within 25 kilometres of inner city Auckland, land prices tripled.[2] As a consequence, land costs now comprise 60 percent of the cost of building a new dwelling in Auckland, compared with 40 percent in the rest of the country.

Low rates of building in Auckland since 2005, combined with ongoing population growth, have led to a physical shortage of houses. Population-per-dwelling is estimated to have risen considerably in Auckland over the past few years (figure 5.1). While the size of the current housing shortage is uncertain, based on data from the 2013 Census of Population and Dwellings, it is now estimated to be between 5000 and 10,000 dwellings. This is lower than earlier estimates, but still material.

Figure 5.1: Population-per-dwelling

5724016_files/gs-speech-9-may-201404.jpg

Source: Westpac, RBNZ estimates.

In Christchurch, a significant shortage of accommodation also exists following the Canterbury earthquakes. Around 10,000 homes need to be completely rebuilt and a further 12,000 require major repairs (greater than $100,000 in value).[3] The physical shortage of housing in Christchurch has been exacerbated by the extra demand for accommodation by construction workers. This has resulted in annual rent inflation of about 11 percent in Christchurch, compared to 3.5 percent in the rest of New Zealand.

The Auckland Accord is helping

Auckland's population is expected to grow markedly over the next 30 years. The Auckland Council estimates that 10,000 new dwellings will be required per year to meet future population increases. This expected future demand for accommodation means it will likely take some time to eliminate the current supply shortage. However, the Housing Accord being implemented by the Government and the Auckland Council is an important step towards achieving a better balance. Under the Accord, consent processes are fast tracked for housing developments within defined Special Housing Areas, consistent with the rules outlined in Auckland's Unitary Plan. The Unitary Plan, developed under the Resource Management Act, sets out the Council's vision for housing development over the next 30 years, but does not come into effect until 2016. The Accord is essentially an interim measure designed to accelerate development ahead of the Unitary Plan coming in to effect.

The Accord aims for 39,000 new building and section consents to be issued over the three years to September 2016, with actual building likely to occur over a somewhat longer period. So far, 63 Special Housing Areas in a mixture of urban and rural locations have been created, with an estimated capacity for 33,500 new dwellings. Over the past year, ‘greenfield' land in the pipeline for housing development (i.e. zoned and with utilities connected) has increased to a level sufficient for about 25,000 dwellings, up 10,000 on a year ago.[4] This is good progress, but acceleration in development activity is needed to meet the Accord and longer-term supply targets.

Construction is up but further boost required

New building in Auckland has increased over the past year, with 6500 new dwelling consents issued in the year to March 2014, compared to 4900 in the previous year. New building activity still remains relatively low and a substantial further ramp up will be required to eliminate the supply shortage while accommodating future population growth.

Figure 7.1: Annual issuance of dwelling consents (March years) 5724016_files/gs-speech-9-may-201405.jpg

Source: Statistics New Zealand.

In Canterbury, residential reconstruction work is now picking up - after a very slow start. To date, most of the earthquake-related residential reconstruction work has related to The Earthquake Commission (EQC) repair program, with around 53,000 residential repairs completed as of April.[5] The required 22,000 larger scale residential repairs and rebuilds (mostly managed by insurers) are now underway but only a small portion has been completed. Dwelling consents in Canterbury reached 6200 in the year to March and we expect to see significant further growth over the coming year. Most of the outstanding shortage should be eliminated over the next three years.

With our two largest cities both seeking to boost housing construction over the next three years, there will be significant pressures placed on building resources nationally. We expect total residential construction to increase by 30 percent over the next three years, well surpassing the historical peak of 2004.

Key risks going forward

There are clearly many moving parts to the overall supply shortage equation – and many risks around the prospects for reducing the imbalance. Probably the major risk at present is the outlook for immigration. After increasing substantially over the past 18 months, net immigration rose to 32,000 in the year to March, in part due to fewer departures of New Zealand citizens. The Reserve Bank forecasts this pattern to reduce gradually as economic conditions improve in Australia, which is the main destination for New Zealanders shifting offshore. However, cyclical turning points are hard to predict and there is a risk that the net migration inflow remains greater for longer, which would underpin the demand for housing.

On the other side of the equation, there is the risk of a continuing slow supply response due to bureaucratic and/or logistical constraints in Auckland and Christchurch. In Auckland more land is being made available but developments need to be approved and implemented, and construction needs to be consented and undertaken in scale. In Christchurch it is apparent that the consenting process has been improved, but risks remain around the slow pace of insurance settlements, particularly in the areas where land damage has been sustained.

A third risk arises from the increasing pressure that will come to bear on construction resources, potentially resulting in further price and wage pressures in the construction sector. As seen in figure 8.1, construction cost inflation in Canterbury is running at around 8 percent per annum and, at a national level, reached 5 percent in the March 2014 quarter. Annual construction cost inflation in Auckland, which was running at below 2 percent a year ago, has accelerated to 6 percent. Rising construction cost inflation is a threat to housing affordability as well as to the overall rate of CPI inflation.

Figure 8.1: Annual construction cost inflation

5724016_files/gs-speech-9-may-201406.jpg

Source: Statistics New Zealand.

Monetary policy perspective

From a monetary policy perspective, the Bank has two concerns: first that construction cost pressures could feed into generalised price inflation; and second that house price increases could cause households to increase their spending, reducing savings and putting additional pressure on overall domestic demand. The OCR increases that commenced in March are aimed at countering emerging inflation pressures in general, but their success, or otherwise, in moderating housing related pressures will be key.

Based on current market interest rate expectations, adjusted for an average mortgage interest margin, floating mortgage rates could be 7 to 8 percent in two years' time, up from current levels of around 5 to 6 percent. While this would be a significant move, figure 9.1 indicates that it would take mortgage rates back to their average of the past 20 years. That is, the outlook is towards a normalisation of mortgage rates which are currently near to 50 year lows. We should not expect to see historically low interest rates at a time when the New Zealand economy is growing strongly.

Having said that, the timing and extent of interest rate increases required over the coming period will depend on a number of uncertain variables. A big uncertainty is the future path of the exchange rate, which has a major bearing on traded goods prices and overall economic activity. The more downward pressure that the exchange rate exerts on prices and activity, the less pressure will need to be exerted by interest rates.

Figure 9.1 Floating mortgage interest rate*

5724016_files/gs-speech-9-may-201407.jpg

*Forecasts based on market OCR pricing (modelled) as at 16 April 2014 and average spread of floating interest rate above OCR over the past 5 years.

Source: RBNZ.

The other big area of uncertainty is around housing. Will inward migration continue to underpin housing demand? Will the dampening impact of the LVR restrictions be long lasting? Will the supply response soon start to have a moderating effect on house prices? Will home owners increase their spending out of capital gains? And how will rising mortgage rates affect spending by borrowers?

With regard to this last question, we expect that interest sensitivity will be greater now than during the last tightening phase in 2003-2007. First, most New Zealand mortgages are currently on either floating or relatively short-term fixed rates whereas in 2003-2007 the bulk of mortgage borrowers were on 2-to-4 year fixed rates. This means that OCR increases will be felt more immediately by existing as well as new mortgage borrowers. Second, a higher starting level of mortgage debt relative to incomes means that any given interest rate increase will have a larger impact on monthly debt service payments, relative to incomes, than in the previous cycle. This latter point is shown in Figure 9.2 where the debt service ratio for a typical new borrower (with a 20 percent deposit) is projected to rise more sharply, relative to the interest rate track, than in the previous cycle.

Figure 9.2 DSR for a representative new entrant to the housing market

5724016_files/gs-speech-9-may-201408.jpg

* new entrant to the housing market purchasing a house at the average residential sales price, earning the average household disposable income, financed using a 25 year mortgage at an 80% LVR. Interest rate forecasts are based on market OCR pricing as at 16 April 2014 and an average spread above OCR over the past 5 years.

Source: RBNZ.

Financial policy perspective

The Reserve Bank's concerns expressed through last year, and motivating the LVR restrictions introduced in October, related to financial stability risks arising from highly geared house purchases at potentially unsustainable prices. The high LVR lending through that period was increasing the likelihood of a house price correction by pushing house prices higher, and also increasing the vulnerability of banks to such a correction.

As discussed earlier, the LVR restrictions have had a moderating effect on the housing market as well as reducing the proportion of high LVR loans in bank balance sheets. We believe that the LVR policy is achieving its purpose of improving the safety of the financial system. The market overall is less vulnerable to the effects of an adverse shock – from international financial markets for example. And banks are now less exposed to potential credit losses as the interest rate cycle turns upwards.

Nevertheless, there remain a significant volume of high LVR loans on bank balance sheets – in the order of 20 percent – and, as seen in figure 9.2, debt service ratios are likely to increase substantially over the coming two years. This could put some stress on highly-leveraged borrowers and their lending institutions.

We have stated clearly that the LVRs are temporary restrictions and will be eased or removed when housing market pressures have abated and when we are confident that a resurgence of high LVR lending will not occur following removal. The LVRs are supporting monetary policy but they will have a reduced effect over time. Also, the efficiency costs of such restrictions will tend to increase the longer they are kept in place. Before removing the LVRs, however, we will want to be confident that the housing market is responding to interest rate increases; and that immigration pressures are not causing a resurgence of house price pressures. It will take some time to gain this assurance. At this stage we consider the earliest date for beginning to remove LVRs is likely to be late in the year.

Conclusion

Housing market and credit cycles are part of the economic landscape and, although they can be moderated through sound policy, they cannot be eliminated. When pressures become excessive the cycle can be damaging for the financial system and for the broader economy. This is why we monitor housing closely and why housing is a key factor influencing our monetary and financial policies.

House price pressures in New Zealand since late 2012 have been driven by supply shortages and by strong demand supported by easy credit.

On the supply side, progress is being made with national dwelling consents now up to an annual rate of 22,000. In Auckland, progress is being made in freeing up the supply of buildable land and improving the consenting process. And in Canterbury, the replacement of severely damaged homes is well in train after a slow start. However, the shortage remains large – with around 50,000 new homes targeted in Christchurch and Auckland over the next three years. This is a task that is likely to take considerably longer than three years to complete.

On the demand side, pressure has been moderated by the LVR restrictions and will be further restrained as monetary policy brings interest rates back to a level more consistent with a strongly growing economy. Interest rate increases are expected to have a greater effect than in the previous cycle due to the high levels of existing debt and a high proportion of floating rate mortgages. However, many uncertainties remain around the housing market outlook, particularly regarding the future impact of the increasing net migration inflow.

Over their first six months, the LVR restrictions have had a dampening effect on housing that is broadly in line with expectations. The easing or removal of the LVR restrictions will depend importantly on the restraining impact of interest rate increases and any renewed pressure arising from the net immigration.

Achieving greater stability in the housing market will contribute greatly to maintaining financial and price stability, and to achieving sustainable economic growth.

References

Canterbury Earthquake Recovery Authority (CERA) (2014) ‘Greater Christchurch Recovery Update – Issue 30', March,
cera.govt.nz/sites/cera.govt.nz/files/common/greater-christchurch-recovery-update-issue-30-march-2014.pdf, accessed 28 April 2014.

Fletcher Earthquake Recovery (EQR), ‘Completed repairs',
www.eqr.co.nz/repairs/completed-repairs, accessed 28 April 2014.

Ministry of Business, Innovation and Employment (2013) ‘Residential land available in Auckland', Research Report, March,
www.dbh.govt.nz/UserFiles/File/Publications/Sector/pdf/residential-land-available-in-auckland-report.pdf, accessed 14 April 2014.

Ministry of Business, Innovation and Employment (2014) ‘Auckland Housing Accord, Monitoring Report #1', March,
http://www.beehive.govt.nz/sites/default/files/Auckland_Housing_Accord_Monitoring_Report_1.pdf, accessed 14 April 2014.

Productivity Commission (2012), ‘The housing affordability enquiry', New Zealand Productivity Commission Final Report, March 2012.

[1] The house price data from Property IQ gives a more lagged picture, but has slowed down even more than the REINZ data, suggesting the change in sales composition may have caused the REINZ index to somewhat understate the slowing in house prices.

[2] Ministry of Business, Innovation and Employment (2013), ‘Residential land available in Auckland'

[3] CERA (2014), ‘Greater Christchurch Recovery Update – Issue 30'

[4] Ministry of Business, innovation and employment (2014), ‘Auckland Housing Accord monitoring report #1'

[5] Fletcher EQR: https://www.eqc.govt.nz/