Long-Run Property Market Trends
in New Zealand and Australia

What history and data tell us about where house prices might head next

Dr Yuqian Zhang · 8 July 2026 · Analytical Brief

1. Executive Summary

New Zealand and Australia have both seen one of the longest housing booms in the developed world over the past 30 years. In New Zealand, real house prices (adjusted for inflation) roughly tripled between 1990 and their 2021 peak, as measured by the Bank for International Settlements (BIS) harmonised index. The index rose from roughly 60 to 178 (2010=100), before starting a correction that has now run for more than two years. Australia, after a smaller post-pandemic dip, returned to positive real price growth by mid-2025.

This brief uses data from the BIS, the OECD, the Reserve Bank of New Zealand (RBNZ), the Reserve Bank of Australia (RBA), the IMF, and academic research to answer four questions. First, are current prices unusually high by historical standards? Second, what sets the NZ and AU housing markets apart from those in other countries? Third, what can past boom-and-bust episodes tell us about what might happen next? Fourth, how likely is each future path for New Zealand?

The central finding is that New Zealand's housing market is undergoing an orderly correction without serious stress on the financial system. This path most closely resembles Sweden's experience from 2022 to 2024, rather than the severe crashes in Ireland (2007-2012) or Spain (2007-2015). Real prices are falling at a moderate pace (roughly 4 percent per year). Bank capital buffers are strong. Credit gaps are negative, meaning lending is not growing too fast. And prudential tools, such as debt-to-income (DTI) limits and loan-to-value ratio (LVR) restrictions, act as guardrails against renewed speculative excess. The most likely outcome is a long period of flat real prices, with nominal prices growing at or below income growth over the next three to five years. The chance of a sharp crash (a total real decline of more than 25 percent) is assessed as low, provided there is no major external shock and prudential discipline is maintained.

2. Long-Run Real House Price Trends

2.1 Fifty-Five Years of BIS Data

The BIS harmonised long series on residential property prices, available from 1970 Q1 for both New Zealand and Australia, provides the best basis for comparing across countries.^[1] The BIS publishes these series with base year 2010=100, adjusted for inflation (deflated by the consumer price index). The data reveal a striking pattern: real house prices in both countries changed sharply around the year 2000, rising much faster after two decades of relative stability.

Between 1970 and 2000, real house prices in New Zealand rose at roughly 2.0 percent per year, broadly tracking productivity and income growth. From 2000 to the 2021 peak, real growth sped up to roughly 4.5 percent per year. The 2000-2021 period saw real house prices rise roughly 150 percent in New Zealand and 110 percent in Australia, far outpacing income growth in both countries.

The post-pandemic cycle has produced a notable split between the two countries. New Zealand real prices peaked in late 2021 (BIS real index roughly 178, 2010=100) and have since fallen by roughly 12 percent in real terms as of Q4 2025 (index at 157.5). The year-on-year real decline sped up to -4.24 percent in Q4 2025, the fifth straight quarter of negative real growth.^[1] Australia's real price dip was smaller (roughly 5 percent real decline from peak) and had fully reversed by Q3 2025, when real prices returned to +2.0 percent year-on-year growth.^[1]

2.2 Policy Regimes and Structural Breaks

New Zealand's housing market has been shaped by several distinct policy regimes since the 1990s. After 1990, the framework of inflation targeting brought structurally lower nominal interest rates. Combined with limits on housing supply, this produced sustained real house price rises. In 2016, the Auckland Unitary Plan replaced the restrictive Metropolitan Urban Limit with a more flexible Rural Urban Boundary, which led to a substantial supply response in Auckland, as shown by Greenaway-McGrevy and Phillips (2023).^[2] But this supply boost was later overwhelmed by the collapse in interest rates and the fiscal stimulus during the COVID era.

3. Valuation Metrics and Deviation Analysis

3.1 Price-to-Income Ratio

The OECD price-to-income ratio for New Zealand (seasonally adjusted, 2015=100) is the most complete single measure of affordability, with data going back to March 1986.^[3] At the December 2021 peak, the ratio reached 142.9, well over two standard deviations above its long-run median of 88.9. As of March 2025, the ratio stands at 105.8, a 26 percent decline from the peak, though it remains above the median.

At 105.8, the ratio is about 0.85 standard deviations above its long-run average. This is high but not extreme by post-2000 standards. For context, the ratio stayed above 100 for every year from 2015 through 2025. This suggests that the period since the Global Financial Crisis may represent a new, lasting norm rather than a temporary blip, driven by persistently lower real interest rates, strong population growth in cities where housing is scarce, and tax rules that favour housing.

3.2 Real Price Deviation from Trend

We apply a one-sided Hodrick-Prescott filter (lambda=400,000, the same approach the BIS uses for credit gaps) to the real house price index. This gives estimates of how far prices stray from their underlying trend. The peak deviation was roughly +38 percent above trend in 2021, and has since shrunk to roughly +8 percent as of 2025.

The current +8 percent deviation is well within the range of past experience, and well below peaks seen before past corrections. However, the HP filter adapts to structural shifts in the data. If the post-2000 era does represent a new norm, the filter treats part of that structural rise as "trend," which may understate how overvalued prices are relative to a fixed pre-2000 benchmark.

3.3 Composite Overvaluation Assessment

Across four valuation measures (price-to-income, price-to-rent, real house price deviation from trend, and household debt-to-income), the average z-score has fallen from roughly 2.44 at the peak to roughly 1.00 now. The market remains above historical norms but no longer shows the extreme overvaluation of a classic bubble. The Demographia International Housing Affordability Survey (2025 Edition) confirms this: Auckland's median multiple has improved from roughly 11 (among the least affordable globally five years ago) to roughly 7.5. Demographia credits this improvement to New Zealand's broader housing supply reforms, including the Auckland Unitary Plan upzoning.^[4]

4. Credit Conditions and Household Leverage

4.1 Mortgage Rates and Monetary Policy

The RBNZ's tightening cycle was among the most aggressive in the OECD: the OCR rose from 0.25 percent (October 2021) to 5.50 percent (May 2023), pushing standard 1-year fixed mortgage rates from roughly 2.8 percent to 7.6 percent by mid-2024.^[5] The easing cycle that followed (nine cuts, OCR to 2.25 percent by November 2025) has brought 1-year fixed rates down to roughly 5.28 percent as of June 2026.

Over 70 percent of New Zealand mortgages were scheduled to reprice during 2025, with many households shifting toward short-term fixed-rate mortgages in anticipation of further easing.^[6] This high share of short-duration fixed-rate debt sets New Zealand apart from markets like the United States, where 30-year fixed-rate mortgages are the norm. It means that changes in monetary policy pass through to household cash flows much more quickly.

4.2 Household Debt and Macroprudential Policy

New Zealand's household debt-to-disposable-income ratio stands at roughly 170 percent, among the highest in the OECD, driven almost entirely by mortgage debt.^[7] Total housing debt reached NZD 388.5 billion as of November 2025, making up 63.8 percent of total private-sector credit.^[8]

The activation of debt-to-income (DTI) restrictions on 1 July 2024 is a structural shift in prudential policy. Banks may lend no more than 20 percent of new loans to owner-occupiers at a DTI above 6, and no more than 20 percent to investors at a DTI above 7.^[9] The RBNZ's October 2025 review kept DTI settings unchanged, noting they are "calibrated to limit high-risk lending in housing upswings and periods of low interest rates, without the need for adjustment."

The BIS credit-to-GDP gap for New Zealand has moved into negative territory as of 2025, in line with Australia's -9.9 percent gap (Q4 2025).^[1] Under the Basel III countercyclical capital buffer framework, a negative credit gap means no buffer is required. This does not mean credit conditions are benign; it simply means credit growth is not currently running ahead of its long-run trend.

5. Housing Supply Dynamics and Demographic Pressures

5.1 Building Activity and Supply Elasticity

New Zealand's long-run price elasticity of housing supply, meaning how much new construction responds to a rise in prices, is estimated at 0.71 by Caldera and Johansson (2013).^[10] A value of 0.71 means a 10 percent rise in house prices leads to only a 7.1 percent increase in new building. This is well below 1.0, placing New Zealand among the less responsive (more supply-inelastic) OECD housing markets, alongside the UK and most continental European countries. Australia's elasticity is estimated at roughly 0.62 nationally.^[11]

Building consents data from Stats NZ show a clear cycle: consents peaked at roughly 50,000 a year in 2022 (the highest since the 1970s boom), fell to roughly 33,500 in early 2025, and have since recovered to roughly 39,700 (year to May 2026, up 19 percent on the year before).^[12] The mix has shifted toward medium-density townhouses, which now rival stand-alone houses in monthly consent volumes.

5.2 Population Growth and Migration

Net migration to New Zealand has dropped sharply from the record +135,500 (October 2023 year) to roughly +12,000-14,000 by late 2025.^[12] NZ citizen departures remain high at roughly 40,000 net loss per year, partly offset by non-NZ citizen arrivals. This sharp drop in net migration removes a major demand-side force that helped drive the 2020-2021 price surge.

Australia's population growth has also slowed, from a peak of 538,000 net overseas migration (June 2023) to roughly 301,000 (December 2025).^[13] However, Australia's overall population growth rate of 1.5 percent remains above New Zealand's, and capital city growth of 1.8 percent continues to put strong pressure on housing demand, especially in Sydney, Melbourne, and Brisbane.

5.3 The Auckland Unitary Plan: Evidence of Upzoning Efficacy

The academic research on the 2016 Auckland Unitary Plan provides strong causal evidence that upzoning (allowing more housing on existing land) meaningfully increases housing supply when done at scale. Greenaway-McGrevy and Phillips (2023) find that the AUP upzoning, covering roughly three-quarters of Auckland's urban land, substantially increased dwelling construction in upzoned areas.^[2] Greenaway-McGrevy (2025) estimates roughly a 24 percent increase in long-run floorspace, implying dwelling prices 15 to 27 percent lower than they would have been without the upzoning.^[14] Related rent evidence suggests Auckland rents are roughly 28 percent lower than they would have been without the AUP, relative to non-upzoned NZ cities.^[15]

The government's Going for Housing Growth programme builds on this evidence, requiring councils to plan for 30 years of housing capacity using "high" population projections. The Medium Density Residential Standards (MDRS) are being made optional for councils once they show they can meet their Housing Growth Target.^[16] The replacement of the Resource Management Act (RMA) by the Planning Bill and Natural Environment Bill, introduced in December 2025, aims to cut regulatory barriers further, though transitional rules mean the RMA will remain in force for 3.5 to 6 years after the new laws take effect.

6. Tax and Regulatory Settings

Several structural features are unique to New Zealand and have a large effect on the housing market. First, New Zealand has no comprehensive capital gains tax, making housing the most tax-favoured major asset class for domestic investors. The bright-line test, even at its tightest (10 years, 2021-2024), was a weak substitute. The return to a 2-year test from July 2024 largely removes any tax disincentive to short-term property speculation.

Second, the restoration of 100 percent interest deductibility for rental properties (from 1 April 2025) brings back the pre-2021 tax treatment. Combined with the absence of a capital gains tax, this makes leveraged property investment especially attractive compared to other asset classes. This is the mirror image of Australia's coming restriction of negative gearing and its capital gains tax discount (effective 1 July 2027). Treasury and Grattan Institute estimates suggest these changes will lower Australian house prices by 1 to 4 percent relative to where they would otherwise be.^[17]

Third, New Zealand's foreign buyer ban (in place since 2018) has removed one channel of external demand pressure, though the scale of foreign buying before the ban was fairly small (roughly 3 percent of transactions nationally, higher in central Auckland and Queenstown). The 2025 targeted exemption for investor visa holders buying properties worth NZD 5 million or above is estimated to affect fewer than 1,000 homes, or roughly 0.05 percent of the housing stock.^[18]

7. International Comparisons: Boom and Bust Episodes

7.1 Comparative Peak-to-Trough Declines

Five past episodes offer the most useful comparisons for New Zealand's current situation:

7.2 Key Distinguishing Conditions

The Jorda-Schularick-Taylor (JST) macrohistory database, covering 17 advanced economies since 1870, provides a helpful framework for understanding these episodes.^[19] Three findings from the JST research programme are directly relevant here:

First, whether a banking crisis occurs is the key dividing line. Ireland, Spain, Japan, and Sweden (1990s) all suffered systemic banking crises alongside their housing crashes. What set them apart was not the size of the housing bubble itself, but how exposed banks were to real estate lending, how strict their lending standards were, and how quickly regulators stepped in. In contrast, Sweden (2022-2024) and Canada have avoided banking crises during their recent corrections, and both have seen relatively orderly adjustments.

Second, the composition of credit matters more than aggregate credit growth. Jorda, Schularick, and Taylor (2016) demonstrate that mortgage lending, specifically, is the primary driver of financial fragility in modern advanced economies, as the share of mortgage loans on bank balance sheets doubled from approximately 30 percent in 1900 to approximately 60 percent today.^[20] Both New Zealand and Australia have banking systems dominated by mortgage lending, which makes this finding directly applicable.

Third, demographic and supply factors determine recovery trajectories. Japan's demographic decline fundamentally prevented price recovery for over three decades, despite ultra-low interest rates. Ireland's post-crisis supply constraint (building virtually ceased during the crash) and population growth enabled a strong rebound to surpass the 2007 peak. Spain's oversupply (5 million units added to a 20 million base during the boom) prolonged the correction. Canada's immigration-driven demand and supply shortage have cushioned the recent downturn.

8. Trajectory Assessment: NZ Housing Market Outlook

8.1 Scenario Framework

We assess three scenarios for the New Zealand housing market over a five-year horizon (2026-2031), drawing on the full weight of evidence assembled above.

8.2 Scenario A: Prolonged Plateau (Central Case)

The central case is that New Zealand real house prices remain broadly flat over the next five years, with nominal prices growing at approximately the rate of income growth (3-5 percent per year) and real prices stable or declining modestly. This scenario is supported by several converging forces:

Continuing real price erosion through inflation. Even with nominal prices stable, CPI inflation at 2-3 percent implies a real price decline of 10-15 percent over five years through the inflation channel alone. This mirrors the post-GFC experience in several European markets where nominal prices were flat but real prices declined steadily.

Binding DTI constraints on the upside. The DTI speed limits (20 percent of new lending at DTI above 6 for owner-occupiers) effectively cap the extent to which falling interest rates can translate into higher prices. As rates fall and households can service larger mortgages, the DTI constraint becomes the binding limit. APRA's activation of similar DTI limits in Australia (February 2026) reflects the same logic: regulators in both countries are explicitly using macroprudential tools to break the interest-rate-to-house-price transmission channel.

Supply response gaining traction. Building consents recovering to approximately 40,000 annually, combined with the structural shift toward medium-density construction, suggests that the supply side is finally responding. The academic evidence from the Auckland Unitary Plan is unambiguous: upzoning works, and the Going for Housing Growth programme extends this logic nationally.

Migration normalisation removing demand pressure. Net migration at approximately 12,000-14,000 is close to the long-run average and well below the 2023 peak. While migration could recover if economic conditions improve, the government's focus on high-skill migration rather than volume suggests a structurally lower net migration trajectory than the post-COVID surge.

8.3 Scenario B: Continued Growth

The conditions for renewed growth are not implausible. If the RBNZ cuts the OCR below the current 2.25 percent (perhaps to 1.5-2.0 percent), mortgage rates would fall toward 3.5-4.0 percent, substantially improving serviceability. If combined with a relaxation of DTI limits (which the RBNZ has the discretion to implement) and a recovery in net migration toward 30,000-40,000, the demand-side boost could be substantial. The restoration of interest deductibility and the shortened bright-line test would amplify investor demand in this scenario.

However, this scenario faces two structural headwinds. First, the RBNZ has stated that DTI settings are calibrated for upswings and would not require adjustment; relaxing them in the face of a housing recovery would be difficult to justify. Second, the RBA's experience (three 2025 cuts fully reversed by May 2026) demonstrates that inflation risks remain live, constraining how far central banks can ease.

8.4 Scenario C: Sharp Correction

The conditions for a crash are not currently present, but they can crystallise quickly. The primary trigger would be an external shock that forces the RBNZ to raise rates while household balance sheets remain stretched. The most plausible candidates are a global recession driven by trade disruption or geopolitical conflict, a sharp rise in global risk premiums, or a domestic inflation resurgence linked to fiscal expansion or commodity price movements.

The critical protection against Scenario C is the strength of bank balance sheets. The RBNZ's 2025 solvency stress test confirmed that banks can absorb loan losses under adverse economic scenarios. Household equity buffers remain substantial: the most recent RBNZ Financial Stability Report records moderate and stable mortgage arrears, and the aggregate loan-to-value ratio of the mortgage book is well below levels that would threaten bank solvency even under a severe price decline. This is fundamentally different from the pre-GFC situation in Ireland or Spain, where loan-to-value ratios were far higher and underwriting standards far weaker. For comparison, in Australia, the RBA's April 2025 Financial Stability Review reported that fewer than 1 percent of households were in negative equity and that even a 30 percent price decline would leave 9 in 10 mortgagors with positive equity.^[23]

The primary vulnerability in Scenario C is not bank solvency but the macroeconomic consequences of household deleveraging. New Zealand household consumption is sensitive to housing wealth effects, and a sharp price decline would reduce consumer confidence and spending, amplifying any recessionary forces. The RBNZ's November 2025 Financial Stability Report explicitly notes that "financial stability risks remain higher than they have been in recent years,"^[21] reflecting concerns about the uneven business recovery and global risk environment.

9. Predictive Uncertainty

Several sources of uncertainty warrant explicit acknowledgement:

Model uncertainty. The valuation metrics used here (price-to-income ratios, HP-filtered trends) embed assumptions about what constitutes "normal." If the structural decline in global real interest rates since the 1990s is permanent, the post-2000 shift in price-to-income ratios may represent a new equilibrium rather than a deviation awaiting correction. The academic literature is divided on this point: the JST research programme emphasises the historical uniqueness of the post-1980 "great mortgaging," while others argue that falling real rates and rising land scarcity in urban areas justify permanently higher price-to-income ratios. We cannot resolve this debate definitively, but the weight of cross-country evidence (including Japan's failure to mean-revert after three decades of ultra-low rates) suggests caution about relying on mean reversion.

Policy endogeneity. The scenarios assume that macroprudential policy remains restrictive during upswings and that monetary policy responds symmetrically to inflation and financial stability concerns. In practice, political pressure to relax LVR and DTI restrictions during a housing downturn is substantial, and the RBNZ's new Financial Policy Committee (operational from 2026) has not yet established its institutional credibility. The Australian experience, where APRA activated DTI limits only after investor lending surged 18 percent in a single quarter, illustrates the tendency for macroprudential policy to be reactive rather than pre-emptive.

Migration uncertainty. Net migration is the single most difficult variable to forecast, and it has historically been the dominant driver of short-run housing demand in New Zealand. The range of plausible outcomes is wide: from sustained out-migration of NZ citizens (continuing the trend of 40,000 annual net departures) to a resurgence of inward migration if economic conditions improve relative to Australia and other destinations.

Climate and insurance risk. An emerging factor not captured in historical data is the impact of climate-related risks on property values and insurability. New Zealand's exposure to coastal flooding and erosion, particularly in Auckland, Wellington, and Christchurch, could materially affect property values in affected areas over the medium term. This is a source of downside risk that is concentrated geographically but potentially significant in aggregate.

10. Data, Sources, and Methodology

The data behind each figure is available for download below. This is an analytical brief rather than a raw data release, so the datasets combine three kinds of series, which should not be confused: indicative annual points read from official long-run series (used to show decade-scale trends and turning points), author computations (the deviation-from-trend and z-score constructs), and illustrative comparisons (the policy-milestone and international-episode charts). For authoritative, full-resolution data, use the official source links in each row of the sources file.

10.1 How the series are constructed

Real house prices (Figures 1, 2, 8). Real house prices are nominal prices deflated by the consumer price index on the BIS harmonised basis (2010=100). Figure 1 plots indicative annual points for New Zealand and Australia. Figure 2 compares the approximate real price change around discrete policy events; the events are not evenly spaced, so it is a comparison rather than a continuous series. Figure 8 reports peak-to-trough real declines by episode, with New Zealand and Australia asterisked because their corrections were ongoing at the time of writing.

Price-to-income (Figure 3). The OECD analytical price-to-income indicator (2015=100); the dashed line is the median of the plotted series (88.9) as a long-run reference.

Deviation from trend (Figure 4). Computed by the author with a one-sided Hodrick-Prescott filter (lambda = 400,000, consistent with the BIS convention for long series) on the real house price series, expressed as the percentage deviation of the series from the filtered trend. Because the filter adapts to structural shifts, it should be read as indicative of cyclical position rather than a precise valuation gap.

Valuation z-scores (Figure 5). For each metric (price-to-income, price-to-rent, real HPI deviation from trend, and household debt-to-income), the z-score is the value minus the series mean divided by the series standard deviation, over the available sample. Two snapshots are shown: the Q4 2021 peak and Q1 2025.

Credit and supply (Figures 6 and 7). Indicative annual points reproducing the profile of RBNZ series (1-year fixed carded mortgage rate; household debt as a percentage of disposable income) and Stats NZ series (dwelling consents issued; net permanent and long-term migration).

10.2 Downloadable data

All source URLs were verified as live on 9 July 2026. The sources file records the source, URL, and a construction note for every individual series.

The primary sources are the Bank for International Settlements (real house price long series), the OECD (price-to-income indicators), the Reserve Bank of New Zealand (mortgage rates and household debt), and Stats NZ (building consents and migration), all listed with their URLs in the reference section below and in the sources file.