NZ housing flat patch nears modern duration record as OCR rises

New Zealand is about 36 months into a post-bust housing grind. BNZ chief economist Mike Jones says the last longer flat spell ran from 1997 through mid-2001—about 45 months. If national values stay flat into autumn 2027, as many bank and Reserve Bank paths imply, the current episode would set the modern duration record.

Cotality data cited in market reporting put August 2026 national dwelling values down 0.4 percent month-on-month. The national median sat at $797,944, about 1 percent lower than a year earlier. Most of the heavy correction landed in 2022 and 2023. Since then prices have drifted, not crashed.

The Reserve Bank raised the official cash rate 25 basis points to 2.75 percent on 2 September 2026. That was the second consecutive hike after July’s move to 2.50 percent. The September Monetary Policy Statement projects nominal house prices to decrease slightly in 2026. Real prices ease near term before rising modestly from mid-2027 with incomes. Rising floating and short-fixed mortgage rates collide with that path. So does election-related tax uncertainty ahead of 7 November 2026.

National medians still mask two markets. Auckland and Wellington sit deep below peak. Canterbury, Otago and Southland have reclaimed or exceeded prior nominal highs. Policy keyed only to a single index risks missing that split.

Why prices stay flat

The drivers stack rather than compete. Mortgage serviceability is tighter after the OCR lift. Major banks passed the full 25 basis points onto floating products. Carded floating rates now cluster near 6.25–6.40 percent. RBNZ B20 averages at end-July already showed floating at 6.32 percent, one-year fixed at 5.34 percent and five-year fixed at 6.13 percent. Fixed rates had drifted higher through mid-2026 on hike expectations.

Labour-market caution is the second brake. Stats NZ’s June 2026 Household Labour Force Survey put seasonally adjusted unemployment at 5.6 percent. That was up 0.2 percentage points on the March quarter and 0.4 points on a year earlier. About 171,000 people were unemployed. Underutilisation sat at 13.8 percent. Employment was 2.905 million. Valocity linked the soft August value print to rate pressure and that labour backdrop.

Cotality chief property economist Kelvin Davidson has repeatedly argued that sellers are not forced. Soft labour conditions have not produced widespread job losses. Vendors can stay patient and hold firm on price. Elevated listings give buyers pricing power anyway. Cotality’s July sales tally was 6,935 deals, down 6.4 percent year-on-year—the seventh consecutive monthly fall. Rolling 12-month sales eased to 89,385.

Property sales volumes have inched lower so far this year, although they're still at a relatively normal level. But the stock of listings remains elevated and this is giving buyers the balance of power when it comes to pricing.

Davidson made that assessment alongside Cotality’s July Home Value Index release, when the national median was $804,303.

Supply has also turned. Stats NZ recorded 40,581 new dwellings consented in the year ended June 2026, up 19 percent. The year to July reached 40,908, up 21 percent. Multi-unit consents outnumbered stand-alones. Auckland consented 17,097 homes, up 20 percent. Canterbury consented 8,647, up 33 percent. Economic indicators spokesperson Michelle Feyen said it was the first time annual new-home consents had exceeded 40,000 since 2023. HUD’s June-quarter Housing Market Update put active listings near 34,100—more than double boom-era levels. A larger stock caps upside even if demand returns.

Election risk adds a demand freeze for some investors. Labour proposes a 28 percent capital gains tax on residential investment and commercial property. The family home is exempt. Gains would apply only after 1 July 2027. Revenue is ring-fenced to health. National and ACT rule out new property taxes. ANZ’s August Property Focus flagged rising rates, election uncertainty and possible housing taxes as investor headwinds.

Macroprudential settings are not loosening the gate. The Financial Policy Committee left LVR speed limits unchanged on 14 August 2026. Owner-occupier lending above 80 percent LVR stays capped at 25 percent of new flows. Investor lending above 70 percent LVR stays at 10 percent. Debt-to-income speed limits also held. Assistant Governor Angus McGregor said housing risks were contained, with prices broadly flat and lending growth modest.

Where the trade-offs bite

Inflation control is the first trade-off. June-quarter CPI hit 4.1 percent on Middle East conflict fuel prices. Ex-fuel inflation was 2.9 percent. Most core measures sat inside the 1–3 percent band. The Committee expects inflation back inside the band by mid-2027 and near 2 percent later that year. Hiking the OCR to clear the fuel spike and anchor core raises mortgage costs. That extends price stasis and weak residential investment in Auckland and Wellington—the channels the September MPS already flags.

The RBNZ media release put the household channel in plain terms. Weak income growth, job insecurity and flat house prices continue to weigh on household spending and residential investment, particularly in Auckland and Wellington. Resilient export demand supports regional New Zealand at the same time. Two-speed recovery is the operating description.

Affordability restoration collides with household wealth. Cotality’s Q2 2026 Housing Affordability Report put the national value-to-income ratio at 6.7, matching its long-term average and down from a 9.8 peak in late 2021. Time to save a 20 percent deposit eased to 8.9 years against a long-term average of 9.0 and a peak of 13.1. Auckland’s ratio fell to 7.2, below its long-term 7.5. Wellington sat at 5.5, below 6.2. First-home buyers took a record 29.0 percent share of July purchases on Cotality figures. Those gains for entrants sit beside real drawdowns. BNZ’s June Eco-Pulse put national prices about 15 percent below the 2021 peak nominally and about 28 percent below in real terms. BIS Q1 2026 real residential prices showed New Zealand down about 4 percent year-on-year among sharper advanced-economy falls.

Credit quality shows the scar without a systemic forced-sale wave. Cotality’s Pain and Gain report for Q2 2026 found 13.1 percent of residential resales made at a loss—the highest share since 2012. Auckland’s loss share was 20.9 percent. Wellington region was 18.4 percent. Apartments were near 45 percent. The median loss was about $60,000 against a median gain near $280,000. Loss-making sellers held about 4.3 years versus 10.4 years for gainers—pointing to the 2021–22 peak cohort. Cotality has also estimated about 3,600 first-home-buyer households in negative equity on a 20 percent deposit assumption, from roughly 29,000 who bought in the peak window. Patient vendors and mid-5s unemployment still limit forced sales. Banks watch the book anyway.

Supply normalisation versus construction lag is another tension. Consents above 40,000 undercut pure shortage rhetoric. Builds lag. Funding costs rise with the OCR. The RBNZ notes strong consents have not fully translated into construction nationwide. Builders and multi-unit developers carry pipeline risk if sale prices lag costs. Local-government development contributions and rating bases in high-consent cities face the same squeeze.

A capital gains tax would trade revenue aims against transaction risk. Labour’s design taxes only post-1 July 2027 gains on investment and commercial property. If Labour leads after 7 November, H1 2027 can produce valuation-date dynamics: deferred purchases and selective listing behaviour. Volume risk exceeds crash risk while stock is already elevated. National-led continuity removes the overhang. It does not erase the rate bite. Ring-fencing CGT revenue to new health spending still leaves the Crown with wider fiscal choices on whether new taxes fund new programmes or repair the balance sheet first.

LVR and DTI stability keeps higher-risk lending contained. It also keeps deposit and income gates tight. Affordability gains from lower prices are partly offset by credit constraints and higher servicing rates. That is the policy tension in plain form: price relief without full access relief.

Second-order effects households and firms should watch

Prolonged real-price softness restrains durables and retail demand in Auckland through wealth and confidence channels. The RBNZ says household saving rates have risen and consumption remains weak. Flat house prices weigh especially in Auckland and Wellington.

Refinancers rolling off cheaper fixed terms onto floating near 6.25–6.40 percent face payment shock. Collateral values are not rising to cushion them. Serviceability tests bite harder on new borrowing capacity. ASB housing confidence for the three months to July showed a net 57 percent expecting higher interest rates in a year—the highest such reading since April 2023.

The negative-equity first-home cohort is a credit-quality watch item, not yet a systemic forced-sale driver. Loss-making resale shares may peak then ease as hold periods lengthen and the 2021–22 cohort either holds or exits through 2027–29. Bank provisioning and mortgage-book stress tests should track that path without importing sharper Canadian-style crash assumptions.

Construction faces a consent-to-build lag into rising funding costs. Fletcher Building, regional multi-unit developers and iwi or community housing providers sit in that pipeline. Soft clearances in Auckland and Wellington stress margins even where consents look healthy. REINZ July figures showed a national median sale price of $760,000, down 0.7 percent year-on-year, with sales of 6,090 down 10 percent and inventory of 33,252 up 9.3 percent. Days to sell stretched to 50. realestate.co.nz put August average asking prices at $849,362, 3.2 percent below August 2023, with 32,908 homes for sale—45 percent more than three years earlier.

Net migration is supportive but modest. Stats NZ put the provisional year to June 2026 net gain near 17,600. Population at 30 June 2026 was about 5.36 million, up 0.7 percent. That is far below 2022–23 migration peaks. Demand heat stays limited even if rates eventually peak.

Regional labour mobility toward stronger South Island markets is a quieter channel. RBNZ conference slides for the September MPS show house prices increasing in the South Island and decreasing in the North Island on REINZ-based measures. Export-exposed regions draw support from commodity prices and trading-partner demand. Auckland’s multi-unit pipeline and Christchurch’s relative balance shape how long the bifurcation lasts.

If Labour forms government, listed property trusts and builders can see election-night volatility around the July 2027 CGT start. If National leads, deferred investor demand may unlock gradually—still against higher mortgage rates. Either way, rate levels—not tax alone—set the medium-term price ceiling while the consent pipeline is full.

How this cycle compares with past flat spells

The duration benchmark is 1997 through mid-2001. Jones cites that stretch at about 45 months. The current sideways phase since the autumn-2023 trough is about 36 months. Forecasters’ flat-to-soft 2026 paths and only modest 2027 lifts imply the record is within reach if labour markets stay soft.

Post-GFC reclaim was slower on the clock but different in shape. BNZ notes it took about 63 months after the GFC to reclaim the previous peak. This cycle is about 55 months in and still roughly 15 percent shy nominally—and far more in real terms. Auckland and Wellington real drawdowns near 35 to 40 percent from peak, on BNZ figures, exceed a typical mild correction. Valocity’s August national index sat only about $11,000 above the March 2023 post-Covid trough, moving inside a narrow band of roughly $35,000 since early 2023.

Davidson’s Covid counterfactual reframes the bust as mean reversion. Strip the spike and the bust and current levels sit about 15 percent above pre-Covid prices on his telling. The boom was policy-distorted. The correction restored a more normal level rather than opening a permanent undershoot. That reading supports a sideways grind over a fresh crash.

Peer economies differ in timing and depth. BIS Q1 2026 real global house prices fell 1.2 percent year-on-year. New Zealand’s real decline near 4 percent sat with Canada and China among sharper moves. Canada’s peak-to-trough path was steeper and faster. Australia’s Cotality August national index fell 0.9 percent—the fifth monthly decline—leaving the median 3.6 percent below the March 2026 peak, with Sydney down 1.4 percent on the month. Australia’s downshift is later and from a hotter 2024–25 run-up. New Zealand’s profile is long duration, medium depth, and unusually split by island. Risk models should not import Australian or Canadian peaks naively.

The counter-read: bottom in for entrants

The strongest opposing case runs like this. Affordability metrics have normalised. Value-to-income is back at the long-term average. Deposit-saving time is near average. First-home buyers hold a record purchase share. Pain is concentrated among 2021–22 buyers and apartment stock. South Island markets already make new highs. The RBNZ Survey of Expectations one-year house-price inflation jumped to 1.47 percent in August from 0.33 percent—more optimistic than many bank paths. On that view, the national median is a lagging composite. Entry-level stock has found a floor.

Elements of that case are true. Affordability has improved on Cotality’s Q2 measures. First-home share is high. Regional reclaim in Canterbury, Otago and Southland is real. Invercargill has been among the stronger annual value gainers on Cotality commentary. ASB household confidence still shows a net share saying it is a good time to buy even as price-growth expectations collapsed to a net 9 percent from a summer peak of 30 percent.

The thesis against an early national rebound rests on three facts. First, the RBNZ and major banks still pencil flat-to-slightly-down national prices through 2026 and only low single-digit nominal lifts in 2027—Westpac near +0.2 percent then +2 percent; BNZ flat then about +3 percent; ANZ little movement over the next year. Second, Davidson’s labour threshold for consistent growth sits well into 2027. Unemployment at 5.6 percent is not that threshold. Third, OCR track averages in the September MPS still rise toward about 2.81 percent by December 2026 and about 3.15 percent by end-2027. Floating pass-through is already complete on the latest hike. Borrowing capacity shrinks as stimulus is withdrawn. Supply above 40,000 consents caps the upside even if sentiment turns. Election CGT risk freezes some investor demand into H1 2027 under a Labour-led outcome.

A buyers’ market for entrants can coexist with multi-year national flatness. That is the resolution, not a contradiction. QV’s three months to end-July 2026 still showed a national average home value of $898,799, down 1.5 percent over three months and 1.2 percent year-on-year, with Auckland and Wellington weaker and Christchurch slightly up. Patchwork is the structure, not a footnote.

Open questions into 2027

Will flatness persist into autumn 2027 and beat the 45-month record? Bank and RBNZ base cases make that the central risk, not a tail.

Where does the OCR peak? BNZ’s mid-2026 vintage was more hawkish near 4 percent. ANZ and ASB clustered closer to the low threes. Westpac’s August overview had 3.00 percent end-2026 and 4.00 percent end-2027. Oil prices and election arithmetic still swing the path. Floating clusters by mid-2027 will set serviceability for the next cohort of buyers and refinancers.

Does unemployment roll over fast enough? Westpac projects a drift toward 4.9 percent by end-2027. Davidson’s consistent growth call needs that labour firming. Soft job security remains the brake he names most often.

What is the exact MPS house-price path in numerical cells? The published prose says nominal prices decrease slightly in 2026 and real prices rise modestly from mid-2027 with incomes. Market commentary sometimes translates the near-term drift as roughly another 1 percent. That is interpretation, not a labelled table cell.

How large is any H1 2027 CGT volume effect if Labour leads? Design taxes only post-valuation-day gains. Behavioural freezes can still cut turnover without forcing a price crash while listings are already high. Fiscal design still matters: new taxes earmarked for new spending do not automatically strengthen the Crown balance sheet.

Full city-level August Cotality medians and methodology remain thinner on the public hub than June and July packs. Attribution to Cotality via secondary market reporting should stay explicit until the primary PDF is public.

Watch the next Cotality and REINZ prints through spring 2026 for whether national drift stays inside a narrow band. Watch Stats NZ’s September-quarter labour release for any turn in the 5.6 percent unemployment rate. Watch the RBNZ’s next Monetary Policy Statement for revisions to the OCR track and the house-price prose path. Watch election-night arithmetic on 7 November 2026 for whether the 1 July 2027 CGT start date becomes live policy or stays a counterfactual. For households, the practical near-term signal is mortgage reprice risk on floating and short fixed terms—not a sudden national crash. For builders and councils, it is consent conversion under higher funding costs. For banks, it is loss-making resale shares and the small negative-equity first-home cohort under patient-vendor conditions. The policy implication is plain. Gradual stimulus withdrawal to clear fuel-led inflation extends a housing sideways grind that national medians understate and regional data reveal. If labour markets do not firm until well into 2027, New Zealand is more likely to claim the modern flat-patch duration record than to stage an early V-shaped price rebound.