First Home Loan: lifting the income cap to $300,000
The deposit problem is real. The question is whether the new settings go too far.
About the Housing CompassThe Housing Compass is a way of testing whether a housing policy will actually do what it says it will. It asks three questions, in order:
Most housing debate stops at the first question. This assessment works through the first two. The third can only be answered once the policy has been in the market long enough to read, and we will come back to it. |
THE INTENTION
The aim is simple: help more first home buyers into a home by fixing the deposit problem, rather than pretending the main problem is whether they can afford the repayments.
At the moment you can only use the scheme if you earn under $95,000 on your own with no children, or under $150,000 as a couple, or as a single buyer with children.
The new policy replaces both of those limits with a single limit of $300,000. Everything else stays the same.
Kāinga Ora guarantees the loan, which means it covers the lender’s loss if the borrower cannot pay. The banks taking part still decide who gets the loan. The buyer still needs a 5% deposit and still pays an insurance fee of 1.2%, which can be added to the loan rather than paid up front.
WHAT THE POLICY GETS RIGHT
This part I agree with. For many first home buyers at the cheaper end of the market, meeting the repayments is not the hardest part. Saving the deposit is.
The policy makes the same distinction: it lowers the deposit hurdle, but it does not soften the test a lender uses to work out whether a borrower can afford the loan.
The buyer numbers show the problem is real. The number of first home buyers taking out loans in the year to July rose from 28,063 in 2024 to 33,572 in 2026. Cotality puts first home buyers at a record 28.3% of all purchases in the second quarter of 2026, against a long run average of about 22%.
So this is not simply a bigger slice of a shrinking market. The number of buyers is rising as well.
It also isn't an argument against the First Home Loan. It is a test of whether lifting the income cap to $300,000, while leaving every other setting untouched, is set at the right level.
THE FIVE CONSEQUENCES
Each of these issues comes from a setting the announcement leaves unchanged. None of them requires a new tax or scrapping the policy, and several could be fixed without passing a new law.
Consequence | Mitigation |
1. The support can follow buyers up the price ladder. | Bring back regional price caps, set against the cheapest quarter of homes in each area. |
2. The insurance fee may no longer match the new risk. | Recalculate the 1.2% fee for the wider group of borrowers and publish the result. |
3. The scheme reaches incomes it was never designed around. | Set the limit using what first home buyers actually earn, and phase it out gradually. |
4. The exemption could weaken the lending controls used when the market heats up. | Add a brake that responds to the market cycle, using tools that already exist. |
5. It adds buyers, but it does not add houses. | Offer a lower fee or a higher income limit for new builds. |
The support can follow buyers up the price ladder
House price caps were removed from the scheme on 1 June 2022, so there is now no limit on what a buyer can pay for the property. Kāinga Ora’s own reporting links the jump in approvals from July 2022 to the removal of those caps. In other words, the price cap had been holding numbers down.
Fair enough.
But put no limit on the house price alongside a $300,000 income limit and a simple question follows: how far up the market can support backed by the taxpayer now reach?
The scheme is no longer automatically limited to the cheaper homes it was designed around.
Australia has effectively run this experiment for us.
In the six months after its scheme was widened, homes priced below its price caps rose 6.7% in value, while more expensive homes rose 3.6%. In Sydney, homes below the cap rose 4.1% while those above it fell 1.1%.
Mitigation:
Bring back regional price caps, set against the cheapest quarter of homes in each area.
Better still, base them on a multi year average or on local incomes, so the cap does not simply rise every time the market does. With a genuine ceiling on entry level prices, the $300,000 income limit matters much less: a buyer on a high income buying a cheaper home is not the same risk.
The Australian evidence needs to be read carefully. A price cap can push extra demand into the price band that qualifies. That is why this fix needs to sit alongside the supply fix in point five.
Is the 1.2% insurance fee still priced for the new risk?
The 1.2% fee is the price the scheme’s insurance maths has produced. Budget 2025 moved the borrower’s share from 0.5% to the full amount. The scheme also came through the 2021 to 2023 fall in house prices with few claims, which is useful evidence, but it is evidence about the borrowers the scheme used to have.
Lift the income limit, leave house prices uncapped, and the scheme may end up covering a different mix of borrowers holding bigger loans. The obvious question is whether the old price is still the right one.
On its current settings, Kāinga Ora already expects the amount it may have to pay out under its Mortgage Insurance Scheme to rise 45% over two years, from $71 million at 30 June 2025 to $103 million in 2026/27.
The announced cost of the expansion, $4 million to $6 million, covers running the scheme. That is not the same thing as how much the Crown could be liable for in future. If a borrower adds the insurance fee to the loan, they start out owning only about 3.8% of the home. House prices do not have to fall far before the government’s guarantee starts to matter.
Mitigation:
Work out the price again for the wider group of borrowers, and publish the result.
Test it against the incomes the scheme will now cover, the bigger loans that can follow from having no price cap, and a lasting fall in house prices in dollar terms.
Few claims in the past shows the old settings worked for the old set of loans. It does not prove the same fee is right for a very different one.
The scheme now reaches incomes it was never designed around
The First Home Loan was built for households that can afford the repayments but cannot save the deposit. That is also the reason it is exempt from the debt to income rules, which cap how much a household can borrow against what it earns.
Those rules were never meant to stop Kāinga Ora lending to lower income households.
A household earning $300,000 sits in the top 10% of household incomes. That does not make those buyers undeserving, but it does mean the scheme now reaches well beyond the group its own exemptions were written for.
The next question is obvious: where did $300,000 come from?
National points to an average first home buyer income of about $146,000 and assumes half of buyers earn more than that. An average cannot tell you that. You need the midpoint, or the full spread of what buyers earn. Even taking the $146,000 figure at face value, a limit of around $180,000 to $200,000, adjusted each year for inflation, is easier to explain than $300,000.
There is also no phase out.
A buyer either qualifies or does not, based on income over the previous 12 months, so someone just under the limit gets help and someone just over it gets none. That is a blunt way to direct support to the people who need it.
Mitigation:
Set the limit using what all first home buyers actually earn, not what current users of the scheme earn, since their incomes are already held down by the present cap. Then phase the help out gradually, rather than cutting it off at a single line.
The exemption could weaken the lending controls used when the market heats up
First Home Loans sit outside two of the Reserve Bank’s limits on bank lending: the limit on loans that are large compared with the value of the house, and the limit on loans that are large compared with the borrower’s income.
There is also no cap on how many First Home Loans can be written.
Kāinga Ora’s 2026/27 Statement of Performance Expectations describes the scheme as demand driven, meaning it writes as many loans as there are buyers who qualify.
In practice, that leaves two things holding the scheme in check: the income cap, and the lender’s own test of whether a borrower can afford the repayments.
That does not mean New Zealand is in a lending boom.
Adjusted for inflation and measured in July 2026 dollars, total new mortgage lending is still about 18.5% below the 2021 peak, but low deposit lending tells a different story.
Lending to people borrowing more than 80% of the value of the home, known as high LVR lending, is about 20% above that peak and is the highest it has been since these figures began. In July 2026, 16.1% of all new lending was above that 80% mark, the highest monthly share since the series began in 2014. The seven highest months on record are the first seven months of 2026. And the trend is spreading beyond first home buyers.
I'm also not suggesting this is a problem for the financial system today.
The Reserve Bank loosened its low deposit lending rules in December 2025, left them unchanged in August 2026, and judged the risks in housing to be contained.
My concern is how the scheme behaves through the next part of the property cycle. The Reserve Bank can tighten the normal limits if the market eventually heats up, but it does not set the rules for the First Home Loan.
Make the exempt channel bigger now and one of the brakes available later works less well, at exactly the point it might be needed.
Mitigation:
Build a brake into the scheme now, using tools that already exist.
The insurance fee could rise and fall with the market. A limit on numbers could be brought back and tied to a published measure, such as the share of lending over the past 12 months that went above the 80% mark, or the larger First Home Loans could count against the Reserve Bank’s normal limits.
The point is not to predict that the market will overheat. It is to avoid having to redesign the scheme in the middle of the next upswing.
It adds buyers, but it does not add houses
This is the simplest issue. Nothing in the policy favours new builds. A guaranteed loan used to buy an existing house helps one buyer compete for the homes already there, but the country still has the same number of houses and one more supported bidder.
If we put the same support behind a new build, it can help create both a homeowner and an extra house.
Australia’s banking regulator has made that distinction plainly. When its debt to income limit came into force on 1 February 2026, loans to buy or build new homes were left out of it, so the rule would not discourage building.
Australian brokers have also reported first home buyers moving toward house and land packages, as existing homes priced below the caps became harder to find. So in practice a price cap already does some of this steering.
Mitigation:
Tilt the support toward new builds.
That could mean a lower insurance fee, a higher income limit, or both. Existing homes do not have to be excluded, but the settings should recognise the difference between funding another bid for the houses already here and helping bring new houses into existence.
THE READING
The policy starts by identifying the right problem. The deposit barrier is real, the First Home Loan is a sensible tool for dealing with it, and the evidence shows first home buyers are using it.
Where I disagree is with the assumption that the income cap can jump to $300,000 while every other setting stays exactly where it was: no limit on house prices, no limit on numbers, the exemptions from the Reserve Bank’s lending rules left intact, and nothing published on what first home buyers actually earn, on whether the insurance is still priced correctly, or on what the wider scheme means for financial stability.
I would not reverse the policy. I would set it properly before the market tests it for us.
Comments