Methodology
How the verdict is calculated.
Eleven steps, all public, all source-cited. We test every NZ property at 100% finance, interest-only, against MBIE bond medians and carded rates. Plain maths, no black box.
The test
100% finance, interest-only, every property.
Most cashflow calculators start by asking your deposit size or LVR. We do not. Borrow less and almost any property looks cashflow positive — partial-LVR scenarios flatter the verdict and nothing stays comparable across addresses. Testing at 100% finance, interest-only, asks the harder, comparable question: does the rent cover the building on its own?
The rate we use is a single representative carded one-year fixed rate, refreshed daily from interest.co.nz. Users on Starter can override the rate, and adjust LVR with a slider for partial-finance scenarios. The default — and what every public verdict is run at — is the 100% test.
The pipeline
Eleven steps, named and sourced.
Each step takes a single named input and produces a single named output. The whole calculation is auditable end to end.
- 01
Weekly rent estimate
Input
Address
Source
MBIE Tenancy Services rental bond data
Formula
median(MBIE bonds where TLA = address.tla and bedrooms = n and type = t)
The bond medians are the only public, regulator-collected source for actual NZ rents. We filter by TLA, bedroom count and property type. If the suburb has fewer than ten bonds in the quarter, we fall back to the regional median and disclose the vintage on the result.
- 02
Annual gross rent
Input
Step 01 × 52
Source
Derived
Formula
weeklyRent × 52
No banks involved yet. Just the gross rental income the property would generate over a calendar year if fully tenanted at the median rent.
- 03
Vacancy allowance
Input
Region
Source
MBIE vacancy data by region
Formula
annualRent × vacancyRate(region)
We use the regional figure rather than a flat national assumption. Vacancy moves the cashflow more than most investors expect, and one number for the whole country flatters or punishes the wrong places. The current value for the property's region is shown on the result page with its asOf date.
- 04
Effective rent
Input
Step 02 − Step 03
Source
Derived
Formula
annualRent − vacancy
The income line we actually use for the rest of the calculation.
- 05
Mortgage interest (100% finance, interest-only)
Input
Purchase price × rate
Source
RBNZ + interest.co.nz carded average
Formula
price × (rate / 100)
We test every property at 100% finance, interest-only, by default. Borrow less and almost any property looks cashflow positive — the test is whether the rent can cover the building on its own. The rate input defaults to a single representative carded one-year fixed rate, refreshed daily. An LVR slider is on Starter for partial-finance scenarios.
- 06
Council rates
Input
Address
Source
Each council's published rate-in-the-dollar (Auckland Council fully covered; other major TLAs covered via published schedules)
Formula
rateableValue × rateInTheDollar(address.tla) + fixed charges
Each council publishes a rate-in-the-dollar plus per-property fixed charges (UAGC, water-quality, refuse, etc.). The figure varies by council, so a flat national assumption would mis-price low-rate and high-rate councils by thousands a year. We keep a TLA-keyed lookup updated from each council's public schedule and disclose the source + asOf date on the result.
- 07
Insurance
Input
Region + replacement cost + property type + hazard layers
Source
Indicative residential premiums published by NZ insurers + EQC / FENZ / NHI levies
Formula
multi-factor model: rebuild rate × base premium × stack + GST + FENZ + NHI
EQ-zone pricing dominates regional insurance variance in NZ — figures are not interchangeable across regions. We use a multi-factor model that takes the property's region, replacement cost band, type (house / apartment / townhouse), and any hazard-layer flags (flood / coastal / slope) and produces an indicative annual premium with the levy stack applied. Output is indicative, not a quote. The result page lists every assumption the model applied (suburb-median floor area, baseline year built, etc.) so the user can override on Investor.
- 08
Property management
Input
Step 04 × 0.08
Source
NZ industry standard
Formula
effectiveRent × 0.08
Eight per cent of effective rent is the standard NZ property-manager fee. The user can toggle this off for a self-managed test on Starter.
- 09
Maintenance reserve
Input
Price × maintenanceRate (default 0.01)
Source
NZ investor practice (range 0.5%–1% of price)
Formula
price × maintenanceRate
NZ investor practice runs a band of 0.5% to 1% of purchase price per year. We default to 1% as the conservative end — a thumb on the scale toward "you can afford this if maintenance comes in heavy". The result page shows both a 0.5% and a 1% verdict so the band is visible, and the dashboard exposes a slider so investors can dial the assumption to the property in front of them. We default conservative and surface the lever, rather than pick a single number that flatters every result.
- 10
Net annual cashflow
Input
Step 04 − (05 + 06 + 07 + 08 + 09)
Source
Derived
Formula
effectiveRent − totalExpenses
The verdict line. Positive means the property pays for itself before tax at the chosen rate. Negative means a weekly shortfall before tax.
- 11
Yields
Input
Step 02 / price · Step 04 / price
Source
Derived
Formula
gross = annualRent / price · net = effectiveRent / price
Gross yield is the often-quoted figure. Net yield is the same number after vacancy. Neither yield includes expenses; both are surface metrics shown alongside the cashflow verdict.
Worked example
12 Example St, Ponsonby · $850,000.
A three-bedroom standalone house in Ponsonby, asking $850,000. Auckland TLA. Inputs are pulled from the live dataset set: MBIE bond medians for rent, the RBNZ B6 weighted-average residential mortgage rate for finance, the council schedule for rates, and the insurance model for the premium. Refresh dates are visible on the result page.
| Line | Annual |
|---|---|
| Annual rent (720 × 52) | +$37,440 |
| Vacancy (4% Auckland) | −$1,498 |
| Effective rent | +$35,942 |
| Mortgage interest (purchase price × RBNZ B6) | −$55,675 |
| Council rates (Auckland Council schedule) | −$2,635 |
| Insurance (multi-factor model — see step 07) | −$1,880 |
| Property management (8% of effective) | −$2,875 |
| Maintenance (default 1%) | −$8,500 |
| Net annual cashflow | −$35,113 |
Maintenance band
At 0.5% maintenance the verdict is −$594 / wk. At 1% it is −$675 / wk. We show the lower end to make the floor visible; the headline number uses the conservative 1% default per Step 09.
The verdict at the 1% maintenance default is cashflow negative before tax. Each line above traces back to one of the published datasets in the source list below — MBIE for the rent, RBNZ B6 for the rate, the council for the rates, and so on. Real result pages show the dataset name and refresh date next to every figure, plus “What would make this positive?” levers (price drop, rent lift, rate drop, deposit) so the path back to break-even is on the page rather than left as an exercise.
Source list
Where every input comes from.
Every figure on a result page links back to one of these. Vintages are surfaced on the result; this is the master list.
MBIE Tenancy Services — rental bond data
Quarterly CSV download
Weekly rent medians by TLA, bedroom count, property type. Vacancy by region.
RBNZ — interest rate series
Daily
Official Cash Rate, wholesale yield curve.
interest.co.nz — carded mortgage rate aggregator
Daily
Single representative carded one-year fixed rate, refreshed daily.
Auckland Council, Wellington City, Christchurch City + top 20 TLAs
Annual (rates set yearly)
Rate-in-the-dollar by TLA. Auckland-specific per-address lookup where available.
Tower / IAG / regional EQ-zone bands
Quarterly review
Insurance base premium, capping, and regional multiplier.
Stats NZ
Weekly
Population, consents, migration data. Used by the dashboard, not the verdict.
What we don't model
The verdict is a screening tool, not the whole story.
Here is what the cashflow verdict deliberately leaves out, and why.
Tax
The verdict is before tax. We do not model individual investor tax positions, deductibility limits, or interest-deductibility phase-ins. A negative verdict before tax is almost always negative after tax.
Depreciation
NZ residential property has had limited depreciation deductibility since 2011. We do not currently fold depreciation back into the verdict, partly because the rules have moved and partly because they vary by chattels split.
Capital appreciation
The verdict is a cashflow test, not a return-on-investment test. Capital appreciation may make a cashflow-negative property a good investment over a long hold. That decision sits outside the verdict.
Property-specific quirks
Body corporate fees, leaky-building remediation, weather-tightness reserve, ground rents, cross-lease title costs, and HRV-required ventilation upgrades are not modelled. These can move the numbers by thousands a year on an individual property and are why the verdict is a screening tool, not a substitute for due diligence.
Bank-specific premiums
A non-bank lender, a low-deposit premium, or a specific bank's pricing for an investment-classified loan is not modelled. We use a representative carded one-year fixed rate as the default; users can override with an LVR slider on Starter.
Reminder
The verdict is an estimate. Property finance has no guarantees. Use the output as a screening tool, not a substitute for professional advice.