Data-Driven Property Investing: How to Research an Investment Property

Data-driven property investing: how to research an investment property in 2026

Data-driven property investing means choosing where and what to buy from a short list of numbers you can check yourself: vacancy rate, days on market, gross and net rental yield, price-to-income, population and jobs growth, the supply pipeline, and funded infrastructure. Almost all of it is published at no cost by the ABS, state government agencies and a few research houses, and the same numbers end up deciding how much a lender will let you borrow against the property.

This guide shows you how to analyse property market data: what each number tells you, where to get it, how it can mislead you, and how it flows into your loan. It finishes with a worked comparison of two hypothetical suburbs in dollars. We don’t pick suburbs for clients, and neither does this article. What we do, every week, is watch investors’ research meet a lender’s assessment, and the two don’t always agree.

One thing changed in 2026 that every investor now has to put in the spreadsheet: whether the property is a new build. More on that in the supply section.

The eight numbers worth checking

Capital growth is not in the table as its own row, but check it: pull at least 10 years of median sale prices from the Valuer General sales files, with houses and units split.

None of these works alone. A suburb can post a low vacancy rate because nothing new has been built for years, and that can change within 18 months if a large apartment pipeline is approved next door.

For the big picture above suburb level, two more sources are worth a monthly look: Cotality’s public home value index releases, which track value movements by capital city and region, and the Reserve Bank’s Chart Pack, which covers interest rates, housing credit and lending, and is updated eight times a year.

Rental demand: vacancy rate, days on market and median rent

Vacancy rate is the share of rental stock sitting empty and advertised. SQM Research publishes vacancy rates by postcode and by region, including Parramatta, Western Sydney and South Western Sydney. Its method counts only listings that have been advertised for three weeks or more, so it picks up stock that is struggling to let rather than homes that leased within a week or two. Read it against two things: the same postcode’s own history, and the wider region on the same chart. A postcode sitting well under its region and falling is telling you something. A single low month is not.

Days on market is how long a listing takes to sell. The major portals publish it on their suburb profiles; realestate.com.au’s, for example, shows it next to the median price, 12-month growth, rental yield and the number of homes sold. Always look at the number sold. A days-on-market figure built from a handful of sales in a year says very little.

Median rent is where NSW investors have an advantage. The NSW Rent and Sales report from the Department of Communities and Justice is built from rental bonds actually lodged, published quarterly by postcode and council area and split by dwelling type and number of bedrooms. It is the rent tenants signed for, not the rent landlords advertised, and the count of new bonds sits in the same row as each median. The department itself suppresses any area with 10 or fewer new bonds and flags results from 10 to 30 bonds to be treated with caution. Take that warning seriously, because it applies to every other small-area number in this guide too.

Rental yield: gross versus net

Gross rental yield is annual rent divided by the purchase price. A $640,000 townhouse renting at $600 a week earns $31,200 a year, a gross yield of 4.88%. That is the figure you see on listings and suburb profiles.

Net yield takes the costs off first. For a typical Sydney investment property the list looks like this:

• Property management and letting fees (often charged as a percentage of rent collected)
• Council and water rates
• Strata levies on units and townhouses, which can run to thousands of dollars a year
• Landlord insurance, including building cover on a house
• Repairs and maintenance
• An allowance for weeks without a tenant
• Land tax once your total land holdings pass the threshold (see our NSW land tax guide for how it is calculated)

Interest is left out of net yield so you can compare properties before finance. Your loan is a separate calculation.

The gap between gross and net is where most yield traps live. High strata, an older building with a repair bill coming, or a rent figure that assumes the property is never vacant can all make a high gross yield look better than it is.

The worst version I’ve seen was a forecast rent that had no connection to the market. A client brought me an off-the-plan apartment in a Melbourne high-rise at about $607,000, marketed at a forecast $835 a week with an $8,000 upfront “finder’s” fee. Comparable units in the same building were renting for $500 to $600, and that unit market had seen almost no capital growth in ten years because of oversupply. The yield was built on a rent nobody had tested against what comparable units were actually leasing for. Half this job is pointing that out before the contract is signed.

Demand drivers: price-to-income, population and jobs

Price-to-income compares what homes sell for with what local households earn. Divide the median sale price by the median annual household income. You can build the price side from actual sales: NSW makes bulk property sales information available from the Valuer General, with files generated weekly for each council area and history back to 1990. Income comes from the ABS through Data by Region and Census area summaries. A high ratio does not mean prices will fall. It does mean growth in that area depends more on buyers arriving with bigger incomes or more equity, because the existing locals are already stretched.

Population growth comes from the ABS Regional Population release, published each year down to small statistical areas (SA2s) and council areas. The latest edition, released 31 March 2026, covers the year to 30 June 2025; Greater Sydney grew by about 75,200 people, or 1.4%, over that year. For the future, the NSW Department of Planning publishes population projections to 2041, which it describes as planning assumptions rather than targets.

Jobs come from the federal government’s Small Area Labour Markets series: quarterly unemployment estimates for SA2s and council areas. The department warns that these are modelled estimates, so look at the direction over several quarters rather than a single reading. What you are testing is whether the area depends on one employer or industry. A town that relies on one large employer can look cheap and high-yielding until that employer cuts back.

Supply: building approvals, the housing pipeline and new builds

Supply is the number most investors skip and the one that most often explains a rent or price that stalls.

• The ABS Building Approvals release comes out monthly, with data by council area and SA2. Compare approvals over the past 12 months with the number of dwellings already there.
• The NSW housing supply dashboard tracks homes lodged and approved (updated monthly) and completions (quarterly), by region and development type. It exists to track the state’s target of 377,000 new homes by June 2029, so expect some council areas to have large targets attached.
• The NSW Planning Portal spatial viewer shows the zoning and planning controls on a specific lot and the streets around it.

Approvals are not completions. Some approved projects are never built, and some take years. But a council area approving new units at several times its historical rate is a council area where rents on existing units may have to compete.

Off-the-plan apartments carry two extra risks that show up in the loan rather than the brochure. The lender values the property close to settlement, not at the contract date, so the valuation can land above or below the price you signed for, and a crowded completion period nearby is one reason it lands below. And on small or high-density units, some lenders lend investors less of the value than they would lend an owner-occupier, around 5% less at some. Our studio apartment investment loans guide covers where the size limits sit.

Why new builds are now a data point

The negative gearing and capital gains tax reforms announced in the 12 May 2026 Budget are now law, and they apply from 1 July 2027. According to the Australian Taxation Office and the Budget tax explainer:

• Negative gearing on residential property is limited to new builds from 1 July 2027.
• Properties held at 7:30pm AEST on 12 May 2026 (including those under contract) are exempt and can keep being negatively geared until sold.
• An established property bought after that time can be negatively geared only until 30 June 2027. After that, its losses can only be offset against residential property income, with any excess carried forward.
• A “new build” has to add to supply. A dwelling that has been sold before does not qualify, unless it was first owned by the builder and not occupied for more than 12 months. Later buyers of a new build lose access to negative gearing on it.

The practical effect for your research: a weekly cash shortfall on an established property bought now is money you carry without a tax offset against your salary from 1 July 2027. Net income after interest is what decides whether that is affordable, so run the numbers at your actual loan amount and rate. Your accountant confirms your own tax position; the point here is that the data you gather now has to include “new or established”.

Infrastructure: funded, not announced

Infrastructure matters when it changes how easily people get to work. Before you pay for it in the price, check that it is funded, under construction and dated:

• Transport for NSW’s project list for rail, metro and road projects and their stages
• The federal infrastructure department for national projects. Its page on Western Sydney International Airport gives 25 October 2026 as the date of the first passenger flights.

A project that has been public for years is usually already in the price of the streets closest to it. The better research question is what has been funded recently and how long until it changes a commute.

How to analyse property market data without being fooled

Small samples

Any figure built from a few sales or a few bonds can swing wildly from one quarter to the next. Check the count behind every median, and prefer a 12-month view to a quarterly one in small suburbs.

Houses and units mixed together

A suburb median can move because the mix of sales changed, not because values did. Split houses and units, and bedrooms where the data allows.

Forecasts dressed as data

A rent estimate from a marketing agent, a growth “prediction” or a suburb page with machine-generated figures is not evidence. Check it against a primary series, and against comparable lodged rents.

Low vacancy with a big pipeline

Tight vacancy today plus heavy approvals nearby is a warning sign. The vacancy figure describes the market before those homes arrive.

Old numbers

Census income data is refreshed only every five years, and regional population estimates are published around nine months after the period they cover. Note the reference date of every number you use.

How the numbers feed your loan

This is the part of the research most investors meet for the first time at application, and it is where a broker is most useful.

Rent has to be evidenced

Lenders want a signed lease or a written rental appraisal from a local agent, and on a newly completed property some want an established rental history first. A printout from a listing site is not accepted as evidence of rental income. I’ve seen an agent’s appraisal come in above expectations and lift a client’s approved loan by about $10,000 at the last minute. Short-stay income generally doesn’t count at all. Lenders want permanent rental income.

Lenders don’t count all of the rent

Most lenders count about 80% of it to allow for vacancy and costs, which is called rental shading. A handful count between 85% and 95%, sometimes depending on the LVR or the postcode. Some lenders also cap the rental yield they will accept for servicing, so an unusually high rent on a cheap property can be cut back in the calculation.

Repayments are tested at a higher rate

APRA requires the banks it regulates to test your repayments at the loan’s interest rate plus a buffer of at least 3%, the level APRA has held since 2021. The property’s rent sits inside that test alongside your other income and your living expenses.

Interest-only lowers borrowing power

An interest-only period cuts the repayment, but lenders assess you as if the whole loan is repaid over the remaining term after the interest-only years. A five-year interest-only loan on a 30-year term is assessed as principal and interest over 25 years, so your capacity drops, not rises.

The 2026 tax changes are in the calculators

Since the Budget, lenders have updated their servicing calculators, and they don’t all treat negative gearing the same way. Some assess an investment differently depending on whether the contract was signed after 7:30pm AEST 12 May 2026 and whether the property is a new build. The same purchase can produce a different borrowing figure at different lenders.

LVR decides LMI

Your loan-to-value ratio is the loan divided by the lender’s valuation. Borrow more than 80% of the value and you will usually pay lenders mortgage insurance, and at least one of the main mortgage insurers adds a loading to the premium on investment loans. Some lenders will lend investors up to 90%.

Structure matters as much as the rate

When an investor uses equity from their home, we usually keep the existing home loan as it is, put the equity release in its own split, and give the new property a standalone loan secured only by itself. Three loans with three clean purposes stops the lender holding one property hostage when you sell the other, and lets your accountant see at a glance which interest is deductible. Our guide to cross-collateralisation explains the risk.

Worked comparison: two hypothetical suburbs

Suburb A and Suburb B are invented to show the method. They are not real places and the figures are illustrations, not market data. Both purchases are established properties bought today with a 20% deposit.

What the numbers say:

1. The yield gap mostly disappears after costs. Suburb A’s gross yield is more than a full per cent higher, but after strata and a longer vacancy allowance the net gap is about 0.2%.
2. A still supports the loan better. Each $100,000 borrowed on A is supported by about $1,055 more counted rent a year than on B. B leans harder on your salary to pass serviceability.
3. B carries less supply risk. A’s council area is approving new dwellings at five times B’s rate, which could hold A’s rent back once those homes are finished.
4. Neither can be negatively geared from 1 July 2027. Both are established and bought after 12 May 2026. If either runs at a loss after interest, that loss can no longer reduce the tax on your salary. It can only be offset against residential property income or carried forward, so you fund the shortfall from your own cash flow in the meantime.

The last step is the one most people skip: put your income, debts and the loan amount for each into our borrowing power calculator at current rates, or have us run it across the lenders that suit your situation. The property that looks right on a suburb profile is sometimes the one your borrowing capacity cannot carry.

Frequently asked questions

What data should I look at before buying an investment property?

Start with eight numbers: vacancy rate, days on market, median rent from lodged bonds, gross and net rental yield, price-to-income, population and jobs growth, the supply pipeline (building approvals and housing completions) and funded infrastructure. Read them together, check the sample size behind each one, and note whether the property is a new build or established, because that now affects negative gearing.

Where can I get property data at no cost in Australia?

The ABS publishes population, income and building approvals data by council area and small statistical area. In NSW, the Department of Communities and Justice publishes median rents from lodged bonds, the Valuer General releases weekly property sales files, and the Department of Planning runs a housing supply dashboard and the Planning Portal. SQM Research shows vacancy rates by postcode, and the major portals’ suburb profiles show median prices and days on market.

What is the difference between gross and net rental yield?

Gross yield is annual rent divided by the purchase price. Net yield takes costs off the rent first: management, council and water rates, strata, insurance, repairs, vacancy and any land tax. Net yield is the more useful number, because a property with a high gross yield and high strata can end up earning less than a lower-yielding house.

How do banks count rental income on an investment loan?

Most lenders count about 80% of the rent when they assess your borrowing power, and a handful count between 85% and 95%. You need a signed lease or a written appraisal from a local agent as evidence; a listing-site printout is not accepted, and short-stay income generally isn’t counted. Some lenders also cap the rental yield they will use.

Is a low vacancy rate always a good sign?

No. A low vacancy rate is good for letting the property today, but if the area also has a large number of new dwellings approved, rents can come under pressure once they are finished. Check the vacancy trend over several years and compare it with building approvals for the same council area.

Do the 2026 negative gearing changes affect which property data matters?

Yes. From 1 July 2027, negative gearing on residential property is limited to new builds, while properties held at 7:30pm AEST on 12 May 2026 are exempt. An established property bought after that time can’t be negatively geared from 1 July 2027, so its net yield and cash shortfall matter more than before. Check the ATO’s guidance and speak to your accountant about your own position.

How much deposit do I need for an investment property?

To avoid lenders mortgage insurance you generally need a 20% deposit plus stamp duty and purchase costs, which means borrowing no more than 80% of the lender’s valuation. Some lenders will lend investors up to 90% of the value, and some allow investors less on small or high-density units. Many investors use equity in their existing home instead of cash.

Talk to us before you shortlist

Research tells you which properties make sense on paper. We check whether they make sense to a lender: how your rent will be counted, what the tax changes do to your servicing, and how to structure the loans so each property stands on its own. See our investment loans page or call us before you make an offer, and we’ll run your shortlist against the lenders on our panel.

This article contains general information only and does not constitute financial advice. Your personal financial situation, objectives and needs have not been considered. Before acting on any information, you should consider its appropriateness to your circumstances. Speak to a qualified mortgage broker for advice tailored to your situation. Mortgage World Australia Pty Ltd is a credit representative (CR No. 396946) of Mortgage Specialists Pty Ltd (Australian Credit Licence No. 387025).

Patrick O'Brien

Patrick is a Director and a Home Loan Specialist. He has been helping Australians with home loans since 2001. Prior to working as a mortgage broker Patrick was employed by Macquarie Bank for 3 years and also worked as an accountant for a publicly listed company. Patrick’s qualifications include:

Bachelor of Business, UTS Sydney. Majored in accounting and sub-majored in Finance and Marketing.
Diploma of Finance and Mortgage Broking Management FNS50310
Certificate IV in Financial Services (Finance/Mortgage Broking) FNS40804