How to find investment properties in Australia — a 10-step analyst’s framework.

Every online bank calculator will tell you one number: the absolute maximum they’d lend you at best-case assumptions. That number is almost never what you’ll actually be approved for, because APRA requires lenders to assess serviceability at your rate plus three percentage points, and most brokers submit applications closer to 80% of the raw max for approval comfort.

Start with the real number. A $600K capacity at current rates becomes a $420K–$480K submission cap once you apply APRA’s buffer plus a 30% debt-to-income guardrail. That’s the ceiling you should be shopping against — not the bank’s marketing figure.

Factor in the drag from existing obligations. HECS reduces assessed income by up to 10%. Credit card limits count as ~3.5% of the limit in potential debt (a $20K card reduces capacity by ~$50K whether you use it or not). Car loans count at full repayment. Close unused cards and refinance car loans to shorter terms before submitting — this step alone can add $70–100K of capacity.

Four deposit strategies make sense in Australia: 5% deposit + LMI, 10% deposit + LMI, 20% deposit + no LMI, or the Home Guarantee Scheme (HGS, a.k.a. FHLDS) if eligible. Each changes your capacity, your repayment, and your resilience in different ways.

Rule of thumb:in a market rising 5%+ pa, waiting 12 months to save an extra 15% deposit costs you more in appreciation than the LMI you’d have paid. In a flat or falling market, waiting wins. Run all four scenarios with today’s numbers and next-year’s savings before deciding.

HGS is a structural edge if you qualify — 5% deposit with zero LMI, up to $900K in Sydney/Melbourne metros, $600K in regional states. Income caps are $125K single / $200K couple. Places are allocated each financial year and run out; applying early matters.

Three clean strategies exist in AU property investing: cashflow-first (high yield, modest growth, often regional or outer metro), equity-first (low yield, high growth, inner metro or gentrifying suburbs), and balanced (middle-band suburbs, 3.5-5% yield + 4-6% growth expectations).

Pick one before you shortlist. Every strategy has different suburb profiles, different acceptable vacancy levels, different depreciation assumptions, different exit horizons. Mixing strategies at the shortlist stage is how buyers end up with a low-yield property in a low-growth suburb — neither wins.

Your tax position pushes the answer. If you’re in the 37-45% bracket with strong other income, equity-first + negative gearing is usually the right play — the refund converts the holding cost into a tax-arbitrage position. Lower tax brackets or irregular income → cashflow-first, where the property pays for itself without needing a refund.

Most online suburb rankings run on scores built from private data — aggregators layer quality-controlled feeds with undisclosed weights and publish a single number. The result looks like analysis but is marketing. Two suburbs with the same “Growth Score 8.5” often have completely different actual growth histories when you pull the raw state sold-price register.

Stick to primary sources. State sold-price registers for median transacted values. Official statistics (census and public table aggregations) for demographics, household structure, dwelling stock. Government records for income distribution by postcode. Licensed market data for rent + vacancy. Crime panels for local offence rates by LGA. Weather and hazard sources for climate and flood data. Every one of these is either free to query or comes with a published API.

The filters that matter: 12-month growth (with the direction of the 24-month), 12-month rent growth, current vacancy < 1.5%, median sale price within your capacity band, and household median income high enough that the ratio of median price to median household income is under 8× (above that is where affordability stress starts biting demand).

For every suburb that survives step 4, run gross + net + break-even rent + +100bps rate-stress at the suburb’s actual median rent and actual current listing prices. Not the median sale price — listings, because that’s what you’d actually pay.

The number that matters is break-even rent: the weekly rent at which net cashflow after all holding costs hits zero. If the current median rent clears break-even by at least $50/week, the suburb’s rent market is paying for itself. If break-even rent is above the current median, you’re subsidising the property every week — only viable if growth assumptions are strong and your marginal tax rate converts the loss into a refund.

Rate-stress at +100bps is the APRA-lite test: if rates rise 100bps from here, does break-even still hold? If the answer is “no, I’d be $4K/yr underwater on a rate rise,” you need either higher capacity or a lower price — not a hope that rates fall.

Data filters the set; feet decide it. Drive or walk every shortlisted suburb at three different times: a weekday morning (commute pressure), a Saturday at 10am (owner-occupier vibe), and a weekday at 8pm (noise, lighting, how safe it feels).

Things aggregator data can’t capture: flight paths (listen), train noise, sinking drains, parking congestion, whether the school-walk is actually walkable, whether the “shops” on the map are occupied or boarded up, how many utes vs family cars, whether the local cafe is busy.

If you can’t walk the suburb in person (interstate buyer), get photos at three times and satellite-map the school-walk and commute. Skipping this step is how buyers end up with a listing that shows perfectly online and is unrentable in practice.

Every candidate property gets a 10-year projection. Not gross yield, not a one-year cashflow — the full table, with rent indexation, loan amortisation, depreciation decay, and CGT at year-10 sale. The output that matters is IRR (internal rate of return) on your actual cash invested.

A leveraged 80% LVR investment property in an Australian capital city historically delivered 8-12% IRR over rolling 10-year windows — the mix of cashflow yield (~3%) and leveraged capital growth (~5-7%). Under 6% IRR is worse than low-cost index funds after risk. Over 12% usually needs above-average growth to sustain — possible but not underwrite-able.

Run three scenarios: conservative (3% growth, 2.5% rent, +100bps rate shock year 3), base (5% growth, 3% rent, flat rates), stretch (7% growth, 4% rent, −50bps by year 6). You’re looking for IRR that clears 7% at conservative, 9%+ at base — that’s a position you can hold through a cycle.

The tax position is half the investment. Two numbers matter: annual refund from negative gearing (if the property runs at a tax loss) and CGT at year-10 sale (assuming you’re not holding forever).

Annual refund = property loss × your marginal rate, capped at tax actually paid on other income. Every deduction line has a specific ITAA 1997 section: interest under s8-1, repairs under s25-10, Div 40 plant & equipment via s40-25 diminishing-value, Div 43 capital works via s43-10 at 2.5%/yr. If you don’t get a quantity surveyor report for Div 43, you’re leaving $3K-7K/yr in refund on the table for a modern apartment.

CGT at exit: accumulated Div 43 reduces cost base (s110-45(2)), 50% discount applies if held 12+ months (s115-25), main residence exemption if PPOR-and-let under the 6-year rule (s118-145), joint ownership split taxes each partner at their own rate. Model both the annual refund and the year-10 CGT before offering — the answer changes the maximum you can pay.

When the shortlist is done and a property pencils, offer fast and clean. Pre-approval current and written, conveyancer briefed, finance clause short (7-10 days, not 21). Building & pest pre-inspected where possible.

Price anchored to comparable sales you pulled yourself from the state sold-price register — not the agent’s guide price. Agents list to “close the gap” between buyer and seller expectations, which usually means the guide is 5-10% below true market. Your offer should be backed by three actual sales in the same suburb + same bed count in the last 6 months.

Stamp duty, conveyancing, building & pest, loan fees, LMI are all baked into the buying-power calculator output — don’t promise more than the number supports.

The decision doesn’t end at settlement. Every 12 months, re-run the 10-year projection with: actual rent received, actual costs incurred, actual maintenance events, current market value from recent comparable sales. The gap between what you projected and what actually happened is the most valuable signal you have — it tells you whether to hold, refinance, rent-review, or sell.

If actual cashflow is tracking base-case or stretch, hold. If it’s tracking conservative-case plus some, refinance to optimise (principal-paid buffer, offset account, rate-hunt). If actuals are materially below conservative and the gap is widening, model a sale against the CGT bill — sometimes rotating capital out of a weak property into a stronger one is the highest-IRR move of your decade.

Annual review is how the 10-year thesis either proves or breaks. Without it, you’re not investing; you’re holding.