Methodology

Modeling principles

• Assumptions first: key drivers (rates, returns, costs, taxes, timing) are inputs, not hidden defaults.
• Cash flow + balance sheet together: affordability and long-run outcomes are modeled in the same run.
• Opportunity cost is explicit: capital tied up in one choice is capital not compounding elsewhere.
• Mechanics over narrative: the model tracks flows and frictions (fees, taxes, transaction costs) across time.
• Sensitivity over certainty: the primary output is “what changes when assumptions change.”

Scope

Tepuy calculators cover Australian personal finance decisions: property vs shares, buy vs rent, offset account vs shares, mortgage repayments, retirement feasibility, stamp duty, LMI and first-home deposit schemes, plus a set of embeddable single-purpose widgets. Where tax treatment is included, it is implemented as a modeling component rather than advice.

What Tepuy does not attempt

• Forecasting property prices, market returns, inflation, or interest rates.
• Optimizing portfolios or constructing recommended asset allocations.
• Personal advice based on your circumstances, risk tolerance, or goals beyond what the inputs encode.
• Replacing a registered tax agent or licensed adviser on the tax positions described below.

Tax year modelled

The site models the FY2026–27 Australian tax year. The financial year is resolved at run time from the browser’s date rather than hard-coded: July–December resolves to that calendar year, January–June to the year before, so a session in September 2026 resolves to an FY start year of 2026.

The site does not contain an income tax bracket table and does not compute your income tax from your salary. Every calculator that needs a marginal rate takes it as an input — a single flat rate you supply (the Property vs Shares default is 37%, bounded 0–47%). This is a real simplification: it ignores bracket creep, the Medicare levy, offsets, and the fact that a large capital gain can push you into a higher bracket in the year you realise it. Where a calculation is sensitive to that, use a rate that reflects your position including the gain, not your usual rate.

Capital gains tax, and the 1 July 2027 transition

The Treasury Laws Amendment (Tax Reform No. 1) Act 2026 (Act No. 49 of 2026) and the Income Tax Rates Amendment (Tax Reform No. 1) Act 2026 (Act No. 50), both given Royal Assent on 26 June 2026, change how individuals are taxed on capital gains. This is enacted law, not a proposal. From 1 July 2027, for Australian resident individuals, trusts and partnerships, the 50% CGT discount is replaced by cost-base indexation for assets held at least 12 months, with a 30% minimum tax rate applied to the net capital gain after indexation.

What the trigger actually is

This is the single point most commonly misstated, so the calculators are explicit about it: CGT does not turn on the purchase date. It turns on when the gain accrues relative to 1 July 2027. A property bought in 2005 and sold in 2030 is not “grandfathered” for CGT purposes. The separate 12 May 2026 date that appears elsewhere on this site governs negative gearing only — see the next section — and has no bearing on CGT.

What survives the reform

• The 50% discount still applies in full to CGT events before 1 July 2027, whenever the asset was bought.
• It remains available by election for eligible new residential dwellings.
• The affordable housing discount of up to 60% is retained in full; the engines cap the election at 60%.
• SMSF, SMSF-pension and company structures are unchanged by the reform and are modelled unchanged (SMSF one-third discount, SMSF pension phase nil, company 30% with no discount).

How a straddling gain is split — a stated modelling choice

A gain that accrues across 1 July 2027 has to be divided between the two regimes. The engines apportion the gain by time held on either side of that date: if 30% of the holding period falls before 1 July 2027, 30% of the gain is taxed on the old basis (50% discount at your marginal rate) and 70% on the new basis (indexed real gain at the greater of your marginal rate and 30%).

Indexation, and why it is indicative

The post-reform slice indexes the cost base by your assumed CPI rate compounded annually over the holding period, then taxes the resulting real gain. Real ATO indexation uses published quarterly CPI index numbers. Annual compounding of a single assumed rate is therefore an approximation of the mechanism, on top of the larger uncertainty that nobody knows what CPI will be over a 20-year hold. Treat the indexed figure as indicative of the shape of the outcome, not as a return-ready number.

The 30% floor

The post-reform slice is taxed at max(your marginal rate, 30%). The floor binds for anyone whose marginal rate is below 30% — a low-income earner realising a large gain pays more under the new rules than their marginal rate alone would suggest.

Known limitation, documented rather than built: the Act exempts income-support and Age Pension recipients from the 30% floor in their disposal year, taxing them at indexed marginal rates instead. None of these calculators models Age Pension status, so there is nothing to hook that exemption onto — the floor is applied unconditionally. If you are in that group, these figures overstate your CGT.

Worked figures, straight from the engine

A $900,000 asset growing at 6.5% for 20 years (sale value $3,171,281, gain $2,271,281), CPI assumed at 2.5%:

• Disposal wholly before 1 July 2027, marginal rate 37% — 50% discount: $420,187
• Disposal wholly after, marginal rate 37% — indexed, floor not binding: $627,715
• Same, marginal rate 20% — the 30% floor binds: $508,958

The third figure is higher than a naive 20%-of-gain calculation precisely because of the floor. It is the case most likely to surprise, which is why it is stated here.

Negative gearing

Negative gearing for established residential property purchased after 7:30pm AEST on 12 May 2026 is abolished from 1 July 2027. Property purchased before that moment is grandfathered indefinitely for this purpose, and new builds (never previously occupied) keep traditional negative gearing.

In the Property vs Shares engine this is implemented as loss quarantining: for an established investment property, a net rental loss cannot offset your other income. The loss is carried forward, applied against future rental profits as they arise, and any unabsorbed pool remaining at sale reduces the capital gain. For a new build, losses offset other income immediately, as before.

Limitation: the calculator has a new-build / established toggle, but no purchase-date input. It therefore models the post-1-July-2027 regime for any established property you enter. If you already own an established property bought before 12 May 2026, you are grandfathered and the calculator’s tax-refund line understates your position — select “new build” to see the un-quarantined treatment, keeping in mind that this also changes the CGT election available to you.

Property acquisition and holding costs

Stamp duty

Duty is computed by a single shared engine covering all eight jurisdictions — NSW, VIC, QLD, WA, SA, TAS, ACT and NT — with separate general, owner-occupier and first-home-buyer schedules where a jurisdiction has them, and a live purpose distinction between owner-occupied and investment purchases. The rate schedules were last checked against the live government calculators on 30 August 2026.

Example: an existing $900,000 NSW investment purchase returns $34,688. Duty schedules change at state budgets; the verification date above is the honest currency of this figure, and it should be re-checked each financial year.

Investment properties in the Retirement Planner

Each property grows at its own rate, its rent grows with inflation, and its loan is serviced from the shared investment loan rate. There is no loan term input: how long a loan runs follows from the balance, the rate and the monthly repayment you enter. A blank repayment is interest-only — the balance never reduces and interest is charged every year until the property is sold. A repayment below the interest causes the balance to grow rather than fall; the planner warns when you enter one.

On sale, the loan is repaid and capital gains tax is deducted through the same CGT engine used everywhere else on this site. The remainder is added to your non-super savings and earns the same after-tax return as the rest of that balance — about 5.0% a year on the default 7% return, 0.85% fees and 30% tax applied to the income portion of the return only. Super contribution tax does not apply to sale proceeds, because they never enter super.

Selling costs are not modelled. No agent commission, legal, or marketing cost is deducted — the proceeds are the sale price less the loan and the CGT, and nothing else. At a typical 2–3% of sale price, a $1.5M sale would really cost $30,000–$45,000 more than the plan shows. If a sale price would not cover the loan and the tax, the planner reports $0 proceeds rather than a residual debt, and warns that your real position is worse by the difference.

Negative non-super balances are held at zero. Where interest exceeds rent by enough to push non-super savings below zero, the planner floors them rather than carrying a debt forward. This is deliberate: the model has no borrowing facility, so a carried negative would compound at the portfolio return rather than at a real borrowing rate — wrong in a different and less visible way. The effect is that a strongly negative-geared plan is shown slightly better than it would really be.

Lenders Mortgage Insurance

LMI is estimated from a nine-band LVR table (80–82% through 95–97%), with three loan-size brackets within each band (up to $300k, up to $600k, above $600k), a ~15% loading for investment purposes, and stamp duty on the premium added at a default 10%. No LMI is charged at or below 80% LVR; above 97% the tool declines to quote rather than guess, and loans above $3M are referred out.

Example: a $600,000 loan at 85% LVR returns $6,534 at a 0.99% base rate. Limitation: LMI premiums are set by individual insurers and lenders and are commercially variable. This repository carries no dated source for the premium table, unlike the duty schedules. Treat LMI as an order-of-magnitude estimate and get a real quote before committing.

Land tax

Annual land tax on investment property is modelled for all eight jurisdictions, on a per-state bracket formula applied to the property value.

Depreciation

Division 43 capital works is claimed at 2.5% of construction cost per year for the remaining eligible period; Division 40 plant and equipment is depreciated on a diminishing-value schedule per item. Both feed the taxable-income line, and therefore the negative gearing pool above.

Australian Government 5% Deposit Scheme

Effective 1 October 2025 the First Home Guarantee was renamed the Australian Government 5% Deposit Scheme. Income tests were removed entirely, the 35,000-place annual cap was removed, the Regional First Home Buyer Guarantee was merged into it, and price caps were raised. Property price caps in the engine, verified against firsthomebuyers.gov.au on 1 September 2026:

• NSW $1,500,000 capital · $800,000 other
• VIC $950,000 · $650,000
• QLD $1,000,000 · $700,000
• WA $850,000 · $600,000
• SA $900,000 · $500,000
• TAS $700,000 · $550,000
• ACT $1,000,000 (single tier) · NT $750,000 · $600,000

The higher tier covers capital cities and designated regional centres. The cap that actually applies to you is set by postcode, and suburbs can span postcodes in different tiers — so a figure here is indicative and must be confirmed against the official price cap tool before you rely on it. The single-parent pathway (2% deposit, single parents and legal guardians) is part of this same scheme and reads the same cap table — it was the separate Family Home Guarantee until 1 October 2025. It carries no income test, no place cap and no waitlist. What differs is the deposit (2% rather than 5%, with the government guaranteeing up to 18% of the property value) and the eligibility test: you must be single, be the parent or legal guardian of one or more dependent children, and apply alone. It is not a first-home-buyer test — prior ownership in the past 10 years is permitted provided no other property interest remains once the new home settles.

Superannuation

• Concessional contributions cap: $32,500 for FY2026–27 (resolved by financial year, not hard-coded — earlier years are carried in the same table: $30,000 for FY2024–25 and FY2025–26, $27,500 for FY2021–22 to FY2023–24).
• Non-concessional contributions cap: $130,000 for FY2026–27.
• Transfer balance cap: $2,100,000 for FY2026–27 (ATO-indexed 1 July 2026).
• Superannuation Guarantee rate: 12%.
• Contributions tax: 15% on concessional contributions; 15% on accumulation-phase earnings; nil in pension phase.
• Pre-60 withdrawals: modelled through a single effective rate you supply (default 17%), rather than by decomposing your balance into taxable and tax-free components.

Caps are indexed and change. These are the figures the engines currently carry; check them against the ATO at the start of each financial year.

Monte Carlo simulation

The Retirement Planner runs 1,000 simulated paths. Annual returns are drawn from a lognormal distribution calibrated so that the median simulated path equals the return you entered — because a return quoted as “8% p.a.” is a compound annual growth rate, not an arithmetic mean. The arithmetic mean of the distribution therefore sits slightly above your input, by roughly σ²/2, which is correct.

Volatility is taken from the portfolio preset you choose: cash 1.5%, conservative 6%, balanced 10%, growth 13%, high growth 16%, and 12% for Custom. Inflation is simulated as a correlated series with its own volatility (0.5% for cash rising to 1.4% for high growth, and 1.0% for Custom), rather than held fixed. A partner, when enabled, is simulated stochastically alongside you — sharing the same annual market draw, but with their own balances, contributions and retirement age. An adaptive withdrawal rule reduces the draw by 10% when the portfolio falls more than 20% below plan.

How the success probability is banded

The proportion of paths that do not run out of money is reported against four bands. These are the engine’s own thresholds, and every surface on the site uses them:

• Strong — 85% and above
• Moderate — 70% to below 85%
• At risk — 55% to below 70%
• High risk — below 55%

A probability is a property of the assumptions you entered, not a measurement of your future. Moving your assumed return by one percentage point will move this number more than most people expect — which is the argument for reading it as a sensitivity, not a score.

The Offset vs Shares calculator also offers a Monte Carlo view; it runs 300 paths, not 1,000.

Cash-flow modeling

The core unit of analysis is a time-series of cash flows. Depending on the calculator, this may include contributions, repayments, rent, dividends, expenses, and tax impacts. The model then tracks how cash flows affect:

• Net cash position: the cumulative cash burden or surplus over time.
• Asset values: property value, share portfolio value, and associated equity.
• Liquidity constraints: whether a scenario remains feasible under the chosen assumptions.

Share parcels are tracked on a FIFO ledger, so a partial sale disposes of the oldest parcels first and each parcel carries its own holding period into the CGT calculation. Dividends are grossed up for franking credits at the 30% corporate rate and the excess credit is refunded or applied according to the structure you selected.

Timing conventions

Tepuy uses consistent timing rules (e.g., year 0 vs year 1 treatment). Where relevant, the calculators avoid “free money” artifacts such as charging interest or earning rent in periods where the model states those cash flows should not yet apply.

Opportunity cost and comparability

Many comparisons are biased because they unintentionally give one option “extra capital” or “extra cash flow.” Tepuy aims to compare options under the same economic constraints by making these quantities explicit:

• Upfront cash: deposit, stamp duty, legal fees, and other acquisition frictions.
• Ongoing cash burden: net costs after rent/income, maintenance, fees, and tax effects.
• Alternative investment return: what the forgone capital could earn elsewhere.

Known limitations

These are the places where a calculator on this site will give you an answer that is narrower, or less current, than it might appear. They are listed because knowing the edge of a model is part of using it.

• Beat the ASX uses one national rental yield. Rent is estimated as a flat 3.5% gross yield of the purchase price everywhere in Australia. Because every other term in that comparison also scales with price, the price cancels out and the breakeven growth rate is identical at every price — 5.49% at the shipped defaults, the same for a $400,000 home and a $3,200,000 one. The tool cannot tell you that one city clears the bar more easily than another, and it is not a suburb ranking. Choosing your suburb changes the dollar figures, not the percentage.
• Beat the ASX suburb medians carry no capture date. The 624 suburb medians are indicative approximations for 2026. There is no dated source for them in this repository, so the page states them as approximate rather than attaching a precision it cannot support.
• Offset vs Shares does not model the 1 July 2027 CGT reform. It applies a flat 50% CGT discount (33.33% inside super) regardless of when the gain accrues. On a post-2027 disposal this understates the CGT payable — by roughly 37% on a 5-year hold, 47% on 10 years and 68% on 25 years. The calculator carries a banner saying so. Use Property vs Shares if the CGT treatment is what you are trying to decide.
• No income tax engine. Marginal rates are inputs, not computed. See “Tax year modelled” above.
• No HECS/HELP engine. Where this site discusses student loan repayments it does so in prose only; no calculator models compulsory repayments or indexation.
• No Age Pension modelling, which is why the CGT 30% floor cannot be waived for income-support recipients as the Act allows.
• No purchase-date input for negative gearing grandfathering. See that section above.
• Nothing here is forecast. Growth rates, returns, inflation and interest rates are all assumptions you supply. The calculators do not predict them and do not have a house view on them.

Sensitivity and interpretation

Tepuy outputs should be read primarily as sensitivity analysis: they show which inputs dominate outcomes. In many scenarios, a small number of parameters drive most of the spread (e.g., interest rates, rent yield, growth rates, time horizon).

Recommended usage

• Run a base case with conservative inputs you are willing to defend.
• Run stress cases (rates higher, vacancy higher, costs higher, returns lower).
• Look for break-even regions rather than a single “winner.”

Definitions (common outputs)

• Equity: asset value minus liabilities (where modeled).
• Net cash position: cumulative cash outlay/surplus over time.
• NPV: present value of a scenario under a chosen discount rate (where included).
• IRR: discount rate that makes NPV equal to zero (where included).

If a specific calculator implements a definition differently, that calculator should be treated as the source of truth for that context.

Disambiguation

Tepuy Solutions is not affiliated with thetepuy.com or other non-financial brands using the name “Tepuy”.

Related links

• Browse calculators
• About Tepuy Solutions
• Articles
• Contact