It is one of the most common questions developers ask, and one of the easiest to get quietly wrong: what number do you actually put in the land line of a feasibility? The purchase price seems obvious enough when you are buying. But what if you already own the site? What if there is still a loan against it? What if a landowner is bringing the land into a joint venture instead of selling it to you?
The figure you choose changes your development margin, your profit on cost and your return on equity, sometimes by a wide margin. Get it wrong and a marginal deal can look bankable, or a strong one can look dead. The good news is that for most situations there is clear professional consensus on the right approach, and it comes straight from valuation standards and appraisal theory rather than from anyone’s opinion. This guide walks through each scenario, explains the reasoning, and points you to the primary sources so you can check the logic yourself.
The question your feasibility is actually answering
Before picking a figure, it helps to be clear about what a feasibility is for. A development feasibility answers one core question: does this development create value over and above every resource it consumes, including the land? Everything else follows from that.
There are really two different questions hiding inside most feasibilities, and they call for different land figures:
- Project feasibility. Does the development stack up on its own merits, independent of how you financed it or what you happen to have paid for the land? This is measured by development margin, profit on cost and project internal rate of return (IRR).
- Owner or equity returns. Given how you funded the deal and what you actually paid, what do you personally make? This is measured by return on equity.
The land figure has to match the question. For a project feasibility, the land needs to be costed at its economic cost to the project, which is its market value, regardless of your personal history with it. The clearest authority on this principle comes from the RICS professional standard on the valuation of development property, which directs valuers to assess site value assuming no debt and to handle any debt analysis separately, reporting the two results apart from one another. The point generalises beyond finance: a project appraisal is about the asset and the scheme, not about your loan or your cost base.
Hold that distinction in mind, because it resolves three of the four scenarios below.
If you are buying the site: use the purchase price
When you are acquiring the site, the land input is the contracted purchase price. This is genuinely uncontested. The agreed price is what a willing buyer and a willing seller have settled on at arm’s length, which is the textbook definition of market value, so it is both the actual cost and the market evidence in one figure.
The only real nuances concern acquisition costs and where they sit in the model.
Keep acquisition costs on their own lines
Stamp duty (transfer duty) is usually the largest acquisition cost, and it varies materially by state. On a $1.5 million residential site, transfer duty for a non first home buyer may run to roughly $60,000 to $65,000 in the eastern states, before any surcharges. High value sites can attract premium rates: in NSW, for example, a premium duty rate applies to residential land above an indexed threshold (around $3.7 million for 2025 to 2026). You can model the exact figure for each state using Feasly’s stamp duty calculators, and the NSW transfer duty guide sets out the current brackets.
Foreign purchaser surcharges can add a significant layer on top: these sit around 7 to 9 per cent of the dutiable value in the states that levy them, with NSW currently the highest. If a foreign person or entity is acquiring, model the surcharge explicitly, because it can be the difference between a feasible and an unfeasible acquisition.
Beyond duty, acquisition typically carries legal and conveyancing fees, due diligence costs (surveys, planning advice, contamination and geotechnical investigations) and finance establishment costs. Whether you group these into the land line or a separate acquisition costs line is a presentation choice, not an economic one, as long as they are captured somewhere in total development cost. The cleaner practice is to keep them on a discrete acquisition line so the raw land value stays visible on its own, which matters for the GST and residual land value cross checks below.
Keep the raw land figure clean for GST
There is one place where the composition of the land figure has real tax consequences: the GST margin scheme. Under the margin scheme, GST on the eventual sale is calculated on the margin, which is broadly the sale price less the original land acquisition cost. The ATO guidance on margin scheme valuations and subsequent case law confirm that the acquisition cost for margin purposes is the land purchase price itself. It does not include stamp duty, legal fees or development costs. Bundling acquisition costs into the land figure would therefore misstate your GST. Keep the purchase price as a clean, separable number, and your margin scheme calculation stays correct.
If you already own the land: use current market value
This is the scenario developers most often get wrong, usually by entering what they paid for the site years ago, or by entering nothing at all because “the land is already mine.” For a project feasibility, both are mistakes. The figure you want is the land’s current market value at its highest and best use.
The reasoning rests on opportunity cost, and it is one of the most settled ideas in investment analysis.
Historical cost is a sunk cost, and sunk costs are irrelevant
What you paid for the land is gone. It cannot be changed by anything you decide today, which is the definition of a sunk cost, and sunk costs do not belong in a forward looking economic analysis. Penn State’s investment analysis course material uses almost exactly this situation as its canonical example: an investor who forgoes the cash they could realise by selling a property, in order to keep and develop it, must include that forgone sale value as an opportunity cost in the analysis. The relevant cost of using the land is what you give up by not selling it, which is today’s market value, not yesterday’s purchase price.
Put concretely: if you bought a site for $5 million and it is now worth $15 million, the cost of putting it into a development is $15 million, because that is what you could sell it for today. Develop it, and you forgo that sale. Entering $5 million (or $0) credits the development with $10 million of land appreciation that has nothing to do with how well the development itself performs.
Market value has a precise meaning
Market value is not a loose term. The Australian Property Institute, adopting the international definition, defines it as the estimated amount for which an asset would exchange between a willing buyer and a willing seller in an arm’s length transaction, after proper marketing, with both parties acting knowledgeably and without compulsion. API standards add that the assessment must reflect the highest and best use of the property, where that use is physically possible, legally permissible, financially feasible and maximally productive. That is the basis you want for the land you already own.
Why $0 flatters the result
If you enter $0 (or historical cost) for land you own outright, your development margin and profit on cost are inflated by an amount that has nothing to do with the development’s performance. The project looks more profitable than it is, because you have effectively gifted the scheme a valuable asset for free. Costing the land at market value isolates the development return from the land appreciation return, so you can see whether the act of building, rather than the act of having bought well years ago, actually creates value. Profit on cost is prized precisely because it is hard to flatter with financial engineering, and a clean land figure is what keeps it honest.
The legitimate second view
There is a defensible reason to also model the land at your actual cost base, but only to answer the other question: what is your total cash return as the existing owner? That is an owner or equity view, and it is perfectly valid in its place. The sophisticated approach is to run both. Use market value for the project test, which tells you whether the development stacks up, and use your real cost base for the personal return test, which tells you what you take home. The error is using the owner level figure to claim the project is feasible. Feasly’s opportunity cost modelling is built for exactly this dual view, so you can see the project return and your personal return side by side without conflating them.
If there is a loan against the land: do not use the amount owing
Some owners instinctively reach for the balance still owing on the land loan and enter that as the land cost. For a project feasibility, this does not hold up, for the same reason historical cost does not: the debt balance is unrelated to the land’s value or its opportunity cost.
A site worth $3 million might carry a $500,000 loan or a $2.8 million loan. Neither figure tells you anything about whether developing the site creates value. Using the loan balance simply understates the true economic cost of the land whenever the debt is less than market value, which inflates the margin in the same way entering $0 does. The amount owing is a financing fact about you, not an economic fact about the asset, and the RICS guidance on assessing site value independent of debt is the cleanest expression of why the two should be kept apart.
The one place the loan balance is relevant is a deliberate cash flow or equity position analysis. A question like “if I roll the existing land debt into the construction facility, what is my net cash position and how much equity am I contributing?” is a real and useful question. But it is a funding and serviceability question, properly modelled in the finance layer through LVR and LTC, debt drawdown and repayment, not on the land cost line. Mixing the loan balance into the land figure produces a result that is neither a clean project test nor a clean equity test.
The rule is simple: enter market value as the land cost, and model the actual loan in the funding structure where it belongs.
Land in a joint venture: value it, then treat it as equity
When a landowner brings their land into a joint venture with a developer rather than selling it, the land still needs a value, and that value still needs to flow into the feasibility. Australian practice is to settle on an agreed land value up front in the development agreement, set at an amount broadly analogous to the market value of the land. Law firm guidance such as Hall & Wilcox on drafting development agreements describes this approach, and it is the figure your model should use.
The more interesting question is how that value enters the feasibility, and here treatment is genuinely context dependent.
Two ways to model it
There are two clean approaches, and which one you use comes down to avoiding double counting:
- Land as a cost line. Enter the agreed land value as the land cost, because it is a real economic cost to the project and a real entitlement of the landowner, and model the landowner’s profit share separately. This keeps the project feasibility view complete and comparable to a straight acquisition.
- Land as equity, with $0 on the cash land line. If your funding stack already captures the land as the landowner’s equity contribution, with their land value and return modelled in the capital stack and profit waterfall, then entering it again as a cash cost would double count it. In that case the cash land line is $0 and the value lives in the equity layer.
Both are correct. The mistake is to do neither cleanly, or to do both at once.
The land contribution usually behaves like preferred equity
In most landowner joint ventures, the land contribution functions as equity in the capital stack. The landowner effectively puts the land up as security, which supplies the equity the financier needs, and the landowner is then reimbursed their land value out of the sale proceeds in an agreed order before profit is split. This contribution frequently behaves like preferred equity: it ranks behind the senior debt but ahead of, or alongside, the developer’s ordinary equity, and it often carries a preferred return or a profit hurdle that must be cleared before the developer’s share is distributed.
This is where modelling the structure properly pays off. Feasly’s funding stack can model a landowner contributing land as a preferred equity partner, with the land value and its return profile sitting in the capital stack rather than on the cost line. That keeps the project cost view honest while the profit waterfall reflects the real commercial deal.
Profit share, fixed return, or a hybrid
Landowner joint ventures sit on a spectrum. At one end, the landowner takes a fixed land payment and the developer keeps the upside. At the other, the landowner takes a pure profit share. Many deals are a hybrid: a set land value plus a share of profit. Indicative Australian structures discussed by finance commentators include landowner and developer profit splits in the region of 50/50 after the developer’s management fees, and capital partner preferred returns commonly pitched somewhere in the low to high teens as an IRR hurdle before profit sharing. Treat these as indicative only. They are deal specific and shift with market conditions, not benchmarks to plug in. Feasly’s joint venture guide covers the structures in more detail.
A structuring caveat worth flagging
The legal and tax structure of a joint venture (an unincorporated JV, a partnership, or a unit trust) materially affects GST, stamp duty and liability. Accounting and legal advisers such as RSM Australia note that a true joint venture can avoid the joint and several GST liability that comes with a partnership, and some states apply duty to a developer’s economic entitlement even without a land transfer. These are reasons to get tax and legal advice before signing, not feasibility modelling choices. But the agreed land value the advisers settle on is the figure that flows into your model.
Acquisition costs, stamp duty and GST: keep the land figure clean
A theme runs through all four scenarios: the raw land value should stay a clean, separable figure in your model. Three things depend on it.
- Stamp duty belongs in total development cost, ideally on its own line, calculated for the relevant state. It is part of project cost, but it is not part of the margin scheme acquisition cost, so it should never be folded into the land value.
- The GST margin scheme calculation is driven by the raw land acquisition cost, so bundling other costs into the land figure misstates GST. The margin scheme also has to be agreed in writing before settlement and is unavailable in certain cases, such as where full input tax credits were claimed on the acquisition.
- Residual land value cross checks rely on a clean land figure to be meaningful, which leads to the final point.
How this connects to residual land value
Residual land value (RLV) and this guide are two sides of the same coin. RLV runs the feasibility backwards to tell you the maximum you can afford to pay for the land given your target return. This guide is about the opposite direction: once you know the deal, whether you are paying a set price, own the site, or are partnering with a landowner, what figure do you enter?
The two work together. Use RLV to test whether the price you are about to pay, or the agreed value in your JV, is justified by the scheme. Then enter the actual figure into the feasibility to model the project. Feasly’s residual land value guide and calculator cover the method in full, so this guide does not repeat it.
Quick reference
| Your situation | Figure to enter | Why |
|---|---|---|
| Buying the site | Contracted purchase price | It is both your actual cost and arm’s length market evidence |
| Already own the land | Current market value at highest and best use | The opportunity cost of developing rather than selling; historical cost is a sunk cost |
| Loan still owing on the land | Market value (not the loan balance) | Debt is a financing fact about you, not an economic fact about the asset; model the loan in the funding layer |
| Landowner joint venture | Agreed land value, modelled as equity | A real cost and a real entitlement; treat as preferred equity in the capital stack to avoid double counting |
The through line is straightforward. A project feasibility should cost the land at what it is worth today, so the result reflects the development on its own merits. Your purchase price does that when you are buying. Market value does it when you already own or when a partner contributes the land. The figures that get developers into trouble, historical cost, $0, and the balance owing, all answer a different question than the one a feasibility is meant to answer. Keep the project test and the equity test separate, value the land at market in the project test, and model your actual finance and deal structure in the layers built for them.
This guide is general in nature and does not constitute financial, legal or tax advice. Stamp duty rates, GST treatment and joint venture structuring should be confirmed against current state revenue office and ATO guidance and with your own advisers.