Yield on Cost and the Development Spread in Australia

Yield on cost is the stabilised net operating income (NOI) a completed project will produce, divided by everything it cost to get there. The development spread is that yield on cost minus the capitalisation rate the finished asset will be valued on. Those two numbers, taken together, are how a developer who intends to hold or to sell to an investor works out whether building something is worth more than buying something already built. On a project selling apartments or lots one by one, the profit metric answers the question. On an income-producing project, an office building, a shopping centre, an industrial estate, a self-storage facility or a build-to-rent tower, the spread does.

Before considering an income-producing development on the strength of its yield on cost, you’ll need professional advice from a registered valuer and your accountant, and usually your lawyer as well. The valuer is the one who forms a view on the exit capitalisation rate and on the sustainable rent behind the net operating income (NOI), and those two inputs decide most of the answer. The accountant is the one who works out whether Goods and Services Tax (GST) is recoverable on the build, how land tax will be assessed on the completed asset, and whether the federal build-to-rent concessions are available on your facts. Those judgements sit outside anything a guide can tell you, because they turn on your asset, your ownership structure and the evidence in your market at the time. The consequences land on you rather than on any of them: an exit capitalisation rate that moves 50 basis points against you, or a concession the asset turns out not to qualify for, comes out of your equity. What follows is written to make that conversation shorter and sharper, not to substitute for it.

Figures, rates and thresholds in this guide were current at the time of writing in September 2026 and they move. Capitalisation rates move with the cost of capital, construction costs move quarter by quarter, and the state build-to-rent concessions have been amended repeatedly since 2020. Each linked primary source is the place to confirm the position before a number goes into a live feasibility.

What is yield on cost in property development?

Yield on cost is a project’s stabilised net operating income (NOI) expressed as a percentage of its total development cost. It is sometimes called the development yield, and the two terms are generally used interchangeably in the Australian market. The formula is:

Yield on cost = stabilised net operating income (NOI) ÷ total development cost

A project with a total development cost of $128.6 million that will produce $6.3 million of stabilised net operating income (NOI) has a yield on cost of 4.90 per cent.

What makes it different from every other return metric is that it is ungeared and indifferent to time. It does not care how you funded the project, what your equity return is, or how long the money was out. It is a single ratio comparing what the asset earns to what the asset cost to create, which is precisely what makes it comparable to a capitalisation rate, because a capitalisation rate is a ratio comparing what an asset earns to what someone paid for it. The capitalisation rate is the market’s price on income; the yield on cost is your price on the same income. Putting the two side by side is the whole point.

How do you calculate yield on cost?

You calculate yield on cost by dividing stabilised annual net operating income (NOI) by total development cost, then multiplying by 100. Both halves need care, and in the Australian market both halves carry traps that the offshore explainers do not cover.

The numerator is stabilised net operating income (NOI), not first-year income and not gross rent. The denominator is total development cost, which means every dollar spent to get to a leased, operating asset, including land, construction, consultants, authority charges, finance costs and any non-recoverable GST. A yield on cost built on a construction figure alone, without land or finance, is not a yield on cost. It is a number that will flatter the project by a wide margin, and it is one of the more common errors in a desktop appraisal.

What is stabilised net operating income, and when does it start?

Stabilised net operating income (NOI) is the annual income the asset produces once it is leased and operating at a normal occupancy level, rather than the income it produces in its first month or its first year. The distinction matters because the gap between practical completion and stabilisation can run from a few months on a pre-committed industrial facility to a year or more on a large residential rental building leasing up unit by unit.

Building the figure up, rather than guessing a net number, is the discipline that keeps a yield on cost honest. Net operating income (NOI) is gross income less the operating expenses the owner cannot recover, before finance, tax and capital expenditure. In practice that means:

• Gross potential income, from base rent plus other income such as car parking, storage, utility margins and signage
• Less a vacancy and bad debt allowance, giving effective gross income
• Less operating expenses, being property management, on-site staffing, repairs and maintenance, insurance, council rates, land tax where it is not recoverable, marketing and amenity operating costs

Two things commonly go missing. The first is a capital expenditure or replacement reserve, which is excluded from net operating income (NOI) by convention but does not stop being a real cost, and a yield on cost that ignores it entirely overstates what the asset will actually deliver to an owner. The second is the difference between face rent and effective rent. Where incentives are being offered, and in a soft office market those can be substantial, the rent written on the lease overstates what the tenant is really paying. Capitalising a face rent that the market will only pay effective rent for produces a yield on cost and an exit value that neither survives a valuation.

The timing question is separate and it is often mishandled. A project that reaches practical completion in March and stabilises in the following February has eleven months of partial income, full operating costs and, generally, full interest. Those months belong in the total development cost as a letting-up allowance and as capitalised holding costs, not in the stabilised net operating income (NOI). Pushing them into the numerator makes the yield look better. Leaving them out of the denominator makes it look better again.

What is the development spread, and what does it tell a developer?

The development spread is the yield on cost minus the capitalisation rate the completed asset will be valued or sold on, expressed in percentage points or basis points. It is the compensation for taking development risk rather than buying a finished, leased, operating building.

Development spread = yield on cost minus the exit capitalisation rate

If the yield on cost is 4.90 per cent and the market will value the finished building on a 4.50 per cent capitalisation rate, the spread is 40 basis points. If the yield on cost is 6.80 per cent against a 5.80 per cent exit capitalisation rate, the spread is 100 basis points. A negative spread means the market would price the completed building below what it cost to create, which is a statement about the project rather than about the developer’s construction skill.

The logic is worth stating plainly, because it is the reason the metric exists. An investor can buy a leased building today on the market capitalisation rate and take almost none of the risk that a development carries: no planning risk, no construction risk, no cost escalation, no programme risk, no leasing risk, no interest during a build. A developer who creates the same income stream at the same yield the investor could simply have bought has been paid nothing for absorbing all of that. The spread is the payment. Whether it is adequate payment is a judgement about the specific project, not a number a guide can hand over.

How much development spread do developers look for?

There is no government or regulatory source that sets a target development spread, and any figure quoted is market practice rather than a rule. Ranges commonly quoted for the Australian market sit somewhere around 150 to 300 basis points, with the lower end associated with pre-committed, low-complexity assets and the upper end with speculative, long-dated or planning-heavy projects. Those are indications of how the market has talked about the number rather than a threshold anyone is bound by, and the place to confirm what is genuinely being underwritten today is current valuation evidence and the terms actually being quoted on comparable projects.

Two qualifications tend to matter more than the headline range. The first is that a spread is only as reliable as the exit capitalisation rate it is measured against, and that rate is a market input the developer does not control. The second is that the same spread means very different things over different programmes. A 150 basis point spread earned over an eighteen-month industrial build and a 150 basis point spread earned over a five-year staged residential rental project are not comparable propositions, which is why the spread is generally read alongside a time-weighted measure such as the internal rate of return (IRR) rather than in isolation.

A worked example: how the spread creates value

Take a 200-apartment residential rental building. All costs below are stated inclusive of GST where GST applies, for reasons covered later in this guide.

And the income side, on stabilisation:

The yield on cost is $6,300,000 divided by $128,600,000, or 4.90 per cent. If the market values a completed building of this type on a 4.50 per cent capitalisation rate, the value on completion is $6,300,000 divided by 0.045, which is $140,000,000. The development spread is 40 basis points, and the value created is $11,400,000 over a $128,600,000 cost, before tax and before any gearing.

Forty basis points is a thin spread on a project of this length, and the example is deliberately drawn that way, because thin spreads have been a feature of Australian residential rental feasibilities rather than an exception. The number that should follow immediately is the sensitivity to the one input the developer does not set. Holding the stabilised net operating income (NOI) constant at $6,300,000 and flexing only the exit capitalisation rate:

The row that explains the whole metric is the 4.90 per cent one. When the exit capitalisation rate rises to meet the yield on cost, the spread is gone and the project breaks even on value alone. That is not a coincidence, it is the arithmetic: value equals net operating income (NOI) divided by the capitalisation rate, and when that rate equals net operating income (NOI) divided by cost, value equals cost. A developer who knows their yield on cost therefore already knows the exact exit capitalisation rate at which the project stops creating value, without modelling anything further. On this project it is 4.90 per cent, and it is 40 basis points away.

A second scenario: what a cost movement does to the spread

Change one input. Assume the construction cost rises by 5 per cent between feasibility and contract execution, from $82,500,000 to $86,625,000, and that the longer drawdown profile adds $250,000 of capitalised interest. Everything else, including the stabilised net operating income (NOI), is unchanged.

Total development cost becomes $128,600,000 plus $4,125,000 plus $250,000, which is $132,975,000. The yield on cost falls to $6,300,000 divided by $132,975,000, or 4.74 per cent. Against the same 4.50 per cent exit capitalisation rate, the spread has narrowed from 40 basis points to 24, and the value created has fallen from $11,400,000 to $7,025,000.

A 5 per cent construction movement, which is a modest assumption across the months between a feasibility and a signed contract, removed close to 40 per cent of the value created. Construction costs keep moving, and the denominator of a yield on cost is a moving number in a way that a purchaser’s price is not. A spread quoted off a cost plan that is twelve months old is a spread quoted off the wrong denominator.

How does yield on cost differ from a capitalisation rate, a development margin and an internal rate of return?

Each of the four answers a different question, and using one where another belongs is the most common way a developer and a funder talk past each other.

A capitalisation rate is the market’s number. It is income divided by price, derived from comparable sales evidence, and it prices the finished asset. A yield on cost is your number. It is income divided by your cost, and nobody else can calculate it, because only you know what the project cost to build. The comparison of the two is the development spread.

A development margin is a profit percentage on a sale. Margin on cost, or margin on revenue, measures the profit from selling the completed product against either total development cost or gross realisation. It answers the build-to-sell question. On a build-to-hold project there is no sale, so the margin has to be calculated against a notional value derived by capitalising the income, which means the margin is really just the spread wearing different clothes. In the worked example above, the $11,400,000 of value created is a margin of about 8.9 per cent on a $128,600,000 cost, and that margin is entirely a function of a 40 basis point spread. Quoting the margin without the spread hides where the number came from.

An internal rate of return (IRR) is time-weighted and geared or ungeared depending on how it is built. It answers the question of what the money earned per year, accounting for when it went out and when it came back. The yield on cost is blind to time. Two projects with identical yields on cost and identical spreads can have very different internal rates of return (IRR) if one takes two years and the other takes five.

The practical division of labour tends to be that the spread tests whether the project should exist, the internal rate of return (IRR) and equity multiple test whether the capital structure works, and the development cashflow tests whether the project survives the middle. None of them substitutes for the others.

What goes into the “cost” in yield on cost?

Everything spent to get to a leased, operating asset goes into the cost. That is the short answer, and the long answer is where most of the error sits, because the denominator is the half of the equation the developer controls and therefore the half most vulnerable to optimistic definition.

A defensible total development cost for yield on cost purposes generally includes land and acquisition costs, transfer duty and any foreign acquirer surcharge actually paid, demolition and site preparation, the construction contract sum, consultants, authority and headworks charges, developer contributions, capitalised interest and finance fees, holding costs through construction and lease-up, letting-up costs and leasing incentives, fit-out and amenity capital, contingency, and any GST that cannot be recovered.

Does land go in at historic cost or at current market value?

Both answers are used, and they answer different questions, so the trap is mixing them within one comparison rather than choosing wrongly.

Land at historic cost measures the actual project: what this entity actually spent, including a site bought cheaply years ago. Land at current market value measures the decision on today’s facts: whether developing this site now beats selling it or doing something else with it. A site bought in 2016 and carried at cost can make a yield on cost look strong while concealing that the same result would follow from simply selling the land.

Consultants and funders frequently assume one convention without saying which, so the question worth asking before comparing anything is which basis the other party’s number is on. A yield on cost built on historic land cost and compared against a market capitalisation rate is not a clean comparison, because the capitalisation rate reflects today’s market and the land does not.

Where does GST sit in an Australian yield on cost?

This is the point at which Australian yield on cost arithmetic diverges from the offshore explainers, and it can be worth more basis points than any other single input.

Where the completed asset will be let as residential premises for residential accommodation, the rent is input taxed. The Australian Taxation Office (ATO) states that if you rent out residential premises for residential accommodation, the rent is input taxed and you do not include GST in the rental charge, and that an entity intending to rent out residential premises is not entitled to input tax credits on the construction and acquisition costs of those premises. The consequence for a yield on cost is direct: the GST embedded in the construction contract, the consultant fees and most of the contingency is a real, permanent cost of the project and belongs in the denominator.

Where the completed asset will be let on a taxable basis, as most commercial, retail and industrial premises are, the GST on construction inputs is generally recoverable, and the total development cost used for a yield on cost should be stated net of it.

The size of the effect is easy to underestimate. In the worked example, the GST embedded in construction ($7,500,000), consultants ($700,000) and contingency ($400,000) totals $8,600,000. If that same $128,600,000 project were on a taxable-supply footing and the GST were recoverable, the total development cost would fall to $120,000,000. On the same $6,300,000 of net operating income (NOI), the yield on cost would be 5.25 per cent rather than 4.90 per cent, and the spread against a 4.50 per cent exit would be 75 basis points rather than 40. The two assets are not otherwise comparable, so treat that as an illustration of the cost-side effect rather than a like-for-like swap. But it does show that a residential rental yield on cost and a commercial yield on cost are not the same measurement, and that comparing them without adjusting is an error of roughly the size of the spread itself.

Which Australian taxes and concessions change the numerator?

Land tax, and whether it can be passed to tenants, is generally the largest single tax line inside operating expenses, and it varies by state in two separate ways: the concession available to the asset, and the recoverability of the tax from the tenant.

How do the state build-to-rent land tax concessions change yield on cost?

Every mainland state has legislated some form of land tax concession for eligible residential rental developments, and because land tax sits inside operating expenses, a concession flows straight through to net operating income (NOI) and therefore to the yield on cost. The regimes are separate from each other and from the federal incentives, and they are not aligned on dwelling counts, dates or duration.

New South Wales. Revenue NSW has published that the Land Tax (Build-to-rent Concessions) Amendment Act 2025 establishes a new ongoing land tax concession scheme for owners of certain new build-to-rent properties while preserving the existing time-limited scheme. Revenue NSW states that the new ongoing scheme has no time limit and still provides for the land value of the parcel to be reduced by 50 per cent where there is a building on the land used and occupied as a build-to-rent property in accordance with the Treasurer’s guidelines. The same Act provides for refunds of surcharge purchaser duty and surcharge land tax for eligible new developments.

Victoria. The State Revenue Office publishes that eligible build-to-rent developments receive a 50 per cent reduction in the taxable value of the land, together with an exemption from the absentee owner surcharge, and that eligibility generally requires at least 50 self-contained dwellings with an occupancy date on or after 1 January 2021 and before 1 January 2032, rented or genuinely offered for rent to the general public under residential rental agreements meeting the required terms, with the eligibility requirements satisfied continuously for at least 15 years from the occupancy date.

Queensland. The Queensland Revenue Office publishes that eligible developments may receive a 50 per cent reduction in the taxable value of land for land tax, a 100 per cent reduction for the land tax foreign surcharge, and a 100 per cent discount on additional foreign acquirer duty. It states that an eligible development must comprise at least 50 dwellings meeting certain requirements, must meet the discounted rent housing requirements including providing at least 10 per cent of dwellings at discounted rent to eligible tenants, and must first become suitable for occupation between 1 July 2023 and 30 June 2030. The concessions run for a maximum of 20 years or until 30 June 2050, whichever comes sooner, and the provisions sit in Part 6A of the Land Tax Act 2010.

One Queensland detail has a direct modelling consequence: the Queensland Revenue Office states that the concessions are not available during construction and become available once the development is operational. A feasibility that applies the concessional land tax rate from site acquisition overstates net operating income (NOI) during the very period when holding costs are heaviest.

Western Australia. The Department of Treasury and Finance publishes that the exemption provides a 50 per cent exemption from land tax for up to 20 years for eligible developments containing at least 40 self-contained dwellings available for three-year residential leases, owned by the same owner or group and managed by the same management entity, and completed between 12 May 2022 and 1 July 2032. Western Australia’s 40-dwelling threshold is the outlier, and a 45-apartment project can qualify there while failing the 50-dwelling tests elsewhere.

South Australia. RevenueSA has published a land tax reduction for eligible build-to-rent properties where construction commenced on or after 1 July 2023, applying as a 50 per cent reduction in land value and running from the 2023-24 financial year to and including 2039-40, with the detailed eligibility criteria dealt with in regulations. The RevenueSA page is where to confirm the criteria as they currently stand.

Tasmania, the Australian Capital Territory and the Northern Territory. The concessions above are state regimes, and the land tax base also differs between Tasmania, the Australian Capital Territory and the Northern Territory. For a project in any of the three, the relevant revenue office’s current published position is the only reliable source for that operating expense line, and modelling a mainland-style 50 per cent concession into the numerator without confirming it would overstate net operating income (NOI) from the first stabilised year onwards.

Back to the worked example to show what this is worth. Suppose the $3,146,400 of operating expenses includes land tax of $260,000. If the development qualifies for a 50 per cent reduction in taxable land value, that line falls to $130,000, operating expenses fall to $3,016,400, and stabilised net operating income (NOI) rises to $6,430,000. The yield on cost goes from 4.90 per cent to 5.00 per cent, and the spread against a 4.50 per cent exit capitalisation rate widens from 40 basis points to 50. Ten basis points, a quarter of the project’s original spread, came out of one line in the operating expense schedule, and that line is conditional on eligibility criteria that have to be met continuously for years after handover.

Can land tax be recovered from the tenant?

This is a second, separate state variation, and it decides whether land tax sits inside net operating income (NOI) at all.

In Victoria, section 50 of the Retail Leases Act 2003 (Vic) renders void a provision of a retail premises lease to the extent that it makes the tenant liable to pay an amount for tax for which the landlord or head landlord is liable under the Land Tax Act 2005 (Vic) or the Commercial and Industrial Property Tax Reform Act 2024 (Vic). For premises caught by that Act, the land tax is the owner’s and it reduces net operating income (NOI) dollar for dollar. Whether particular premises are retail premises for the purposes of the Act is itself a legal question and not always obvious from the use.

Other jurisdictions take different positions on recoverability, and the position also differs between retail leases legislation and ordinary commercial or industrial leases that fall outside it. The consequence for a yield on cost is that two otherwise identical buildings in two states can carry materially different net operating income (NOI) on identical rent, purely because of who bears the land tax. Anyone building a yield on cost on a retail or mixed-use asset should treat recoverability as a question for the lawyer reviewing the lease precedent, not as a modelling assumption.

How do the federal build-to-rent tax incentives affect the return?

The federal incentives do not change net operating income (NOI), because net operating income (NOI) is measured before tax. They change what the owner keeps, and they carry a clawback that a yield on cost does not show at all.

The ATO states that the build-to-rent development tax incentives give owners and investors in eligible developments access to an accelerated deduction of 4 per cent for capital works and a concessional final withholding tax rate of 15 per cent on eligible fund payments from a managed investment trust to a foreign resident of an information exchange country. The 4 per cent rate shortens the write-off period for construction expenditure from 40 years to 25, against the ordinary capital works deduction rates that apply to a build-to-hold asset.

The eligibility criteria the ATO publishes are exacting, and each one is an operating constraint for years after the building is finished:

• The development consists of 50 or more residential dwellings made available for rent to the general public
• The dwellings are residential premises, taxable Australian real property, and not commercial residential premises
• The dwellings and common areas continue to be owned by a single entity for at least 15 years, though the development can be sold to another single entity during that period and remain eligible
• Dwellings are tenanted or made available to be tenanted by way of lease for 5 years or more, with the ATO noting that this requirement does not apply where a tenant requests a shorter term provided the owner offered at least 5 years
• At least 10 per cent of the dwellings are available as affordable dwellings, and the number of comparable non-affordable dwellings is greater than or equal to the number of comparable affordable dwellings

To access the accelerated deduction, the ATO states that construction must have commenced after 7:30 pm Australian Eastern Daylight Time on 9 May 2023, and that the owner must notify their choice to opt in by lodging the approved form. Notification of a relevant event must be lodged on or before 28 days after the event, and the ATO states that the Commissioner does not have the discretion to extend that 28-day period.

The affordability standards have already moved once. The ATO states that the initial legislative instrument applied to assessing events from 1 January 2025 through to 26 March 2026, and that an amended instrument applies from 27 March 2026, requiring among other things that a dwelling be either a moderate or lower-income dwelling, that at least 2 per cent of dwellings be lower-income dwellings, and that where dwellings are tenanted the tenant was identified by an eligible community housing provider engaged by the owner.

Then there is the misuse tax. The ATO states that where an active build-to-rent development ceases to be one during the 15-year compliance period, the owner is liable for a misuse tax equal to the total of the capital works deduction amount and the build-to-rent withholding amount, and that the capital works component is the accelerated deduction claimed plus 8 per cent of that amount, with the withholding component similarly multiplied by 1.08. The ATO also states that a build-to-rent owner cannot claim a deduction for misuse tax paid.

The structural point for a yield on cost is this. The concessions improve after-tax returns and can be part of what makes a thin pre-tax spread acceptable to an equity investor, but they are conditional on operating decisions stretching fifteen years past practical completion, and failing them reverses the benefit with an uplift. A spread that only works because of the concessions is a spread with a fifteen-year operating covenant attached, and that is a materially different proposition from a spread that works on the rent alone. Whether any of this applies on your facts is a question for your accountant, because eligibility turns on the ownership entity, the trust’s status and the dwelling mix.

How sensitive is the development spread, and which inputs move it most?

The spread is more sensitive to the exit capitalisation rate than to almost anything else, and the exit rate is the input the developer has least control over. That asymmetry is the argument for treating the spread as a range rather than a point.

In sensitivity analysis, three variables deserve to be flexed together rather than one at a time, because they tend to move together in practice:

The exit capitalisation rate. Capitalisation rates respond to the cost of capital, and the cost of capital responds to the cash rate, which the Reserve Bank of Australia publishes and updates after each Monetary Policy Board meeting on its cash rate page. A project underwritten at the tightest capitalisation rate in recent evidence is a project that depends on the market not moving during the years it takes to build.

Construction cost. Covered above, and worth flexing at both a modest and an unpleasant level, because the second scenario in the worked example showed a 5 per cent movement taking close to 40 per cent of the value created.

Stabilised rent and lease-up duration. These move in the same direction as each other and often in the opposite direction to the developer’s assumption. A market soft enough to push the exit capitalisation rate out is usually soft enough to slow leasing and soften rents, so testing a capitalisation rate expansion while holding rent and lease-up constant understates the downside. The scenario worth running is all three moving adversely at once, because that is the state of the world in which they usually move.

The output that answers the question is not a single spread. It is the exit capitalisation rate at which the spread reaches zero, how far that is from today’s evidence, and how much of the gap a plausible cost movement consumes.

How does yield on cost work in New Zealand?

The arithmetic is identical in New Zealand: stabilised net operating income (NOI) divided by total project cost, compared against the capitalisation rate a completed asset would be valued on, with the same break-even identity when the two meet.

Two differences are worth building into a New Zealand feasibility rather than carrying an Australian assumption across.

The Goods and Services Tax position on residential rental is similar in effect. Inland Revenue publishes that long-term residential rental is an exempt supply, that GST cannot be charged on the rent for a residential dwelling, and that a landlord cannot claim GST on dwelling expenses. As in Australia, that makes the tax on construction inputs a permanent cost sitting in the denominator of a residential rental yield on cost, and it means a New Zealand commercial yield on cost and a New Zealand residential rental yield on cost are measured on different bases.

The holding cost structure is different. The state-by-state land tax concessions described above are creatures of Australian state legislation and have no New Zealand counterpart, so the operating expense schedule behind a New Zealand yield on cost is built from a different set of lines, with local authority rates and insurance doing much of the work that land tax does in an Australian model. Carrying an Australian assumption of a 50 per cent land tax concession into a New Zealand numerator would overstate it.

Market capitalisation rates differ from Australian levels and move on their own cycle, so a New Zealand exit rate should come from local sales evidence and a registered valuer rather than from an Australian comparable.

Where does yield on cost mislead developers?

Yield on cost misleads most often when the numerator and the denominator are not on the same basis, and the errors are consistent enough to list.

The denominator is incomplete. Land left out, or finance costs left out, or the lease-up period’s holding costs left out. Each omission lifts the yield on cost and the spread with it, and each is invisible in the headline percentage. The question to ask of any yield on cost handed to you is what is in the cost, not what the percentage is.

Day-one income used as stabilised income. A building that will stabilise in two years is not producing stabilised net operating income (NOI) on the day the occupation certificate issues, and treating the two as the same overstates the numerator while understating the denominator at the same time.

Face rent capitalised instead of effective rent. Incentives are a real cost and the market looks through them. A yield on cost built on face rent will not survive the valuation that prices the exit.

GST treated as recoverable when it is not. In the worked example above, the difference between the two treatments was 35 basis points of yield on cost, which on a 40 basis point spread is most of the deal.

Land at historic cost compared against a current market capitalisation rate. A stale land number against a live market rate is two different points in time in one ratio.

A geared return compared against an ungeared one. Yield on cost is ungeared by construction. Comparing it against a geared equity return, or comparing a geared equity return against a capitalisation rate, produces a spread that means nothing.

The exit capitalisation rate treated as fixed. The single most expensive assumption on a build-to-hold project, and the one the sensitivity table exists to test.

Concessions assumed rather than confirmed. Each state regime has dwelling counts, date windows, affordability requirements and continuing compliance obligations. A yield on cost carrying a concession the asset will not qualify for is simply wrong, and it will be wrong in the direction that makes the project look viable.

The spread quoted without the programme. A 200 basis point spread earned over six years and a 200 basis point spread earned over two years are different investments described by the same number.

What to ask your valuer

These are the questions that decide the exit half of the spread, and they are not questions a guide can answer on your facts.

• What exit capitalisation rate would you adopt for this asset on completion, and what specific sales evidence supports it? How old is that evidence and what has moved since?
• What range would you consider plausible for that rate over the delivery programme, rather than a point estimate at today’s date?
• Are you valuing on effective rent or face rent, and what incentive level are you assuming for the lease-up?
• What lease-up period and stabilised occupancy would you adopt for this asset type in this submarket, and what letting-up allowance does that imply?
• What capital expenditure or replacement reserve would a purchaser deduct when pricing this asset, and how would that affect the value they would pay?
• If the asset is held rather than sold, what would a financier’s valuation for refinancing purposes likely differ on compared with a market value for sale?

What to ask your quantity surveyor

• What is the current cost plan, and what escalation allowance does it carry to the expected contract execution date and through the construction period?
• Which cost lines carry the most estimating uncertainty at this design stage, and what is the realistic range on each?
• Is the contingency sized for the design stage we are actually at, and what is it intended to cover?
• Which authority, headworks and developer contribution amounts are confirmed and which are estimates?
• What are the cost consequences of the design features the build-to-rent concessions or the affordable dwelling requirements impose?

What to ask your accountant

• Will the completed asset’s rental income be input taxed or taxable, and what GST is recoverable on construction, consultants and fit-out on that basis?
• What GST amount should sit permanently in the total development cost, and when does it need to be funded?
• How will land tax be assessed on the completed asset in this state, on what basis, and what is the estimate for the stabilised year?
• Does the asset and our ownership structure meet the eligibility criteria for the state land tax concession, and what are the continuing obligations if it does?
• Do we meet the federal build-to-rent eligibility criteria on our facts, including the single-entity ownership, the 15-year period, the 5-year lease offer and the affordable dwelling requirements, and what are the notification deadlines?
• What is the misuse tax exposure if the development ceases to be an active build-to-rent development, and at what points in the next 15 years is that most likely to arise?
• On what basis should the capital works deduction be modelled, and how does that flow through to the after-tax hold return?

What to ask your lawyer

• Are these premises retail premises for the purposes of the relevant retail leases legislation, and can land tax and outgoings be recovered from tenants under the lease precedent we are using?
• What operating restrictions do the state and federal build-to-rent regimes impose on the lease documents, and do our standard terms comply?
• What happens to the concessions on a sale of the asset, and what does a purchaser need to do to preserve them?
• Do any planning agreements, affordable housing obligations or Section 173 style agreements on the site affect the income the asset can produce?

The short version

Yield on cost is the stabilised net operating income (NOI) over the total cost of creating the asset, and the development spread is that yield less the capitalisation rate the market will value the finished building on. The spread is what a developer is paid for taking risk an investor could have avoided by buying something already built.

Three things decide whether a spread is real. The denominator has to be complete, including land, finance, lease-up and any GST that cannot be recovered, because in the Australian market that last item alone can move a residential rental yield on cost by more than a third of a percentage point. The numerator has to be stabilised, effective and net of the operating costs the owner genuinely bears, which in turn depends on state land tax concessions the asset has to keep qualifying for and on whether the tenant can be made to carry the tax at all. And the exit capitalisation rate has to be treated as a range, because the arithmetic guarantees that when it rises to meet the yield on cost, the value created is exactly nil.

A developer who knows their yield on cost already knows the exact market rate at which their project stops being worth building. The useful question is how far away that rate is, and how much of the distance a plausible round of construction cost escalation would close.