S-Curve Construction Cost Phasing for Australian Developers

An S-curve is the shape your construction spend makes when you plot it cumulatively over the build programme. It starts slowly, steepens through the middle, then flattens out as the job finishes. Draw it and you get something close to a stretched letter S, which is where the name comes from. In a feasibility, it is the alternative to spreading the construction budget evenly across the build months.

Two people hold the inputs. The quantity surveyor’s cost plan is where the shape of the curve comes from for a specific job, and the payment terms in the building contract are what decide when money actually leaves the account, which is a different question and often the more expensive one. This guide is about what to do with both, and what changes when the shape or the lag is wrong.

Figures, thresholds and legislated payment periods in this guide were current at the date of writing and they move. The security of payment rules in particular have changed materially in several jurisdictions in recent years. Each linked primary source is where to confirm the current position before you rely on it.

What is an S-curve in construction cost phasing?

An S-curve is a cumulative distribution of construction cost across the build programme. Instead of assuming an equal amount is spent each month, you assume a low spend rate at the start, a high spend rate through the middle, and a low rate again at the end.

The logic is physical, not financial. In the opening months you are establishing the site, running demolition, doing bulk earthworks and getting a small number of trades on the ground. Very few people are being paid. By the middle of the job you may have structure, services rough-in, facade and early fit-out running concurrently, with the largest number of subcontractors on site at once. That is the steep part. At the end you are down to defects, commissioning, landscaping and cleaning, with a handful of trades finishing off. The rate of spend falls away again.

Plotted month by month, the periodic spend looks like a bell. Plotted cumulatively, it looks like an S. They are the same data.

The word “S-curve” is also used on site for progress reporting, where planned value is plotted against earned value and actual cost. That is the same shape used for a different purpose. This guide deals with the feasibility and cashflow use: how much money you need, and when.

Why does the shape of the cost curve matter to a feasibility?

Because a feasibility that gets the total right and the timing wrong will still misprice the deal. Three things move when you change the shape of the curve, and they do not all move in the same direction.

Peak monthly funding requirement. This is the biggest and most reliably overlooked effect. A construction budget spread evenly needs the same amount every month. The same budget on a realistic S-curve needs substantially more in the busiest months. If your facility has a monthly drawdown cap, or your equity is being contributed in tranches, the straight-line version will not show you the month where you run short.

When you cross funding thresholds. Construction loan drawdowns are usually released against certified progress claims and are constrained by a facility limit and by cost-to-complete tests. The month you cross a given cumulative spend figure is different under an even spread than under a curve, which matters if you have staged equity, a limited pre-sale release, or a facility that steps up.

Interest. Interest on a construction facility accrues on the drawn balance, so the shape of drawdown drives the interest bill. This is the effect most commonly claimed and, as the worked example below shows, the one most commonly overstated.

There is a fourth, quieter effect. Everything you phase off the back of construction inherits the shape. Construction contingency is usually distributed with the works it protects. Superintendent and project management fees often follow site activity. Get the curve wrong and a chunk of your soft costs land in the wrong months too.

Does an S-curve actually change your interest bill?

Less than most people expect, and sometimes not at all. This is worth understanding properly because a lot of published material implies that switching from a straight line to an S-curve is primarily an interest accuracy exercise. It usually is not.

Interest on a facility that draws progressively is driven by the average drawn balance over the term. A symmetric S-curve, one where the spend profile is a mirror image around the midpoint, has almost exactly the same average cumulative balance as a straight line. The extra money you avoid drawing in the early months is offset by the extra money you draw in the middle. The average is unchanged, so the interest is unchanged.

The interest saving only appears when the curve is asymmetric, specifically when it is back-weighted so that cumulative spend sits below the straight line for most of the programme. Many real construction curves are mildly back-weighted, because the finishing trades in the second half of a job carry a lot of value and the early enabling works carry relatively little. But a job with a deep basement, significant piling, or a large early procurement commitment can be front-weighted instead, in which case the S-curve produces more interest than the straight line, not less.

The honest summary: model the S-curve for the peak funding number and the drawdown timing. Treat the interest movement as a secondary output that could go either way depending on the shape.

Worked example: $18 million over 18 months

Take a construction contract of $18,000,000 across an 18 month programme. Interest is illustrated at an indicative 9.5 per cent per annum, which is used here only to make the arithmetic visible. Current construction facility pricing is a matter for a term sheet, not a guide. Interest is calculated monthly on the closing drawn balance, which is a simple convention and slightly conservative.

Three profiles for the same $18,000,000:

Two things fall out of that table.

The symmetric S-curve produces exactly the same interest as the straight line. Not approximately: the same, to the dollar, because the average balance is identical. What it does change is the peak monthly draw, which rises from $1,000,000 to $1,493,827, roughly 49 per cent higher. That is the number that breaks a funding plan.

The back-weighted curve saves $167,279 of interest against the straight line, about 12 per cent of the construction facility interest. It also pushes the peak monthly draw to $1,647,986, about 65 per cent above the straight-line figure.

For reference, the monthly percentages behind those two curves, rounded to one decimal place:

Note the crossover in the funding test. If your facility limit for the works component were $12,000,000, the straight-line model shows you reaching it at the end of month 12 with $12,000,000 drawn. The symmetric S-curve shows you at $13,338,000 by the same point, which is over the limit. Same total cost, same programme, different month you have a problem.

What the same deal does to margin

Put those two interest outcomes into a small feasibility so the effect on margin is visible. All figures are illustrative.

Every input other than the interest line is held constant between the two scenarios. The development margin moves 0.7 percentage points on cost. On a thin deal that is not nothing, but it is not the reason to build the curve. The reason to build the curve is the $647,986 gap between what the straight-line model says you need in your busiest month and what you actually need.

What shape does a real Australian construction programme follow?

It depends heavily on the building type, and the honest answer is that the curve should come out of the construction programme and the cost plan rather than out of a formula. Still, some patterns tend to hold.

The four phases that bend the curve

Site establishment and early works. Site sheds, hoarding, demolition, remediation, bulk excavation, shoring and piling. On a slab-on-ground townhouse job this is a small slice. On a tower with two basement levels it can be a large and front-loaded slice, and it can flatten the early part of the curve considerably less than the textbook shape suggests.

Substructure and structure. Footings, basement, in-situ or precast frame, slabs. Spend rate climbs steeply here and stays high. Precast introduces a wrinkle: panels are often paid for on delivery or on manufacture, not on erection, which can pull spend forward of the programme.

Facade, services and early fit-out. Typically the peak. The largest number of trades on site simultaneously, and the highest value trades running concurrently. Long-lead items such as lifts, switchboards and curtain wall may involve deposits and milestone payments well before installation, which again shifts money earlier than the physical programme implies.

Completion and handover. Final fit-out, commissioning, landscaping, external works, defects rectification, cleaning. Spend rate falls but does not go to zero, and the tail can be longer than the programme suggests because of the retention and defects arrangements dealt with below.

Where townhouses and apartments differ

Programme length is the first variable, and it varies more by product than by state. The Australian Bureau of Statistics has published national average completion times showing new townhouses at 3.36 quarters, roughly ten months, and new flats, units or apartments at 6.66 quarters, close to 20 months, in 2018-19, against 2.22 quarters for detached houses. Those figures come from the ABS analysis of average dwelling completion times, sourced from Building Activity, Australia. They are dated and predate the cost and labour conditions of recent years, so treat them as an indication of the relative gap between product types rather than as a current benchmark. The current release of Building Activity, Australia is where to check the latest data cube.

The practical consequence for the curve: a shorter programme has a flatter, less pronounced S because there is less room for a ramp-up and ramp-down. A ten month townhouse job spread on a textbook symmetric curve can look wrong, because two of those ten months are effectively site establishment and two are effectively handover. On short programmes it is often more honest to build the profile from the programme stages directly than to fit a mathematical curve to it.

How do you build an S-curve in a cashflow model?

The three-input method

The most common approach needs only three inputs per cost line: the total amount, the start month and the duration. A curve function then distributes the amount across those months. It is quick, it is transparent, and it is good enough for early-stage appraisal work when you do not yet have a programme.

The trap with the three-input method is that it invites you to apply the same shape to every cost line. Construction and contingency belong on a curve. Land, acquisition costs, statutory contributions and settlement-linked items generally do not.

Choosing a curve

There is no standard S-curve. Several formulations are used, and the differences between them are smaller than the difference between any of them and a straight line.

The simplest is the smoothstep, where the cumulative fraction spent at time t (with t running from 0 to 1 across the programme) is:

C(t) = t² × (3 − 2t)

That produces the symmetric curve in the worked example above. It is easy to implement in a spreadsheet, easy for a financier to follow, and symmetric, which as shown means it will not move your interest number.

To skew the curve, a beta distribution cumulative function is the usual tool. Two shape parameters let you push the peak earlier or later. The back-weighted curve above uses shape parameters of 2.6 and 2.0, which moves the peak spend month past the midpoint. A Gompertz or shifted logistic function does something similar with one parameter.

For most feasibility work, a pragmatic alternative is the thirds method: allocate a nominated share of the budget to the first third, middle third and final third of the programme, then spread evenly within each third. Something like 20 per cent, 50 per cent, 30 per cent gives a recognisable S shape with no mathematics at all, and it is easy to override when the quantity surveyor tells you the middle third is heavier than that.

Front-loaded and back-loaded variants

Some situations argue for pushing the curve deliberately in one direction.

A job with substantial early procurement, precast off site, imported facade, or a large lift and vertical transportation package, tends to be more front-loaded than the shape of the physical programme suggests, because money leaves against manufacture and delivery rather than installation.

A job with a heavy external works, landscaping and civil component tends to be more back-loaded, because those packages run late and can be significant on a townhouse or land subdivision product.

A job under a guaranteed maximum price contract or a design and construct arrangement may have a payment schedule negotiated to a milestone table rather than to measured progress, in which case the contract schedule is the curve and there is nothing to model.

What not to put on the curve

Costs that are event-driven rather than progress-driven should not be smeared across a curve. Land settlement, transfer duty, developer contributions payable at a defined stage, insurance premiums paid up front, connection and headworks charges billed on application, and sales commissions payable at settlement all land in a specific month. Putting them on a curve produces a cashflow that looks smooth and is wrong.

The general test: if you can name the month it gets paid, name the month. Curves are for costs whose timing you can only estimate.

When does the money actually leave your account?

This is where a lot of otherwise careful cashflow models come unstuck. The S-curve models when work is done. Money leaves when a payment claim is made, assessed and paid, and the gap between those two things is set by your contract and by the security of payment legislation in your jurisdiction. On a monthly claim cycle the lag is typically several weeks, which shifts your entire cost curve to the right relative to the programme.

Every Australian state and territory has security of payment legislation, and the detail varies. The following is the current position in the larger jurisdictions.

New South Wales

Under the Building and Construction Industry Security of Payment Act 1999 (NSW), a payment claim can generally be served from the last day of each month in which work was carried out, unless the contract provides for an earlier date. Building Commission NSW summarises the position on its making a payment claim page: contracts may specify when payment claims can be made, and if no dates are stated, it is the last day of the month.

Section 11 of the Act sets the due date. A progress payment from a principal to a head contractor becomes due and payable on “the date occurring 15 business days after a payment claim is made”, or an earlier date if the contract provides one. For a subcontractor the equivalent figure is 20 business days, and for an exempt residential construction contract it is 10 business days where the contract is silent. Subsection 11(8) states that a contract provision “has no effect to the extent it allows for payment of a progress payment later than the relevant date”. The full text is at section 11 of the Act.

The respondent also has up to 10 business days to serve a payment schedule if they are not paying the full claim. For a developer modelling cash, the practical effect is that work done in month 6 is claimed at the end of month 6 and paid roughly three weeks into month 7.

Victoria

Victoria’s regime changed substantially on 15 April 2026, when the Building Legislation Amendment (Fairer Payments on Jobsites and Other Matters) Act 2025 (Vic) amended the Building and Construction Industry Security of Payment Act 2002 (Vic). The Building and Plumbing Commission sets out the detail on its changes affecting construction contracts page.

Three of those changes matter to a cashflow. First, a contract provision has no effect to the extent it provides for payment later than 20 business days after a payment claim is served. Second, a contract cannot require a payment claim to be served later than the last day of each named month, and cannot require milestone payment claims to be served less frequently than monthly. Third, the amendments apply retrospectively to contracts entered into before 15 April 2026, subject to transitional provisions.

The Commission also notes that a business day in Victoria now excludes the period from 22 December to 10 January, and that a contract may provide that a payment claim for work carried out between 22 December and 31 December is not served until 31 January the following year. If you are modelling a build that runs through a Christmas shutdown, that is a real gap in your December and January cost rows. The current in-force version of the Act is on legislation.vic.gov.au.

Queensland

Queensland runs progress payments, payment disputes and the trust account framework through the Building Industry Fairness (Security of Payment) Act 2017. Business Queensland summarises the framework on its payments in the building industry page, and the Act itself is on legislation.qld.gov.au.

The Queensland feature that most affects a developer’s cash timing is the trust account framework. Where a contract is eligible, the contracted party, generally the head contractor, must operate one project trust account for each eligible contract and one retention trust account holding cash retentions across their eligible contracts. Contract payments are made into the account and subcontractors are paid out of it. Whether a given contract is eligible, and when, depends on criteria and a phased commencement that the Queensland Building and Construction Commission publishes on its trust accounts page. Because those criteria have been staged in over several years, confirm the current position there before assuming a contract is in or out.

Western Australia

Western Australia moved from the Construction Contracts Act 2004 to the Building and Construction Industry (Security of Payment) Act 2021 (WA). The Western Australian Government’s security of payment guide states that the 2021 Act applies to contracts entered into on or after 1 August 2022, and that contracts entered into before that date remain under the earlier legislation. The Act sets maximum payment periods and introduces a statutory right to progress payments; the current text is on legislation.wa.gov.au. Because Western Australia’s maximum payment periods could not be verified against a government source at the time of writing, the specific number is not stated here. It is one to confirm from the Act before you set the lag in your model.

South Australia, Tasmania, the Australian Capital Territory and the Northern Territory

Each has its own regime. South Australia and Tasmania operate under a Building and Construction Industry Security of Payment Act 2009, the Australian Capital Territory under the Building and Construction Industry (Security of Payment) Act 2009, and the Northern Territory under construction contracts legislation dating from 2004 that follows a different model to the eastern states. The Tasmanian Act is on legislation.tas.gov.au. Default payment periods, claim windows and adjudication timeframes differ in each, so the lag you build into a cashflow for a Hobart job is not necessarily the lag for a Sydney one. Confirm the position from the relevant jurisdiction’s legislation before setting it.

New Zealand

New Zealand’s Construction Contracts Act 2002 runs a payment claim and payment schedule regime with default timeframes that apply where the contract does not deal with the matter. The Ministry of Business, Innovation and Employment maintains guidance on the Act at building.govt.nz.

The New Zealand retention regime is stricter than most Australian equivalents and directly affects the tail of the curve. Retention money is trust property and must be held on trust, kept separate from other money, and held in a bank account with prescribed ledger accounts. The party withholding it must provide quarterly reports to each subcontractor, a report after each transaction, and pay interest on late payments. The requirements apply even where the contract says otherwise. Those rules were tightened again by the Construction Contracts (Retention Money) Amendment Act 2023. The Ministry sets out the detail on its retention money requirements page.

How do retention and the defects liability period change the tail?

Retention is the reason the last few per cent of your construction budget can sit unpaid long after practical completion, and the reason a naive curve that finishes at 100 per cent in the final construction month overstates late-stage cash outflow.

In New South Wales, retention money held by head contractors on projects valued over $20 million must be held in a trust account with an authorised deposit-taking institution, under the Building and Construction Industry Security of Payment Regulation 2020. Building Commission NSW notes on its retention money page that usually up to 5 per cent of the total contract value is held until the subcontractor has completed the job and fixed any defective work, that the $20 million threshold is assessed on the head contract, and that head contractors who fail to comply face fines of up to $22,000.

That regime governs the relationship between the head contractor and its subcontractors rather than between you and the head contractor. Your own retention position sits in your building contract. Typical Australian building contracts provide for a percentage of each progress payment to be retained, with a portion released at practical completion and the balance at the end of the defects liability period, commonly some months later. Because those figures are contractual rather than statutory, they vary and there is no primary source to quote for a market norm. Read the contract.

For modelling, the effect is a small deferred lump. If your contract retains 5 per cent capped at 2.5 per cent of the contract sum, with half released at practical completion and half at the end of a 12 month defects liability period, then a meaningful slice of the construction budget does not leave your account in the construction window at all. On the $18,000,000 contract in the worked example, a 2.5 per cent cap would be $450,000, with $225,000 released at practical completion and $225,000 a year later.

Two consequences follow. Your modelled peak debt is slightly lower than a curve that pays 100 per cent within the build window would suggest. And you have a cash obligation sitting past your last settlement, which is a problem if your facility has been repaid and the entity has been wound down. Neither is large, but both are the kind of thing that surfaces at the wrong moment.

Does GST timing move the curve?

Yes, and it is a separate curve. GST on progress claims is paid out with each claim and recovered through your Business Activity Statement, so there is a timing gap between paying it and getting it back. On an $18,000,000 contract that is $1,800,000 cycling through the model.

Division 156 of the A New Tax System (Goods and Services Tax) Act 1999 deals with supplies and acquisitions made on a progressive or periodic basis, treating each periodic component as a separate supply for attribution purposes. The section text is at section 156-5. How that interacts with your reporting cycle, your accounting basis and the treatment of retention amounts is an accountant’s question, and it is a real one: the difference between monthly and quarterly reporting can be a meaningful working capital swing on a large build.

For a first-pass model, the workable approach is to run the construction curve exclusive of GST for feasibility purposes, and run a separate GST cash line that reflects your actual reporting cycle. Blending them produces a cashflow that is neither.

How does the S-curve interact with your construction facility?

Three connections are worth being deliberate about.

Peak debt. Peak debt is a function of the cost curve, the revenue curve and the order in which equity and debt are drawn. Because the S-curve concentrates spend in the middle months, it usually brings the peak debt month forward relative to a straight line, and raises the monthly funding requirement around it. If a financier is sizing a facility off your cashflow, the curve is what they are reading.

Capitalised interest. Where interest is capitalised rather than serviced, it accrues into the facility and reduces available drawdown, which means the interest calculation and the drawdown calculation feed each other. A curve that draws more in the middle capitalises more interest in the middle, which consumes headroom in exactly the months you need it. A straight-line model hides that circularity because it never produces a spike.

The average drawn balance shortcut. A commonly used shortcut assumes a fixed proportion of the facility is outstanding on average across the term, often quoted around 55 per cent, and applies the rate to that. It is a reasonable approximation for a quick appraisal. The point of building the curve is to test whether that shortcut holds for your programme. In the worked example above the average drawn balance was 52.8 per cent of the construction budget on the straight line and symmetric curve, and 46.3 per cent on the back-weighted curve. Close enough to the shortcut in one case, materially different in the other.

What breaks an S-curve assumption on site?

The curve is a forecast, and the forecast tends to fail in specific and fairly predictable ways.

Programme slippage. A job that runs two months long does not simply stretch the curve. Work that was going to be claimed in month 16 gets claimed in month 18, but preliminaries and site overheads keep accruing at close to their monthly rate throughout, and interest keeps accruing on the drawn balance for two extra months. Slippage costs more than the stretched curve suggests.

Variations. Variations are typically approved and claimed in the second half of a job, which back-weights the curve in a way the original profile did not anticipate. If your contingency is sitting on the same S-curve as the base contract, it is being spent earlier in your model than it will be in reality.

Front-loaded claim schedules. Some contractors negotiate a payment schedule weighted towards the early stages, either through preliminaries recovery or a milestone table. Where that has been agreed, the money curve is not the works curve, and the contract schedule wins. A construction lawyer reading the payment schedule will spot this faster than a quantity surveyor reading the cost plan.

Offsite manufacture and long-lead deposits. Precast, modular bathrooms, facade systems, lifts and switchboards may be paid on manufacture or delivery. Vendor finance and offsite materials provisions in the contract determine whether you are funding these before they arrive on site. If you are, the curve moves left.

Shutdowns. The Christmas and New Year period stops most Australian sites for two to four weeks and slows the claims that follow it. Victoria now formally excludes 22 December to 10 January from the definition of a business day in its security of payment regime. A monthly model that spreads costs evenly through December and January is overstating both months.

How do you sense-check the curve you have been given?

Four checks, none of which take long.

Against the programme. Line the cumulative curve up against the construction programme. The steepest part of the curve should coincide with the period where the most trades are on site. If your curve peaks in month 5 of an 18 month job and the programme shows structure not topping out until month 10, one of the two is wrong.

Against the trade breakdown. Take the five largest trade packages by value and check when each is programmed. If structure, facade and services together are 55 per cent of the contract sum and they all run in months 6 to 13, your curve should be putting roughly that share of the money in that window.

Against claim history, once you have some. After three or four progress claims you have real data. Plot actual cumulative spend against your modelled curve. A persistent gap in either direction in the first quarter of a job usually widens rather than closes.

Against the total. The monthly percentages must sum to 100. This sounds trivial and it is the single most common error in a hand-built curve, particularly after someone has adjusted a few months without rebalancing the rest.

If you are building this in a development cashflow model, it is worth putting the sum check on the face of the sheet rather than leaving it to be discovered.

What to ask your quantity surveyor and your construction lawyer

Ask your quantity surveyor:

• What spend profile would you expect for this building type, on this site, over this programme, and how does it differ from a generic S-curve?
• Which trade packages are heaviest and in which months do they run, so the peak of the curve can be tied to the programme rather than assumed?
• Are there packages likely to be paid on manufacture or delivery rather than installation, and how much value sits in them?
• How front-loaded are the preliminaries in the contractor’s proposed payment schedule, and how does that compare to the works curve?
• Where would you expect variations and contingency to be claimed across the programme, and should they be phased differently to the base contract?
• If the programme extends by two or three months, which cost lines keep accruing at their full monthly rate and which simply move?
• What proportion of the contract sum would you expect to still be outstanding at practical completion once retention is taken into account?

Ask your construction lawyer:

• Under this contract, what is the reference date or claim date for progress claims, and what is the due date for payment?
• Does the contract attempt to set a payment period longer than the maximum permitted in this jurisdiction, and if so what is the effect?
• What is the retention percentage, the cap, and the release mechanism at practical completion and at the end of the defects liability period?
• Does this contract or project trigger a statutory trust account or retention trust obligation in this jurisdiction, and whose obligation is it?
• Are there milestone or advance payment provisions that move money ahead of physical work, and what security do we hold if the contractor fails?
• What happens to the payment cycle over the Christmas shutdown under this contract and under the applicable security of payment legislation?
• If a payment claim is disputed and goes to adjudication, what is the realistic cash timing and what is our exposure while it runs?

Ask your accountant:

• On our reporting cycle and accounting basis, what is the actual cash gap between paying GST on a progress claim and recovering it?
• How should retention amounts be treated for GST attribution, and does that change the timing?

The short version

The S-curve is a better description of construction spend than a straight line, and it belongs in any cashflow you are going to show a financier. But the reason to build it is not primarily the interest number. A symmetric S-curve produces the same interest as a straight line, and a skewed one moves it by a few per cent in either direction. What the curve genuinely changes is the peak monthly funding requirement, the month you cross a facility limit, and the honesty of your equity call schedule.

Beyond that, the curve models when work is done, not when money leaves. The gap between the two is set by your contract and by the security of payment legislation in your jurisdiction, and it has been moving. Getting the shape right and the lag wrong will still put your cash in the wrong month.