Building Insurance Valuations: Your Sum Insured Is for an Event That Almost Never Happens

The sum insured on your building policy is a total-loss number. It answers the question “what if the whole thing is gone?” — and the professional guidance for how it is calculated says that event is “extremely rare. Most claims involve a partial loss.”

That is the part almost nobody plans for. If your cover is short, the shortfall does not politely wait for a catastrophe. It shows up on the ordinary claim — the fire in one room, the storm that takes half a roof — because a reduction for underinsurance applies to partial claims too, not only total losses. This guide walks through how a building insurance valuation is actually put together, which costing basis your policy is using, and the specific traps that quietly leave Australian owners short.

First: it is not a valuation

The Australian professional guidance for this work is ANZVGP 104 Insurance Cost Estimates, issued by the Australian Property Institute jointly with the New Zealand bodies, effective 1 July 2025. Its opening definition is worth reading closely, because the language matters:

“an insurance cost estimate is the result of a calculation by a Member of the cost of replacing, or reproducing, the tangible asset as at the relevant/assessment date.”

A calculation of cost. Not an opinion of market value. The insurer is asking a genuinely different question from a buyer, a lender or the tax office, which is why the answer sits so far from the numbers those parties use — and why “what is my property worth” is the wrong question to bring to an insurance renewal.

Which costing basis is your policy actually using?

This is the single most consequential thing most owners do not know about their own cover. ANZVGP 104 sets out distinct costing bases, and two of them produce materially different numbers for the same building.

Reinstatement cost is capped. The guidance describes it as the cost to reinstate the asset at the same location to “a condition equal to, but not better or more extensive than its condition when new”, notionally assuming a total loss.

Replacement cost is forward-looking. It is the cost of an equivalent asset at the same location providing similar function and utility, “but which is of a current design and constructed or manufactured using current materials and techniques”.

Read those two side by side and the gap becomes obvious. Reinstatement points backwards to the building as it was when new. Replacement points at what you would build today, with today’s materials, to today’s methods. For a 1970s brick veneer house, those are not the same construction and they are not the same price.

Indemnity value is a third thing again, and lower than both: restoration to substantially the condition immediately before the loss, expressly taking into account the age, condition and remaining useful life of the asset, with depreciation applied. An indemnity policy on an older building can settle for a fraction of what rebuilding actually costs.

If you do not know which basis your policy is written on, that is the first thing to establish — before the sum insured figure is even discussed.

The partial-loss mechanism, which is where the damage happens

ANZVGP 104 is direct about the mismatch between how the number is calculated and how claims actually arrive:

“the insurance cost estimate provided by Members is for the total cost of the tangible asset insured in the event of a total loss, which is extremely rare. Most claims involve a partial loss.”

Now put that against how underinsurance is enforced. ASIC’s Moneysmart guidance for home insurance states that most policies carry a coinsurance clause — also called an averaging clause — which “limit[s] the insurer’s liability on claims to the proportion of the replacement value that has been insured”, and that “the limitation applies to all claims under the policy, not just total loss or complete rebuild.” Moneysmart adds that most policies engage the clause “whenever there is underinsurance below 80% - 90%”.

The statutory limits nobody mentions

Here is what almost every discussion of average clauses in Australia leaves out, and it materially changes the picture for a homeowner. Section 44 of the Insurance Contracts Act 1984 (Cth) constrains these clauses in three separate ways.

An insurer cannot rely on the clause unless it told you about it properly. Section 44(1): an insurer “may not rely on an average provision” unless, before the contract was entered into, it “clearly informed the insured in writing of the nature and effect of the provision including whether the provision is based on indemnity or on replacement value”.

For an owner-occupied home at 80 per cent or better, the clause cannot reduce your claim at all. Section 44(2) is the provision worth knowing. Where the building is “used primarily and principally as a residence for the insured” (or for family) and the sum insured is “not less than 80% of the value of the property”, then “the liability of the insurer in respect of loss of or damage to the property is not reduced by reason only of the operation of an average provision included in the contract.” That is a statutory safe harbour, not a market convention.

Even below 80 per cent, the reduction is measured against an 80 per cent benchmark — not against full value. Section 44(3) sets a floor: the insurer’s liability cannot be cut below A × S ÷ P, where A is the amount of the loss, S is the sum insured, and P is 80% of the value of the property. The practical effect is that a homeowner insured at 60 per cent of value is not simply paid 60 per cent — the arithmetic runs against the 80 per cent figure, which produces a materially better outcome.

So the popular formulation — “underinsured by a third means your claim is cut by a third” — is wrong for an owner-occupied Australian home, and worked examples that show a straight proportional cut can understate what the Act guarantees. If you see one, check it against section 44.

Two further points matter. Section 44(4) defines “value” as at “the time when the relevant contract was entered into” — and because a home policy is re-entered at each annual renewal, the 80 per cent test resets every year. That is the real argument for a fresh estimate at renewal rather than an indexation bump: the test you have to pass is measured against this year’s value, not the value when you first took the policy out.

And the safe harbour is residential only. For commercial, industrial and investment property there is no 80 per cent protection in section 44 — only the disclosure requirement in 44(1). Average clauses are therefore a far sharper risk for commercial owners than for owner-occupiers, which is close to the opposite of how the topic is usually discussed.

The GST trap, which costs exactly 10 per cent

This one is specific, mechanical, and entirely avoidable. ANZVGP 104 splits the treatment by property type:

“Insurance cost estimates for destructible improvements to non-residential (retail, industrial and commercial) property, do not usually include GST in the calculations as either the insured and/or the insurer will generally be able to obtain the benefit of the input tax credit for any GST paid. This is not the case for insurance cost estimates for home insurance which should include any GST payable as a cost component in the calculation.”

So a commercial estimate is normally quoted excluding GST, because the input tax credit recovers it. A home insurance estimate must include GST, because the homeowner cannot claim it back. Carry a GST-exclusive figure onto a residential policy and the cover is short by ten per cent before anything else has gone wrong.

The guidance also requires the report to say which basis it used — Members “should clearly state whether the insurance cost estimate(s) reported is on a: plus GST (if any) or inclusive of GST (if any) basis.” If your report does not say, that is a question worth asking, because the two figures differ by a tenth.

What has to be inside the number

The rebuild cost of the structure is the starting point, not the total. ASIC’s list of what owners routinely omit is a useful checklist: demolition, site clean-up, asbestos removal, council fees, architect and surveyor services, and temporary accommodation.

To which practice adds the things that are physically on the site but mentally separate from “the house” — sheds, carports, driveways, fencing, retaining walls, pools and landscaping. Each is destructible, each costs money to reinstate, and each is easy to leave out of a per-square-metre estimate of the dwelling.

And then there is time, which is the largest omission of all and the least visible. ANZVGP 104 treats the lead time before rebuilding can start and the reconstruction period itself as components of the estimate, and requires a provision for cost escalation across the period of insurance and the rebuilding process. The sequence for a total loss is not short: make safe, assess, demolish, clear, design, obtain approvals, tender, build. Construction costs continue moving throughout, and the sum insured has to have been set with that movement already inside it.

The guidance defines those periods but does not put a number on them, and neither will we — a “twelve to eighteen month rebuild” is a practice observation, not a standard. What is not in doubt is the direction: a sum insured that prices today’s build cost with no escalation provision is short by the time the slab is poured. This is also the item that dwarfs the code-upgrade allowances discussed below, and it gets a fraction of the attention.

Rebuilding to today’s code — and there is no single “today”

Where a building is rebuilt, the new work is generally built to the rules in force at the time of that work rather than the rules the original was built to — which is the mechanism behind the code-upgrade gap, and why ANZVGP 104’s replacement cost basis speaks of “a current design” using “current materials and techniques”. Two honest qualifications: how far current standards reach into a partial repair is a matter for the approval pathway and the extent of the damage rather than an automatic rule, and ANZVGP 104’s Australian sections do not themselves impose a code-upgrade requirement. The obligation comes from the NCC and from your jurisdiction’s adoption of it.

Here is the part almost every discussion of this gets wrong: as at July 2026 there is no single National Construction Code edition in force across Australia. NCC 2025 was published on 1 May 2026, but adoption is a matter for each state and territory, and they have not moved together:

So a rebuild starting today in Sydney, Brisbane or Adelaide is governed by NCC 2022, not NCC 2025 — and in the ACT and WA the transition period means either edition may apply depending on when the approval lands. The Australian Building Codes Board puts the warning on its own adoption page: “Some NCC requirements may apply at different times or not at all due to state or territory variations.”

The variations run deeper than timing. New South Wales did not adopt NCC 2022’s housing energy provisions as such — it runs enhanced BASIX instead, from 1 October 2023 — so a NSW rebuild meets an energy standard set by a different instrument entirely. The Northern Territory capped its housing energy requirement at 5 stars rather than 7, and Tasmania did not adopt the housing and apartment energy provisions at all. (Those state variations are drawn from the ABCB’s NCC 2022 variations table, which carries its own “correct as at 22 September 2023” note, so confirm the current position for your jurisdiction.)

The practical consequence for a sum insured is unglamorous but real: the code-upgrade allowance in your rebuild cost is jurisdiction-specific, and it moves when your state adopts a new edition. A figure set for a Queensland building in 2026 is priced against NCC 2022; the same building reassessed after 1 May 2027 is priced against NCC 2025.

How big is the energy-efficiency step, honestly?

The move from 6 to 7 stars is the code change most often cited as the driver of higher rebuild costs. The evidence says it is real but modest, and it is worth being straight about the numbers rather than repeating the largest one.

The ABCB’s Decision Regulation Impact Statement for the NCC 2022 energy provisions — modelled by ACIL Allen with Energy Efficiency Strategies and Tony Isaacs Consulting, reported to the Board on 21 July 2022 — put the additional capital cost for a house (Class 1) under its preferred option at $2,199 nationally, ranging from $710 in Queensland to $6,762 in the Northern Territory, with NSW at $3,319 and Victoria at $3,310.

Apartments need stating carefully, because the figure that circulates is the wrong one. The RIS preferred a different option for Class 2 buildings, under which apartments were modelled at $579 per dwelling — lower than houses, not higher. The $4,283 apartment figure widely quoted comes from the option the RIS did not prefer for that class.

All of these are 2021 dollars, discounted at 7 per cent, modelled over a 2022–2051 life cycle for a dwelling built in 2022, with retail costs net of rebates. They are not 2026 construction quotes.

Expressed per square metre, the RIS’s modelled average works out at about $11.65, within a range of roughly $4.50 to $22.00. The same table also carries two much higher figures — $23.63 and $37.05 per square metre — which circulate as evidence that the real cost is triple the model. They are worth reading properly. Both are costings of upgrade solutions designed by one builder’s assessor, on two atypically large houses (386 and 146 square metres), deliberately using the worst-case orientation. And the same table re-costs those identical houses using the RIS’s own techniques at $18.03 and $16.77, which the document notes “fall within the range of dwellings assessed in the RIS for Melbourne”. The headline gap is largely geometry and method, not a hidden cost.

The RIS does concede a genuine limitation: its modelling “assumes a reduction in window size”, and “the impacts of these reductions on amenity or dwelling value were not quantified”. The cheap path to 7 stars is smaller windows, and the model priced that path without pricing what it costs you in the finished house.

Set even the highest of those figures against the ABS average of about $1,967 per square metre for a new Australian house and the star-rating step is roughly two per cent of the build. It belongs in the number. It is not the reason sums insured fall short.

Bushfire construction, where the money actually is

If a rebuild attracts a bushfire attack level, that is a far larger code-upgrade cost than the energy provisions — and it is the item most likely to have appeared since the destroyed building was constructed, because BAL mapping changes.

The best figures we could find are not from a regulator. They come from a named professional study — Project BAL Build, reported by Kathryn Kinnear (a Level 2 BPAD bushfire practitioner) and Julie de Jong (a Level 1 BPAD BAL assessor) in Fire Australia, published around 2020 against the 2018 fourth edition of AS 3959. Expressed as a percentage uplift on a reference house:

The authors’ own reading is that “major cost impacts are likely to be experienced only for BAL-40 and BAL-FZ”, that building to BAL-12.5 through BAL-29 “is not as significant a cost as previously thought”, and that BAL-40 “surprisingly, added less than 10% cost to the reference house”. The spending concentrates in gap and join sealing, upgraded glazing, screens, bushfire shutters or rated window systems, and lining eaves, verandahs and subfloors.

Four qualifications have to travel with those percentages. They come from one house — a single-storey four-bedroom brick veneer and weatherboard home at Tambellup in regional Western Australia — not a national average. The reference house began from a low glazing standard, which inflates the glazing share of the uplift. Material availability is a real constraint the study notes, with some timber species for BAL-40 hard to source, which feeds straight back into lead time. And most importantly for anyone doing this arithmetic:

Do not add the bushfire uplift to the energy uplift. The study is explicit that “many of the upgrades required to comply with AS 3959 are already required in order to comply with Section J – Energy Efficiency provisions of the NCC and good practice generally”. The two overlap substantially, and stacking them double-counts.

We are publishing percentages rather than current dollars deliberately. Percentages transpose to whatever your rebuild cost actually is and they do not go stale; the dollar figures circulating for BAL construction trace back to an unnamed insurer quoted in a newspaper in 2018, which is not a standard we would rely on in a report.

The finding that should change how you think about this

There is one older dataset worth reproducing, because of what it reveals rather than what it costs. An Australian Building Codes Board Regulatory Impact Statement from 2009 — as reproduced in the Fire Australia study — priced bushfire compliance across three different house types, and the spread between them is far more interesting than the spread between BAL levels:

Those are 2009 dollars and seventeen years old, so read the ratios rather than the amounts. And the ratio is the point: at Flame Zone, the elevated lightweight house costs 3.7 times the slab-on-ground house to bring up to standard. At BAL-40 it is 3.6 times. The driver is the raised subfloor, which has to be enclosed and protected.

Two houses on adjoining lots, with an identical bushfire attack level, can therefore carry fire-compliance costs differing by a factor of three or four — purely because one sits on a concrete slab and the other on timber stumps. Note also that BAL-12.5 and BAL-19 cost the same across all three house types; the first genuine cost step is BAL-29.

This is the sharpest argument there is against an indexed or desk-calculated sum insured. No percentage uplift applied to a generic rebuild rate captures it, because the variable is not the hazard — it is how your particular building is put together. It is exactly what ANZVGP 104 has in mind when it lists “specific materials used in the building” and “design of building(s) including soil type, special footings” among the factors an estimate has to weigh.

Why a rebuild can attract a rating the original never had

The mechanism is not that the standard got stricter. It is that the map moved. Western Australia’s Building and Energy guidance puts it directly: “An area might be designated as bushfire-prone after your original home was built. If you plan to rebuild or upgrade your home, you may need to bring it into line with the requirements of the BCA and standards.” On rebuilding after a fire specifically, it notes that where an area was designated bushfire prone after the original building was built, “the new work may need to comply with the bushfire requirements of the current BCA”.

And the procedural point that matters for a sum insured: “the process for approval to upgrade, repair or rebuild will be similar to applying for building approval for a new building.” A rebuild is assessed like new work. There is no grandfathering of the original construction standard.

Two qualifications. That guidance is Western Australian — the designating authority, the mapping and the transition arrangements are state-specific, even though the underlying principle that a rebuild is approved against the current code and the current map is general. And a mapped location does not automatically mean a construction requirement: the same source lists “BAL-LOW — very low risk (bush fire resistant construction is not required)” among the possible assessment outcomes, so the widely repeated claim that every home in a bushfire-prone area must be built to at least BAL-12.5 is not correct. The assessment is site specific, weighing slope, surrounding vegetation and proximity to other buildings.

On cyclonic wind classification we have no comparable figure and will not invent one. The governing wind standard moved to AS/NZS 1170.2:2021 with housing classified under AS 4055, and that revision changed regional boundaries in ways that increase requirements on some sites and reduce them on others. The principle still holds — ANZVGP 104 lists building design, soil type and special footings among its cost factors, and geotechnical requirements for a replacement building can differ from those of the original — but there is no honest “+X per cent for cyclone” to publish.

Where a code change genuinely moves a rebuild

A cleaner example sits in NCC 2025 itself: it introduces a renewable energy generation requirement for Class 3 and Class 5 to 9 buildings, through mandatory on-site solar photovoltaic systems. (Class 3 is residential accommodation such as hotels and boarding houses; Classes 5 to 9 are the commercial, industrial and public-building classes.) A building in one of those classes destroyed today and rebuilt in an adopting jurisdiction may carry a PV installation the original never had — a cost that exists purely because the code moved between construction and reinstatement. That is what a code-upgrade allowance is for, and it is the kind of item a sum insured set from the original construction cost will always miss.

How the number is built, and why method matters

ANZVGP 104 draws a clear quality distinction between two approaches, and states a preference:

“An elemental cost estimate is likely to produce a more robust cost estimate than an estimate based on building cost guides.”

Cost guides — published rates per square metre — are described as “intended to provide indicative cost estimates”. They are a reasonable sanity check and a poor primary method for a specific building, because a rate per square metre carries none of the things that actually move a rebuild price: access, slope, finish level, structural complexity, heritage fabric, or the particular way this building was put together. An elemental estimate prices the components.

This is the practical reason an insurer’s online calculator and a professional estimate diverge. The calculator is a cost guide with a postcode. Our construction costs per square metre figures are useful for orientation and for testing whether a number is plausible — they are not a substitute for pricing the building in front of you.

The catastrophe caveat the guidance states plainly

This is the paragraph most insurance discussions leave out, and the API says it out loud. Sums insured are calibrated for an isolated loss — one building, one event, normal market conditions for labour and materials. Section 9.0 of the guidance addresses large-scale catastrophes directly: estimates are not prepared “in the context of a more widespread catastrophe or disaster such as an earthquake, flood, or bushfire”, and such events “may result in shortfalls between the insurable sums and the total cost of reinstating”.

In a regional event, every builder is quoting at once, materials are scarce and trades are travelling. The cost to rebuild one house when a thousand are being rebuilt is not the cost to rebuild one house.

Australia has measured this. ASIC’s Getting home insurance right report recorded that “rebuilding costs reportedly increased by 75% following Cyclone Tracy in Darwin in 1974, and by 35% in Newcastle after the 1989 earthquake”, and drew the conclusion that follows from it: “even a prudent consumer cannot predict a need to increase the amount of cover to meet jumps in price of this size.”

The post-event evidence matches. After the 2003 Canberra bushfires, the Insurance Disaster Response Organisation reported that destroyed structures “were underinsured, on average, by 40% of the replacement cost”, and the Insurance Council told Parliament the figure was “about 40% for property and 30% for home contents”. ASIC’s own survey of affected owners who rebuilt like for like put it at 27 per cent.

An accurate sum insured is not a guarantee against a regional cost spike. It is the floor beneath it — and the difference between funding a 35 per cent surge and funding a 35 per cent surge on top of an existing shortfall.

How many Australian homes are actually underinsured?

Nobody credibly knows, and that is worth saying plainly rather than filling the gap with a number.

The reference Australian study is ASIC’s, and it is from September 2005. It found that “between 27% and 81% of consumers were underinsured by 10% or more against current rebuilding costs” — a range that wide because it aggregates two very different surveys. A cost-estimating bureau’s 2000 survey of 1,000 homes found 87 per cent underinsured by some amount and an average shortfall of 34 per cent; an Insurance Council survey in 2002, drawing on insurers representing 80 per cent of the market, found 27.5 per cent underinsured by 10 per cent or more.

That threefold divergence is more informative than either figure. Whether a home counts as underinsured depends almost entirely on who estimates the rebuild cost — which is the case for having it estimated properly rather than accepting a default.

Two cautions. First, those numbers are two decades old and no Australian study of comparable scope has replaced them, so treat any confident modern percentage sceptically — including figures circulating with an APRA attribution. APRA’s March 2026 climate vulnerability assessment states explicitly that “underinsurance was not directly modelled in this stress test due to its complexity”. Second, what APRA does report is the affordability squeeze underneath all of this: between 2010 and 2025 “Australian home insurance premiums rose by an annual average rate of 7.2%, while wages grew annually by 3.1%”, with construction cost inflation named among the drivers.

One boundary worth knowing

An insurance policy is a financial product. A valuer who does not hold an Australian Financial Services Licence cannot advise you on whether your policy provides adequate cover, and ANZVGP 104 requires Members without a licence to make clear that their report is limited to the insurance cost estimate itself.

The division of labour is clean: the valuer establishes what it costs to rebuild, and your broker or insurer advises on the policy that responds to it. A report that tells you both is a report to read sceptically.

Methodology

• Professional guidance: ANZVGP 104 Insurance Cost Estimates, Australian Property Institute (issued jointly with the New Zealand Institute of Valuers and the Property Institute of New Zealand), published 28 April 2025, effective 1 July 2025, read directly. Quotations are from paragraphs 1.2, 3.2, 4.1, 6.2.1, 6.2.2, 6.2.4, 6.3, 6.10 and 9.0, with the lead time, reconstruction period and cost escalation provisions drawn from paragraphs 5.0, 6.2, 6.8 and 6.9.
• Scope note: ANZVGP 104 is a joint Australia–New Zealand paper whose later addenda apply to New Zealand only, and those addenda contain its most detailed material on regulatory upgrade, inflation allowances and common property. Nothing in this article is drawn from them. Where the Australian sections are silent, we say so rather than importing New Zealand practice.
• Underinsurance mechanics, the coinsurance threshold and the forgotten-costs list: ASIC Moneysmart, Underinsurance: what it is and how to avoid it, read directly.
• NCC adoption dates: read directly from the ABCB’s NCC 2025 state and territory adoption information table on 30 July 2026, cell by cell. State-variation examples come from the ABCB’s NCC 2022 variations page. The NCC 2022 variations detail carries the ABCB’s own “correct as at 22 September 2023” note and is presented as that snapshot, not as the current position.
• Energy-efficiency cost figures: ABCB Decision Regulation Impact Statement for the NCC 2022 residential energy efficiency provisions (modelling by ACIL Allen with Energy Efficient Strategies and Tony Isaacs Consulting; report to the Board dated 21 July 2022). Per-dwelling capital costs are the preferred-option Class 1 figures, in 2021 dollars discounted at 7 per cent over a 2022-2051 life cycle for a dwelling built in 2022, with retail costs net of rebates and subsidies. We publish the RIS’s per-square-metre range and the builder-measured figures from the same table rather than the modelled point estimate alone, because the document itself records a threefold gap between them and attributes part of it to an assumed reduction in window size whose amenity impact it did not quantify.
• Statutory position on average clauses: Insurance Contracts Act 1984 (Cth) section 44, subsections (1) to (4), read directly. Note that ANZVGP 104 references only section 44(1) and does not mention the 80 per cent safe harbour in 44(2) or the floor in 44(3); we cite the Act rather than the guidance paper’s summary. Note also that the statutory exclusion protecting dwelling houses in New Zealand sits in different legislation and is not transposable.
• We have deliberately not reproduced published worked examples of a straight proportional reduction on an owner-occupied home, because section 44(3) sets a floor those examples can fall below.
• Underinsurance prevalence: ASIC Report 54 Getting home insurance right, September 2005 — the reference Australian study, now two decades old, with each figure carrying its survey year in the text. Post-catastrophe figures and the Cyclone Tracy and Newcastle earthquake cost increases are from the same report. Premium and wage growth, and the protection-gap figure, are from APRA’s Mind the Gap: An Insurance Climate Vulnerability Assessment, 24 March 2026, which states that underinsurance itself “was not directly modelled in this stress test”.
• A caution for anyone researching this topic: a figure claiming APRA confirmed in 2026 that 83 per cent of Australian homes are underinsured by an average of 34 per cent is in circulation and is not APRA’s. The 34 per cent originates in a survey from the year 2000. We have not published it.
• We have not claimed that ANZVGP 104 requires rebuilding to current codes. It does not say that. The link between replacement cost and current standards runs through its replacement cost definition (“a current design … current materials and techniques”), and the code requirements themselves come from the NCC and each jurisdiction’s adoption of it.
• Bushfire attack level uplifts: Project BAL Build, reported by Kathryn Kinnear (Bio Diverse Solutions, Level 2 BPAD bushfire practitioner) and Julie de Jong (H + H Architects, Level 1 BPAD BAL assessor) in Fire Australia, published c. 2020 and written against the 2018 fourth edition of AS 3959. This is a named professional study of a single reference house in regional Western Australia, not a regulator’s dataset, and it is attributed as such. We publish its percentages rather than dollar figures because the dollar figures in circulation trace to an unnamed insurer quoted in a newspaper in 2018. The study’s own warning against adding AS 3959 and NCC Section J uplifts together is reproduced, because that double-count is the obvious error to make with these numbers.
• The BAL cost table by house type is an Australian Building Codes Board Regulatory Impact Statement from 2009, as reproduced in Figure 1 of the Kinnear and de Jong study — we did not obtain the RIS itself, since the ABCB’s online library does not reach back that far. It is a second-hand reproduction of a primary regulator document, in 2009 dollars, and is used for its ratios rather than its amounts.
• The rebuild trigger and the BAL-LOW correction come from the Government of Western Australia, Building and Energy, Building in bushfire prone areas. Attributed to WA because the designating authority, mapping and transition arrangements are state-specific; the general principle that a rebuild is approved against the current code and map is not.
• Cyclonic wind: no cost differential published. AS/NZS 1170.2:2021 and AS 4055 govern, and the 2021 revision changed regional boundaries in both directions, so no defensible single percentage exists. The principle that a replacement building can carry structural or geotechnical requirements the original did not is supported by ANZVGP 104’s cost-factor list; the quantum is not.
• We have not attributed any rebuild duration or lead-time figure to ANZVGP 104. The guidance defines the reconstruction period but does not quantify it, and a “12 to 18 month rebuild” is a practice observation, not a standard.
• We have not published anything on Australian strata or common property from this source. ANZVGP 104’s common-property material sits in its New Zealand addendum and addresses New Zealand tenure. For the Australian position see our strata insurance requirements by state guide.

Frequently asked questions

Is a building insurance valuation the same as a property valuation?

No, and the guidance is explicit that it is not even the same kind of exercise. ANZVGP 104 defines an insurance cost estimate as “the result of a calculation … of the cost of replacing, or reproducing, the tangible asset” — a cost figure rather than an opinion of value. A market valuation answers what a buyer would pay; an insurance cost estimate answers what rebuilding would cost. The two numbers routinely differ by a wide margin in both directions.

What is the difference between reinstatement and replacement cost?

Reinstatement cost is the cost to return the building to “a condition equal to, but not better or more extensive than its condition when new”. Replacement cost is the cost of an equivalent asset “of a current design and constructed or manufactured using current materials and techniques”. Reinstatement looks back at the building as built; replacement looks at what you would build today. On an older building they produce different numbers, and your policy is written on one of them.

Does underinsurance only matter if my house burns down?

No — the opposite. ANZVGP 104 notes that total loss is “extremely rare. Most claims involve a partial loss”, and ASIC’s guidance confirms that a coinsurance clause “applies to all claims under the policy, not just total loss or complete rebuild”. A reduction on an ordinary repair is the common case; the total loss is the rare one. How large that reduction can be depends on the property type: on an owner-occupied home section 44 of the Insurance Contracts Act 1984 (Cth) caps it, and removes it entirely at 80 per cent of value or better. Commercial and investment property have no such protection.

Should my sum insured include GST?

For home insurance, yes. ANZVGP 104 states that home insurance cost estimates “should include any GST payable as a cost component in the calculation”, because a homeowner cannot recover it. Non-residential estimates usually exclude GST, since the input tax credit recovers it. The report should state which basis it used — if it does not, ask, because the difference is ten per cent.

Which building code will my rebuild have to meet?

The one in force in your jurisdiction when the rebuild is approved — and as at July 2026 that is not the same edition everywhere. NCC 2025 was published on 1 May 2026 and adopted from that date in the ACT, Tasmania, Victoria and Western Australia, but New South Wales, Queensland and South Australia do not adopt until 1 May 2027, and the Northern Territory has not adopted it. A rebuild in Sydney, Brisbane or Adelaide today is therefore built to NCC 2022. The ABCB warns on its own adoption page that “some NCC requirements may apply at different times or not at all due to state or territory variations”.

Did the 7-star energy requirement blow out rebuild costs?

Not by much, on the published evidence. The ABCB’s Decision Regulation Impact Statement put the additional capital cost for a house under its preferred option at $2,199 nationally in 2021 dollars, ranging from $710 in Queensland to $6,762 in the Northern Territory. Per square metre that is roughly $4.50 to $22.00 by the RIS’s own modelling, against builder-measured figures in the same table of $23.63 and $37.05. Even at the top of that range it is a low single-digit percentage of a rebuild at current rates. It belongs in a sum insured; it is not the reason sums insured fall short.

How much does a bushfire rating add to a rebuild?

On the best figures we could find — Project BAL Build, a professional study of one reference house in regional WA reported in Fire Australia around 2020 — the uplift runs about 3.8% at BAL-12.5, 7.7% at BAL-29, 9.6% at BAL-40 and 20.1% at Flame Zone. Treat those as indicative of scale rather than as a national average, since they derive from a single house. Two cautions: the study warns that many AS 3959 upgrades are already required by the NCC’s energy provisions, so the bushfire and energy uplifts must not be added together; and material availability for the higher ratings can extend the rebuild period.

Does the same bushfire rating cost the same on every house?

No, and the difference is larger than the difference between ratings. An ABCB Regulatory Impact Statement from 2009, reproduced in the Fire Australia study, priced Flame Zone compliance at about $20,885 for a slab-on-ground house and $76,679 for an elevated lightweight house — a factor of 3.7 at the same rating, driven by the need to enclose and protect a raised subfloor. Read those 2009 figures as ratios rather than amounts. The practical implication is that a generic per-square-metre uplift cannot capture bushfire compliance, because the variable is how your building is constructed rather than the hazard it faces.

Are insurer online calculators good enough?

They are a cost guide with a postcode. ANZVGP 104 describes published cost guides as “intended to provide indicative cost estimates” and states that “an elemental cost estimate is likely to produce a more robust cost estimate than an estimate based on building cost guides”. A calculator cannot see access constraints, slope, finish level, structural complexity or heritage fabric. It is a reasonable plausibility check and a poor basis for the sum insured on a specific building.

Will an accurate sum insured protect me in a bushfire or flood?

Not completely, and the guidance says so. Section 9.0 notes that estimates are not prepared in the context of a widespread catastrophe, and that such events “may result in shortfalls between the insurable sums and the total cost of reinstating”. When many buildings are rebuilt at once, costs move. An accurate figure is the floor, not a guarantee.

Can my valuer tell me if my policy is adequate?

Only if they hold an Australian Financial Services Licence. An insurance policy is a financial product, and ANZVGP 104 requires unlicensed Members to make clear that their advice is limited to the cost estimate. The valuer establishes the rebuild cost; your broker advises on the cover.

---

Sources:

• Australian Property Institute — standards and guidance papers (ANZVGP 104 Insurance Cost Estimates, effective 1 July 2025)
• ASIC Moneysmart — Underinsurance: what it is and how to avoid it
• NCC 2025 state and territory adoption information — Australian Building Codes Board
• NCC 2022 state and territory adoption dates and variations — ABCB
• Insurance Contracts Act 1984 (Cth), section 44 — average provisions
• ASIC Report 54 — Getting home insurance right (September 2005)
• APRA — Climate Vulnerability Assessment (Mind the Gap: An Insurance Climate Vulnerability Assessment, 24 March 2026)

This article is general information about insurance cost estimates and valuation practice — it is not financial, insurance or legal advice, and it does not address the adequacy of any policy. Costing bases, guidance and policy terms change; confirm the position for your building and your cover with your valuer and your broker. Last verified 30 July 2026.

See also: Building Insurance Valuations · Replacement Cost vs Market Value · Strata Insurance Valuation Requirements by State · Construction Costs per Square Metre · How Plant and Equipment Is Valued · Strata Insurance Valuations · DRC Valuations