The Co-Insurance Clause: What Every Building Owner Must Know

Written by James Kirk, Compliance Manager and William Paull, Co-Founder & Account Director · 20 min read

What is a co-insurance clause in commercial insurance?

A co-insurance clause, also called an average clause, is the fine print that lets your insurer reduce a partial claim when you are underinsured. Insure your building for less than its true rebuild cost, commonly under 80% of it, and the insurer can scale a partial payout down in proportion.

Most building owners meet this clause for the first time in the worst possible place: a renewal document they skim, or a claims letter that pays out far less than they expected. In this guide from Consolidated Insurance Brokers you will see exactly how the clause works, the one distinction almost nobody gets right, and how to make it irrelevant to you before you ever need to claim.

This is not a rare trap. Most owners never realise they are exposed: when businesses are asked, only about one in ten think they are underinsured (Insurance Council of Australia research, 2015; Vero SME Insurance Index, 2025), yet when valuers actually measure it the average building comes up around 24% short, and 31% short for industrial property (MCG Quantity Surveyors, 2024). Rebuild costs have climbed hard while most sums insured sat still, so a number that was right a few years ago is very likely wrong today. The same gap sits behind many of the commercial building insurance claims that pay less than the owner expected. If you want the full story on how you ended up underinsured, we cover it in Underinsurance: The Biggest Risk to Commercial Building Owners. This article is about the clause that turns that gap into a smaller cheque.

How does the co-insurance clause work on a partial loss versus a total loss?

The clause behaves in two different ways depending on the size of the loss, and blurring them is the most common mistake people make. On a partial loss, the clause scales your payout down. On a total loss it scales nothing: you receive your full sum insured, and the shortfall to the real rebuild cost is yours.

Partial loss: the payout gets scaled down

Say your building would genuinely cost $1 million to rebuild today, but you have it insured for $500,000. Most commercial policies test your sum insured against at least 80% of the true rebuild value, and some wordings set that test at 85%. Take the 80% case. You needed to be insured for $800,000, and you carried $500,000. That is 62.5% of the required amount, so the insurer can pay 62.5% of a partial claim. A $100,000 roof replacement then pays $62,500, and the missing $37,500 is yours to find, even though the loss was nowhere near your sum insured.

Here is a detail that even professional firms routinely get wrong, and getting it right is worth money to you. The real clause is kinder than the rough version most articles show. The common shortcut divides your sum insured by the property's full value: $500,000 divided by $1 million is 50%, which would pay only $50,000. The actual Australian wordings divide by the required percentage of value, usually 80%, which is why the honest answer is $62,500, not $50,000. One major Australian insurer publishes the formula plainly in its business pack wording: the claim payment is the loss multiplied by your sum insured, divided by the co-insurance percentage times the building's full rebuild value. That insurer's own worked example pays $937,500 of a $1 million loss on property genuinely worth $2.88 million that was insured for $1.8 million, because that wording tests a figure 20 per cent higher again than the sum insured.

Real determinations show the clause biting in exactly this way. In one published complaint decision (case 795519, as reported by insuranceNEWS.com.au), a commercial building insured for $300,000 against an assessed rebuild value of about $568,000 had a roughly $35,280 fire repair claim scaled back to around $23,285 under the average clause. The Australian Financial Complaints Authority upheld the clause because the insurer had clearly disclosed it, though it did add 15% to the settlement to account for the repair quote being old.

Total loss: you get the full sum insured, and no more

Now burn the same building to the ground. The insurer pays the full $500,000 sum insured, with no scaling at all. The problem is that the building costs $1 million to put back. You are $500,000 short, standing on a vacant block with a mortgage that did not disappear. There is a stubborn myth, repeated even by some professional firms, that average slashes a total-loss payout below the sum insured. It does not. On a total loss you receive the full sum insured. The clause hurts you on partial losses, and partial losses are the vast majority of claims, as the industry figures below show.

That is the whole reason we never let the two cases blur together. Underinsurance does not announce itself at the moment you sign. It waits, quietly, and then decides how big your cheque is on the day you can least afford a surprise.

Does the co-insurance clause apply to contents, fitout and lost income too?

Yes. The clause is not a building-only problem. Contents, stock and fitout are not tested on their own: on the property side they go into one combined total with the building, where you insure one, tested on the same rough 80% basis. The cover that replaces your income does carry a separate test of its own. A tenant who fitted out a shop years ago and never revisited the number is exposed the same way a landlord is.

Say you lease a shop, so the building is your landlord's and is not on your policy. Your fitout, contents and stock would genuinely cost $500,000 to reinstate, and you insure them for $250,000 in total. That $250,000 is everything you have insured at the address, so it is the figure the clause tests. A kitchen fire does $200,000 of damage, a partial loss. Tested against 80% of the true value, you needed at least $400,000 of cover and carried $250,000, which is again 62.5% of the requirement. The insurer can pay 62.5% of the claim: $125,000 on a $200,000 loss, even though your sum insured was $250,000. In a total loss the rule is the same as it is on a building: you receive the full $250,000, but reinstatement costs $500,000, so the gap is yours regardless.

The cover that replaces your income while you rebuild, business interruption cover, carries an underinsurance test of its own, and this is where the biggest recent shortfalls have surfaced. In a 2025 determination (case 12-00-1023207), the Australian Financial Complaints Authority upheld an insurer's use of an 80% underinsurance clause on a business interruption claim, a COVID-era claim by a Sydney salon whose income cover was set as low as $150,000 (and $350,000 for the earlier claim periods) against a true insurable figure of more than $1.15 million. The assessed loss was cut to a settlement of $143,512. Same clause, same maths, a very different asset.

Why do insurers use a co-insurance clause?

The clause is not a trap; it is a fairness mechanism. Your premium is priced on the value you declare. Declare half the true value and you pay about half the premium, while expecting a partial loss covered in full. Average restores that balance, so an owner who insures accurately is not subsidising one who does not.

Understanding why it exists tells you how to beat it. Insurance educator Professor Allan Manning frames it the same way: underinsuring nearly halves the premium while leaving the insurer carrying almost the full exposure on the partial losses that make up most claims, and average is what corrects that imbalance. Read the clause that way and its logic points straight at the solution. The clause rewards accurate declared values and penalises guesses. It is not asking you to insure for more than the building is worth. It is asking you to insure for the truth.

Why is the co-insurance threshold 80% and not 100%?

Because the clause is not asking you to be perfect, it is asking you not to be badly wrong. The gap between the test figure and full value works as a tolerance: rebuild costs move, valuations date, and nobody can name the exact figure on the day. But that tolerance is not identical on every policy. Across the business pack wordings we place, the test is generally set at 80%, and some are set at 85%. So a sum insured at 80% of rebuild value clears the clause on one wording and falls short on another, which is why the figure that matters is the one in your own policy rather than the one in an article. Insure to what the building would genuinely cost to rebuild and the clause never touches you, whichever test your wording uses.

That much is fairly well known. What almost nobody knows is that on several of the wordings we place, the figure the clause tests is not the number you declared. It is a figure 20% higher again.

The number the clause tests may not be the number on your schedule

Business pack wordings often define a "limit of indemnity" or "limit of liability" that sits above your sum insured, commonly at 120% of it, and several of them run the co-insurance calculation against that higher figure rather than against the sum insured itself. Others run it against your sum insured exactly as declared. Both approaches are clearly written, both are lawful, and the difference between them is real money on a claim.

Here is what that difference looks like, using the worked examples the wordings publish themselves. Several of the wordings we place print an example built on the same numbers: a property worth $200,000, so the 80% test sits at $160,000, and a $100,000 partial loss. They all arrive at the same $90,000 payout. What differs is where the tested figure of $144,000 comes from. In one wording's published example the owner declared $120,000 and the policy lifted it to $144,000 for the test. In another's, $144,000 is what the owner had to declare in the first place. Same building, same loss, same payout, and a difference of $24,000 in what you had to insure for to get it.

So the honest answer to "how much do I need to insure for" is: it depends on which of those two structures your wording uses, and you cannot tell from the schedule. It is one of the clearest examples on this whole page of why a policy is not a commodity, and why the cheapest quote is not automatically the same cover.

The test looks at everything at the premises, not just the building

One more feature of the 80% rule surprises people, and it can catch an owner who thought they had the building right. The clause tests your total against the total. On every business pack wording we place, where your policy settles claims on a reinstatement or replacement basis, the comparison is between what you insured for all the property at that address, buildings, contents, stock and any specified items together, and 80% of what all of it is genuinely worth. A policy that settles on an indemnity basis instead, paying the depreciated value of what you lost rather than new for old, words the same test a little differently, so it is worth knowing which of the two your policy uses.

That means the parts talk to each other. An owner whose building figure is spot on but whose contents or stock figure was set years ago is not partly protected. They are tested on the combined number, and a shortfall anywhere in it can scale down a claim on the part they got right. The reverse is true too: being generously insured on one item can help carry a slightly light figure on another, because the test never looks at a single line in isolation. Be careful how far you push that, though. It only helps with the proportion the clause calculates. It does not lift a limit, and at least one wording still caps what it will pay against the sum insured shown for each individual item, so a generous building figure will never pay out a short stock figure.

Stock is the item that moves most, which is why one wording on the panel expressly counts the automatic seasonal uplift on stock into this very calculation. If your stock swings through the year, the seasonal stock clause is the companion piece to this one.

How this sits with the worked example above

The $1 million building example earlier on this page uses the stricter of the two structures: it tests the $500,000 sum insured exactly as declared, which is why it lands on 62.5% and a $62,500 payout on a $100,000 roof claim. Keep that as your working assumption, because it is the outcome you should plan for and it is genuinely how several of the wordings we place operate. On a wording that tests a buffered figure, the same facts would pay more, because $500,000 lifted by 20% is $600,000 against the same $800,000 requirement. The margin is welcome when you have it. It is not a plan, and it is not a substitute for a sum insured that reflects what the building would actually cost to rebuild.

When does the co-insurance clause not apply?

The clause has more off-switches than most owners realise, and knowing them is genuinely reassuring. It does not bite when your sum insured is high enough, it usually stands aside on small claims, some costs sit outside it entirely, and the law puts one condition on the insurer before it can rely on the clause at all.

When you are insured to the required percentage. If your sum insured meets the threshold in your wording, average does not apply. Smaller business pack policies commonly test at 80% of full value. Some larger, individually underwritten programs may apply a higher threshold again. Either way, hit the mark and the clause is switched off. Note too that adequacy is measured at the start of your policy period, not on the day of the loss, so mid-year construction-cost inflation does not itself trigger average. Starting the year underinsured is what does.

On small claims. Most wordings waive average entirely below a threshold, so a minor claim is paid without any underinsurance calculation. The catch is that the threshold is measured differently by different products, so there is no single "standard" number: many business packs waive it where the loss is under 10% of the sum insured, while some larger programs use a different threshold again. This is a practical reassurance point that often goes unmentioned, and the answer is always the same: check your wording.

On extra costs that sit outside the calculation. Statutory extra costs of reinstatement, the added expense of rebuilding to current building codes, are quarantined from the average clause in most wordings. They are neither scaled down by average nor counted in the values used to test whether you are adequately insured. So you should not inflate your declared value to allow for them, but you should set the amount you insure for those extra rebuilding costs deliberately.

Unless the insurer told you about it in writing first. Under section 44 of the Insurance Contracts Act 1984, an insurer cannot rely on an average clause at all unless, before you entered the contract, it clearly informed you in writing of the clause and its effect. In practice insurers meet this by putting a prominent notice, usually with a worked example, in the policy document, and the Australian Financial Complaints Authority has accepted that such a notice satisfies the obligation. So the law is a genuine precondition, not a loophole to rely on. The dependable control is still your sum insured.

There is even pressure to narrow the clause further. A 2024 parliamentary inquiry into flood insurance recommended that small and medium businesses be given clearer guidance on how averaging provisions work, and asked the government to consider prohibiting them for small business policies altogether. Nothing has changed in the wordings yet, but the direction of travel is worth watching.

Can a broker negotiate an underinsurance settlement?

Sometimes, yes. The average clause itself is rarely negotiable, but the numbers inside the calculation are, and they are worth checking. In a 2025 storm claim we handled, our review of how the loss adjuster had valued the building changed the payable proportion, and the insurer amended its settlement to our figure exactly. The clause still reduced the claim. The point is that the reduction has to be calculated on the right numbers.

Here is that claim, told honestly, because it shows both what a broker can move and what nobody can. In late 2024 an owner of a retail building in regional New South Wales asked us to arrange cover. We measured the rebuild properly and recommended a sum insured of at least $1.99 million. The owner instructed us to place $500,000, about a quarter of that figure, and we set out the risk in writing before cover started, with a worked example showing a partial claim could pay as little as 25 cents in the dollar.

In January 2025 a storm damaged the building. A partial loss, the exact case this clause is built for. The insurer's loss adjuster applied the average clause, as the policy entitled it to do, and the client engaged a paid claims advocate to fight the reduction. The two points that actually moved the settlement did not come from the advocate. They came from our broking team going through the adjuster's calculation against the policy wording, line by line.

First, GST. The client's company was registered for GST, so any GST the insurer paid out would come back to them as a tax credit. That meant the building had to be valued without GST when testing how far short the sum insured fell. Second, the policy paid certain costs, including debris removal and professional fees, from a separate allowance sitting on top of the sum insured, so those costs did not belong in the building value used for the test either. Correcting both brought the tested value down to about $1,308,000, and a lower tested value means a higher proportion of the claim gets paid.

The adjuster reviewed its earlier offer, adopted our building value, and the settlement was amended and paid on that basis.

Now the part we will not dress up. The clause still did its job. The client recovered less than half of a partial loss and funded the rest, on a building they had been advised in writing to insure for roughly four times as much. Our review made the reduction honest; it did not make the underinsurance safe. If your own claim has been reduced for underinsurance, the checklist in the FAQ at the end of this page is where to start, and a line-by-line review like this one is precisely what to ask a broker for.

That claim also changed how we place cover. On the business pack policies we place, we will no longer put a commercial building on cover below 60% of the measured rebuild figure from the desktop building replacement valuation we commission. If you believe our figure is wrong, you can commission your own valuation from a registered valuer and we will work from that instead. That 60% is a floor, not a recommendation, and below your policy's threshold the average clause still applies. The floor exists so that no client of ours can drift to a quarter of the real number without making a deliberate, documented decision.

How do you make the co-insurance clause irrelevant?

The clause only matters if you are underinsured, so insure to a proper rebuild figure and it never touches you. That figure is not what you paid, not the council valuation, and not last year's number nudged up a few percent. It is today's genuine rebuild cost, assessed rather than guessed.

The catch is that "assessed rather than guessed" is exactly where most owners fall down. A rebuild figure carried over from a purchase price or an old council notice is a guess dressed up as a number, and it is the guess the average clause tests you against.

That is why we do not let clients guess it. For commercial building owners, Consolidated Insurance Brokers commissions a desktop valuation at no cost to you, so your sum insured is set by a registered valuer's desktop assessment rather than a figure that quietly drifts out of date. We commission it for our purposes as your broker, to inform the advice we give you. Where we find a gap, we close it at a pace you can manage, with options at every step rather than a take-it-or-leave-it, including where a higher excess keeps full cover affordable.

Accurate valuations also unlock the clause's best-kept escape hatch. Some insurers now switch off the average clause entirely for property that was insured to full value on an approved valuer's valuation updated within the previous 12 months. In other words, on several of the wordings we place, insuring at the full valuation figure can remove the clause from your policy, not just help you satisfy it. That is a concrete reward for getting the number right, and it is exactly the sort of thing a policy nobody re-tests will never pick up.

The habit is the hard part, not the arithmetic. The 2025 Vero SME Insurance Index found that only 42% of small and medium businesses review their sums insured every year, which is precisely how a building that was correctly insured in 2021 quietly falls behind. A desktop valuation you refresh at renewal, not a number you index and forget, is what keeps the co-insurance clause permanently irrelevant to you.

Not sure where your own sum insured stands? Check your underinsurance risk in two minutes.

Get your rebuild figure checked

Frequently asked questions

Is co-insurance the same as the average clause?

Yes. In Australian commercial insurance, "co-insurance clause", "average clause" and "underinsurance clause" are three names for the same mechanism, and the Insurance Contracts Act calls it an "average provision". Different policies simply label it differently, so do not assume a policy without the word "co-insurance" is free of it.

The wording proves it: some larger tailored programs literally title the provision "Co-Insurance", while others head the identical clause "Average" or "Underinsurance". One word of caution, because it trips people up: "co-insurance" has a second, unrelated meaning in the market, where two or more insurers each take an agreed share of one large risk. That is a placement arrangement, not the underinsurance penalty this article is about. The health-insurance "coinsurance" you see on overseas websites is a third, different thing again and does not describe Australian commercial policies.

Does the co-insurance clause apply to home insurance?

Almost never, but not because the law shields your home. Standard Australian home building and contents wordings simply do not include an average clause, and an insurer cannot apply a clause that is not in its wording and was not disclosed in writing before you bought. Commercial wordings do include it.

The statutory protection for homes adds less than most people think. For a building used mainly as a residence, the Insurance Contracts Act does cap an average clause: at 80% of value or better it cannot reduce a claim at all, and below that a formula limits the reduction. But that formula is the same 80% calculation commercial wordings already apply, so if a home insurer did write the clause in, an underinsured homeowner would be cut back just as hard. What actually forgives your home policy is that the clause is not there - and that is exactly the forgiveness your commercial building policy does not give you.

What percentage does the co-insurance clause use?

Commonly 80% of your building's true replacement value on a standard business pack policy, meaning you can be up to 20% short before the clause bites. Some larger, individually underwritten programs may apply a higher threshold. Your own percentage lives in your policy wording, so check it rather than assume.

Three points make the percentage less frightening than it sounds. First, adequacy is tested at the start of your policy period, not on the day of the loss. Second, most wordings waive the clause on small claims, commonly under 10% of the sum insured on business packs, while larger programs may use a different threshold. Third, several of the wordings we place do not even test the figure you declared: they test a limit of indemnity set at 120% of it, which means the figure you have to declare to clear the test drops from 80% of the real value to roughly two thirds of it. Others test your declared figure exactly, which is why the safe assumption is the stricter one. The percentage that really decides your outcome, though, is how close your sum insured sits to a genuine rebuild figure.

The insurer reduced my claim payout because of underinsurance. Can I do anything?

Possibly. An underinsurance reduction is not always the last word, and there are specific, checkable points before you accept it. Work through them in order, and get a broker to review the file, because the assessed value and the clause mechanics are both open to challenge.

First, check that the insurer clearly disclosed the average clause in writing - for example, a notice in your policy documents - before you bought, as section 44 of the Insurance Contracts Act requires. Second, challenge the insurer's "full replacement value", because the reduction is only as valid as that figure, and your own valuer's number can move it. Third, confirm the small-claim waiver in your wording does not switch the clause off for a loss your size. Fourth, ask whether an allowance should be added because the repair quote is old, as the complaints authority has done. If it is not resolved, the pathway is the insurer's internal dispute resolution first, then the Australian Financial Complaints Authority, which reviews both the value and the clause and does not always find for the insurer.