The question behind the question
Nothing in Division 152 of the ITAA 1997 says 'obtain a formal valuation'. The provisions say something quieter and more demanding: the eligibility tests turn on market value, the taxpayer self-assesses, and if the ATO asks, the taxpayer must be able to show how the value was arrived at. The ATO's market valuation for tax purposes guidance is explicit that taxpayers can determine market values themselves — and equally explicit that the value must be supported by appropriate evidence, whoever determines it. So the honest answer to 'do I need a formal valuation?' is: legally, not always; practically, it depends on which eligibility gateway you are relying on and how much room for error the numbers leave you. That is a decision, not a habit — and it deserves a framework. This article sets one out. It is a companion to our separate piece on what evidence these claims need; this one is about whether you need the formal work at all.
Where the $6m test forces the issue
There are two gateways into the concessions. If aggregated turnover is under $2m, you are a CGT small business entity and the basic gateway asks nothing about asset values. If turnover is $2m or more, the only route in is the maximum net asset value test in section 152-15: the net value of the CGT assets of the taxpayer, entities connected with the taxpayer, and affiliates must not exceed $6m just before the CGT event. Three features of the test do the damage. It is market value, not book value — goodwill that has never appeared on a balance sheet counts in full. It sweeps in connected entities and affiliates, so the valuation question multiplies across the group. And the $6m figure has never been indexed, so ordinary asset growth pushes more taxpayers toward the line every year. A formal valuation becomes practically unavoidable in the following scenarios.
- ·Turnover is $2m or more, so the MNAV test is the only gateway — every dollar of the eligibility position now rests on market values.
- ·The business is a profitable trading entity with material goodwill. Book value tells you nothing; only an earnings-based valuation establishes what the goodwill adds to the net asset position.
- ·The position sits anywhere near $6m. If the margin between your estimate and the threshold is smaller than the honest error band around that estimate, the estimate cannot carry the claim.
- ·There are multiple connected entities or affiliates — trusts, holding companies, related-party loans — whose assets have never been valued at market and whose inclusion or exclusion is itself contestable.
- ·The assets being counted include private company shares or trust units, which need their own valuation rather than a look-through guess.
When a reasonable estimate is legally sufficient
Self-assessed values have a legitimate place. The ATO's guidance accepts that not every market value requires an external expert — what it requires is that the process behind the value be reasonable and documented. A reasonable estimate is generally sufficient where the turnover test does the work: if aggregated turnover is under $2m, the MNAV test is simply not engaged and no balance-sheet valuation is needed for the basic conditions. It is sufficient where the net asset position is demonstrably far below $6m on any defensible view — if the group's assets could not reach the threshold even with aggressive values on every line, a working paper showing exactly that calculation is a reasonable position. And it is sufficient where the material values rest on observable evidence: a contemporaneous independent appraisal of the premises, a genuine arm's-length offer for the business, recent comparable sales. Notice what each of these has in common — a written calculation in the file, prepared before the claim, that a reviewer could follow. 'The accountant and I discussed it and thought we were under' is not an estimate. It is a hope.
Where the estimate becomes reckless — and who carries the risk
The line between reasonable and reckless is about foreseeable risk. An estimate is heading toward reckless where goodwill is material and nobody has tested what it is worth; where the total sits close enough to $6m that a plausible revision tips it over; where a connected entity's assets were excluded on an untested assumption or included at book value; or where the 'estimate' was produced after the CGT event, with the threshold in view and the incentive pointing one way. The risk allocation matters. Under self-assessment the taxpayer — not the ATO — bears the burden of showing an assessment is excessive, which in a concession dispute means proving the market values that existed at the event date. The client carries the tax, penalties and interest. The accountant carries the file: if the only support for a concession claim worth hundreds of thousands of dollars is an unsupported figure in the working papers, the adviser's position is uncomfortable in a way no engagement letter fully cures. And penalty outcomes track the quality of the process — a documented, honest estimate that turns out wrong is a different case from a number that was never really estimated at all.
What happens at ATO review if you self-assessed
A review typically starts with one question: show us how the market value was determined. If the answer is a defensible working paper or a formal valuation, the inquiry has something to engage with. If the answer is thin, the sequence runs roughly as follows. The ATO can test the values itself, including by commissioning its own valuation. If the revised values push the group over $6m, the concessions do not shrink — they disappear, and the amended assessment brings the entire capital gain back into assessable income, not just the margin above the threshold. Amendment periods generally allow this for two years for many small business taxpayers and four years for others, with no time limit where fraud or evasion is alleged. Administrative penalties then apply to the shortfall: 25% for failure to take reasonable care, 50% for recklessness, 75% for intentional disregard. Shortfall interest accrues on top — and interest charges incurred on or after 1 July 2025 are no longer deductible, so the carrying cost of a failed claim is now worn in full. Then the practical sting: to contest any of this, you must prove market value at a date now years in the past. That is a retrospective valuation, bounded by contemporaneous evidence, commissioned at exactly the moment the evidence is hardest to assemble.
A decision framework accountants can quote
Stripped to its logic, the decision looks like this.
- ·Yes — commission a formal valuation if the MNAV test is your only gateway and the margin between your estimate and $6m is smaller than the plausible error in that estimate.
- ·Yes — if goodwill or unbooked intangibles are a material part of the net asset position; book value cannot answer a market value question.
- ·Yes — if connected entities or affiliates hold assets that have never been valued at market, or whose inclusion in the calculation is itself arguable.
- ·Yes — if the CGT event has already occurred and the claim may be reviewed; the valuation will need to be retrospective, the evidence discipline is stricter, and the file gets harder to build with every year that passes.
- ·No — if aggregated turnover is under $2m and the CGT small business entity gateway applies; the MNAV test is not engaged for the basic conditions.
- ·No — if the net asset position is demonstrably below $6m even with aggressive values on every asset, and that calculation is written down in the file before the claim is made.
- ·Unsure — treat unsure as yes. The fee for a formal valuation is small against a failed concession claim; the cost asymmetry only points one way.
What the formal valuation actually has to do
A valuation for the MNAV test is a specific piece of work, not a generic business appraisal. It maps the group first — taxpayer, connected entities, affiliates — because the test applies to the aggregate, and getting the perimeter wrong invalidates everything inside it. It then values each CGT asset at market value: the willing-but-not-anxious principle from Spencer v Commonwealth, the market value basis in IVS 104, prepared under APES 225. For trading entities that usually means an earnings-based methodology to establish what the goodwill is worth, cross-checked against net assets; for passive holdings it means current market evidence per asset class; for private company shares and trust units it means valuing the interest itself, discounts and all. It nets off the liabilities and provisions the legislation allows and documents the reasoning at every boundary — related-party balances, personal-use assets, superannuation. We have written separately about the evidence file these claims need to survive review, so we will not repeat it here. The narrow point is this: once the framework above says yes, the engagement should be scoped to the test — which is exactly what our maximum net asset value valuation service is built for.
Where Prismi fits — and where we stop
Prismi prepares independent market valuations; we are not registered tax agents and nothing in this article is tax advice. Whether the concessions apply to your circumstances is your accountant's call, made on the valuation evidence. For concession matters we typically recommend the Defensible Valuation File (from $8,995 + GST), because these claims carry genuine review risk and the file needs to hold up if queried. Where the event date is historical, the retrospective surcharge ($495 per date) applies, and additional entities are $750 each — a real consideration under a test that aggregates connected entities. Fees are fixed at engagement and never contingent on outcome, the report is senior-reviewer signed under an independence statement, and the working file is retained for ten years. Two things we will not do. We will not predict or guarantee what the ATO decides — the report is prepared with the ATO's market valuation expectations in mind and documented so the position is defensible if reviewed, which is the most any honest valuer can offer. And we will not engineer a value under $6m. If the evidence supports a position above the threshold, the report will say so — and if a client asks us to conclude under it regardless, we will say so and decline the engagement on those terms.