Observation #7: The Cliff at Ten Years
Australia's 2026 budget raised the tax on the capital startups rely on, then proposed a patch to reduce the damage with a concession that, by its own design, cannot fill the gap.
The sector is settling on a name for the place where Australian ventures die between a first cheque and a scaled one: The Missing Middle. The term describes the stretch where a company has proven enough to be expensive but not yet familiar, where family offices have stopped, and infrastructure investors have not yet started.
And the diagnosis is widely shared, but a more useful question is what the state has done to it, because the most consequential recent policy change does not bridge the gap. On the other hand raises the cost of equity funding for the ventures within it.
The 2026-27 budget removes the 50% capital gains discount, which comes into effect on 1 July 2027. The proposed replacement is a cost base indexation and a 30% minimum tax. In the aggregate, the change may have a defensible case.
A flat 50% discount was itself a distortion, and indexing so that only real gains are taxed is arguably more neutral. The incidence, though, is where it bites. It impacts the patient, illiquid equity that startups and their backers depend on, and hardest of all the founders and employees. This is because their equity is built through labour and time rather than cash, leaving a cost base close to zero. Indexation lifts a near-zero base by almost nothing, so a successful exit is taxed on close to the full nominal gain. The budget paired this with a few measures pointing the other way, higher caps on early-stage venture partnerships and a review of the superannuation performance test. Still, its centre of gravity raised the cost of exactly the capital that the conversion stage runs on.
But then came the backlash and the subsequent proposed patch. Facing discontent from founders and investors, the government proposed a carve-out to shield start-up equity from its own change, the Innovative Business CGT Concession. It is the instrument most directly aimed at the ventures the Missing Middle describes. It is also the one whose design most clearly misses them.
A cliff in the wrong place
As set out in the Treasury consultation paper released on 18 June, the concession would allow founders, employees and early-stage investors to retain the 50% discount rather than move to the new indexation and minimum-tax regime. Qualification requirements include that the equity must generally be new, issued on or after 1 July 2027, in a company under ten years old when the shares were issued, with turnover under 50 million dollars, held for at least five years, subject to a potential 10 million dollar lifetime cap and a principles-based innovation test.
A longer 15 year age limit is contemplated for sectors that take longer to commercialise, with biotech and medtech named directly and deep tech raised as a possibility whose definition practitioners already call unresolved and contested.
Hold those conditions against the ventures they are meant to serve, and the patch faces the wrong way.
The age test is measured at issuance, not at sale, so there is no deadline to exit within a decade. The problem is subtler and more consequential as fewer ventures reach an exit within ten years than the design assumes, and not only in hardware. Cut Through Venture’s data puts the median age at Series B at close to ten years, up from around five in 2021, so many companies are still raising growth capital, let alone reaching a liquidity event, well after the concession’s clock has run out. Software firms increasingly stay private for more than a decade, too, and a capital-intensive business that has to be built rather than coded takes even longer. Tie the concession to a company’s youth at the moment equity is issued, and every capital-intensive round raised after year ten falls outside it, benefiting only the earlier, cheaper rounds while forgoing the later, larger ones.
The concession becomes back-loaded to the wrong end of the venture. It rewards the equity that was, some may argue, the easiest to fund and withholds relief from the equity that is the hardest to raise, which is the equity in the middle.
There is also a five-year holding rule, which compounds the effect by penalising interim liquidity. A founder or investor taking a partial exit before the mark loses the concession on those shares. An instrument meant to reward patience is built on a clock that, paired with the age cliff, risks nudging behaviour toward earlier, faster plays and rushed exit windows, if there are exits at all, rather than the long, lumpy commitment the conversion stage requires. This is a design risk to flag while the consultation is open, not a settled outcome, and industry submissions are already pressing the related point that the carve-out, as drafted, may exclude existing founders, with only transitional rules for earlier shares.
The shape of the incentive is visible now, and it is calibrated for the next cohort’s first equity, not for carrying an existing venture across the gap.
The eligibility gate, like any, carries structural risk. A principles-based test defines a target, and a defined target invites structuring toward the definition rather than toward the innovation it represents, so the concession may select for the companies best able to qualify rather than those actually in the middle.
What a case shows
Goterra makes the cliff concrete. The Canberra agtech, which used black soldier fly larvae to turn commercial food waste into feed and fertiliser, entered voluntary administration in early June 2026, having failed to secure the capital it needed to scale. It attributed the outcome to a funding shortfall rather than product or market failure.
Adjudicating that while the administration is in flight is not the task here.
The narrower point is what a concession like this could ever have offered a venture in its position. Records show Goterra was registered in June 2016, so it was almost exactly ten years old at administration, sitting on the very boundary of the concession built to help ventures like it. As agtech and hardware rather than biotech or medtech, whether it would even reach a contemplated fifteen-year extension is undetermined.
And the deeper point is what a capital gains concession is, and is not. It works on the after-tax return an investor keeps years after an exit, not on the capital a venture raises this quarter. Even a well-formed version could not have financed Goterra’s round, because that was a financing problem, and the concession is a returns instrument. This version compounds the mismatch with an age notch and a new-equity limit that would likely have excluded a decade-old company anyway.
One objection runs the other way, perhaps Goterra was not a fundable business, and the market was right to withhold. Fundable or not, the concession cannot tell the difference. A returns instrument with a ten-year notch does not sort a good bet from a bad one; it excludes both. The cliff is a property of the design, not a verdict on the company.
Where the gap actually is
Place the concession beside the other measures, the caps and the superannuation review, and the pattern is the same. The venture capital caps work on the shape of capital. The concession works on its price. The superannuation review works on the benchmark that governs the largest pool. Two of them genuinely touch the question of fit. Neither touches the two conditions on which a conversion-stage raise actually turns.
The first is whether the capital present can read a first-of-a-kind venture as real. Investors are selected and incentivised to recognise shapes they have already watched succeed, and a genuinely novel, capital-heavy company presents no such shape at the moment it needs the largest cheque. That is a general problem, but a thin market makes it acute. With few funds pricing few theses, an unfamiliar bet has almost no one positioned to underwrite it. The second is whether that capital can move at the speed of a venture’s runway. Willingness that takes the better part of a year to diligence an unfamiliar hardware business arrives after the business is gone. These are matters of recognition and speed, of legibility and process, and price touches neither. A concession makes a vehicle more attractive to hold. It does not teach thin domestic capital to price an unfamiliar venture, or to close before the runway ends.
The independent data points the same way, though not by hollowing out the middle. Funding tracked by Cut Through Venture through 2025 and into 2026 shows the split at the two ends: the smallest rounds, under five million dollars, fell to their lowest level since 2020, while the top ten deals took close to 60% of all capital, the most concentrated the market has been in years. The 2026 report from Side Stage Ventures and Dealroom fills in the early-stage picture, with 18 Australian funds completing 5 or more seed deals a year, compared with 598 in the United States and 516 in Europe, and 41% of smaller raises now coming from offshore.
There is a puzzle in that data. The same report, on its own figures, shows Australia producing more billion-dollar companies per dollar invested than almost anywhere else, with strong venture returns. High output per dollar is not what a failing market looks like. But those returns are earned on a thin base, a small domestic capital pool sitting on outsized output it has not built the funds to finance. That is the more detailed reading of the gap. It is not that the signal cannot be read anywhere, because offshore investors are reading it and supplying a large and rising share of early-stage dollars. It is that domestic capital is too thin and too concentrated on the familiar to price an unfamiliar, capital-heavy bet at scale. The illegibility is local, not universal, and the foreign inflow is the market beginning to correct it rather than proving it cannot be corrected. The aggregate capital base is not small either, since Australia holds one of the world’s largest pools of long-duration capital in superannuation. What is thin is the slice of it organised into funds that price early, unfamiliar bets.
None of this will settle every case. From the outside, a real signal that domestic capital cannot yet price looks identical to a weak one it has priced correctly and declined, and some ventures die because the market read them right. The position here is the narrower one; the signal is real, the domestic market is too thin to fund it at the speed and scale the conversion stage needs, and the instruments on the table do nothing to change that.
The private market has begun to form the missing tier, with new patient-capital vehicles raised specifically to fill the gap. That is a positive response in principle, and it also carries the timing problem in miniature, because a fund still raising cannot catch a venture already falling. The tier the economy needs is forming too slowly to help the ventures whose failures are proving its necessity.
The unfinished work
The state has not overlooked the gap; it raised the cost of the capital that would carry ventures across it, then reached for a patch, and the patch works on the price of capital when the gap is made of something else. It is made of capability, legibility and speed, whether domestic capital can recognise an unfamiliar signal as real, and act on it before the runway ends. Those are harder to legislate than a threshold, which is presumably why the response has clustered where it has. The Missing Middle is not, at its core, a shortage of money. It is the point in a venture’s life where a real signal is least legible to the domestic capital around it, and no concession yet drafted changes what that capital can see.
The consultation on the concession closes on 10 July. Sharpening its age and holding conditions would make it less perverse, and might be worth doing. But it would not reach the gap, because the gap was never priced. It was unread.
And on current settings, the conversion tier is increasingly financed from offshore, with roughly two in five early-stage dollars now coming from foreign investors, about double the share in the US or Europe. Imported capital is not itself a loss, as capital mobility is how thin markets get funded. The cost sits in what tends to travel with it. When the money that carries a venture across the gap is foreign, the intellectual property, the operating talent, and the knowledge that compounds around a scaling company are more likely to compound offshore as well. Australia can keep producing world-class output per dollar and still capture a shrinking share of the value that output generates. That is the gap worth closing, and no concession yet drafted reaches it.
This essay is based on the author’s professional experience and interpretation of publicly available information. It is provided for general informational purposes only and does not constitute financial, tax or legal advice, and it references proposals under public consultation that may change. Views are the author’s own. Where organisations or proposals are named, the commentary is not a comment on the conduct of any individual or entity.
References & Further Reading
Prime Minister of Australia, “Tax reform implementation for small business and startups,” 18 June 2026. https://www.pm.gov.au/media/tax-reform-implementation-small-business-and-startups
Treasury, “Capital gains tax reforms – arrangements for innovative start-ups” (consultation, closes 10 July 2026). https://consult.treasury.gov.au/c2026-779186
business.gov.au, “Early Stage Venture Capital Limited Partnerships (ESVCLP)” (2026-27 Budget cap changes, from 1 July 2027). https://business.gov.au/grants-and-programs/early-stage-venture-capital-limited-partnerships
MinterEllison, “Proposed changes to the superannuation performance test” (Treasury consultation, 2026). https://www.minterellison.com/articles/proposed-changes-to-the-superannuation-performance-test
BDO, “Practical implications of CGT reforms for start-ups” (IBCC design, $10m lifetime cap, transitional rules). https://www.bdo.com.au/en-au/insights/tax/articles/australian-capital-gains-tax-(cgt)-reforms-for-start-ups-what-the-new-proposals-could-mean-in-pract
ABN Lookup, “Current details for ABN 97 612 974 688 (Goterra Pty Ltd),” registered 14 June 2016. https://abr.business.gov.au/ABN/View/97612974688
ASIC Published Notices, “Goterra Pty Ltd 612 974 688, Notice of appointment of administrators,” appointment date 3 June 2026. https://publishednotices.asic.gov.au/browsesearch-notices/notice-details/Goterra-Pty-Ltd-612974688/4987ea73-3df6-47b8-9d40-2376f6f8986b
Cut Through Venture, “Cut Through Quarterly, Q1 2026” (Australian startup funding report). https://www.cutthrough.com/insights/cut-through-quarterly-1q-2026
Side Stage Ventures with Dealroom, “Australia Venture & Startup Report 2026” (full report by registration). https://www.sidestage.vc/outliers-report-2026
Forbes Australia, “Australia’s VC sector is the fastest growing in the world, but one key pool is drying up,” June 2026. https://www.forbes.com.au/news/entrepreneurs/australian-vc-sector-is-the-fastest-growing-in-the-world-but-early-stage-funding-is-slowing/
Matthew Williams (AxleTree Capital), “Australia’s problem isn’t innovation, it’s the investment architecture for the ‘missing middle’,” Startup Daily (opinion), 18 March 2026. https://www.startupdaily.net/advice/opinion/australias-problem-isnt-innovation-its-the-investment-architecture-for-the-missing-middle/
Notes
Treasury Laws Amendment (Capital Gains Tax Reforms): Feedback on proposed CGT changes for innovative start-ups, including replacing the 50% CGT discount with cost-base indexation closes July 10th.
https://consult.treasury.gov.au/c2026-779186