Good evening, reader.

In today's Innovation Commons: the founders are furious and they are justified. We look at the budget measures not being picked up by the media and make a proposal.

Let's get into it.

The Anger

The reaction from the technology sector to the 2026–27 Budget has been the most cohesive industry response to a tax change in a decade. Paul Bassat, co-founder of Seek and Square Peg Capital, told the ABC the proposal was "a hammer blow to the people building the country's next generation of companies" and warned it would push Australian founders to incorporate in Singapore and Delaware (ABC News, 8 May 2026). Scott Farquhar described the changes as "a tax on the willingness to start something hard" (Forbes Australia, 13 May 2026). Daniel Petre called it "the worst signal an Australian government has sent to the venture industry since the 2014 ESS debacle" (Forbes Australia, 13 May 2026).

The numbers behind the anger are not soft. On the Forbes Australia analysis, a successful founder exit moves from an effective tax rate of roughly 23.5 per cent to approximately 46 to 47 per cent, close to a doubling (Forbes Australia, 13 May 2026). Baker McKenzie, PwC and Pitcher Partners have all flagged the same technical mechanism: founders typically hold equity with a nominal or zero cost base, which means cost-base indexation (the instrument the Budget reintroduces) does nothing for them. For a founder whose company is worth $100 million on exit and whose cost base is approximately $1, indexation lifts the cost base to roughly $1.40 (PwC, 13 May 2026; Baker McKenzie, 13 May 2026; Pitcher Partners, 13 May 2026). The pre-1999 logic that indexation is the "fair" treatment of capital gains relies on a non-trivial cost base. Founder equity is the limit case where the logic breaks.

The broken promise compounds the technical injury. The Government campaigned in 2025 with an explicit commitment that the 50 per cent CGT discount and negative gearing arrangements would not change (ABC News, 12 May 2026). Across the Atlassian, Canva, SafetyCulture and Square Peg founder cohort, a meaningful share of capital commitment decisions, ESS grant designs, and personal exit timelines were made in reliance on that commitment. The argument that you cannot run an innovation economy on a tax regime that changes inside the first year of a parliamentary term is a serious argument, and it is being made in good faith.

The sharper problem is that the capital being redirected is post-tax income.

A founder, angel, or operator who reinvests in another venture is deploying money that has already passed through PAYG at marginal rates of up to 47 per cent. The 50 per cent discount, whatever else it was, functioned as a partial correction to the double-taxation arithmetic on that recycled capital.

Removing it without a targeted replacement leaves the recycler facing roughly 47 per cent on the wage, indexation-plus-30-per-cent-minimum on the eventual gain, and effective rates close to the marginal income rate on the realised return. The arithmetic of recycling private capital into the next venture changes from "keep three-quarters" to "keep just over half" (Forbes Australia, 13 May 2026).

The chilling effect is not on the headline founder exit. It is on the second, third, and fourth ventures that the same capital was statistically going to fund. Australia's angel and seed ecosystems run on this recycling. A tax change that reduces the after-tax return on recycled capital reduces the pool that the next cohort of founders draws from. That harm is structural and slow-acting and very hard to undo.

The second-order effect is uncertainty itself, which behaves as a tax. A capital-gains regime that moved within twelve months of an election the Government won on a no-change commitment widens that variance for every founder, every VC committee, and every family office board looking at a five-to-ten-year hold. The investor question shifts from "what is the tax" to "what will the tax be when I exit". This is priced into deal terms as a discount on valuations, into LP commitments as a reduced allocation to Australian funds, and into capex decisions as a deferral. Succession is the limit case of the same problem.

A founder of a 20-year-old advanced-manufacturing or agritech business considering a generational handover now faces a transition window in which the tax treatment of the transaction is materially different depending on whether settlement closes before or after 1 July 2027.

Below the small business CGT concession thresholds the four concessions are preserved and the transaction is largely insulated (Treasury factsheet, 12 May 2026). Above the $6 million net asset value cap and the $2 million aggregated turnover cap, the founder is in the new regime, with no rollover yet drafted, no consultation outcome published, and a transitional window measured in months. That is the opposite of the policy intent.

This the anger. We take it seriously. We are also going to look at what the instrument the anger defends was actually doing.

The Proposal

The Treasurer's 2026–27 Budget proposes, from 1 July 2027, to replace the 50 per cent CGT discount with cost-base indexation, impose a 30 per cent minimum tax on real capital gains, limit negative gearing for residential property to new builds, and apply a 30 per cent minimum tax to discretionary trust distributions (Budget Paper No. 1, Statement 4, p. 146–158; Treasury factsheet, 12 May 2026). The reform is forecast to raise $3.6 billion across the forward estimates and $77.2 billion across the decade, with the negative gearing and CGT components contributing slightly over $40 billion of the latter and the trust changes contributing the rest (ABC News, 12 May 2026; Chalmers, Insiders, 17 May 2026).

Sitting alongside it, in the same Statement 4, is a second package of measures the country is barely talking about: a permanent two-year loss carry-back for companies with up to $1 billion in turnover (BP1, p. 168), loss refundability for small start-up companies in their first two years (BP1, p. 169), R&DTI core-rate increases of roughly 25 to 50 per cent (BP1, p. 174), ESVCLP and VCLP cap uplifts of between 35 and 110 per cent depending on the parameter (BP1, p. 171), and a permanent $20,000 instant asset write-off for businesses with up to $10 million in turnover (BP1, p. 177). The four small business CGT concessions are untouched, as is the 60 per cent CGT discount on qualifying affordable housing and superannuation's one-third discount (Treasury factsheet, 12 May 2026; Pitcher Partners, 13 May 2026).

A government that solely wanted to punish builders would not be expanding their instruments while removing a discount that mostly accrued to other people. The reform is doing something more. It is separating two policy domains that have been fused for a generation.

The Disguise

A policy instrument's identity is its causal effect, not its stated rationale. The discount became the instrument it became.

The 50 per cent CGT discount was introduced in 1999 following the Ralph Review of Business Taxation. The Review's justification was that a more generous treatment of capital gains would "encourage investment in the Australian share market, spur investors to buy and sell assets more frequently, and increase tax revenues" (Review of Business Taxation, A Tax System Redesigned, 1999, cited at BP1 p. 146). Before 1999, capital gains were indexed for inflation, the same instrument the 2026 Budget now reintroduces. The discount was never a builder's policy. It was a share-market participation policy, justified on grounds of churn, liquidity and revenue.

What actually happened over the following 25 years was the opposite of the Ralph rationale. The share of tax filers receiving dividend income fell, while the share with rental income grew substantially (BP1, Chart 4.7, p. 147). Treasury data on 1.5 million Australian investment-property transactions between 2008 and 2025 shows that detached houses were "overcompensated" for inflation by a considerable margin under the flat discount, while units and medium-density housing were undercompensated (BP1, Box 4.2, p. 148). The instrument designed to broaden share ownership ended up subsidising leveraged property speculation against the very housing types (units, medium density, regional) that the country needs more of.

A second mechanism amplified the distortion. Treasury's lifecycle modelling on 2022–23 sales shows that almost one in three negatively-geared property investments delivered the investor an effective subsidy: the tax benefit from rental loss deductions exceeded the eventual CGT paid on sale, despite the property turning a nominal profit (BP1, Chart 4.12, p. 152). The combination of negative gearing, the 50 per cent discount, and the option to time realisation in a low-marginal-rate year made established housing the most concessionally taxed asset class in the Australian system, alongside concessional superannuation. The average marginal rate paid on net capital gains income between 2009–10 and 2022–23 was approximately 25 per cent, lower than the rate paid by a full-time minimum-wage earner (BP1, p. 150).

Under-Discussed New Switches

Once the property-and-timing function of the discount is set aside, the question becomes whether the surrounding package supports entrepreneurial risk-taking. On the evidence in Statement 4, it does a lot of lifting that’s been missed in the messy middle of entrepreneurship.

Loss carry-back, permanent and broad. From 1 July 2026, all companies with turnover up to $1 billion can carry losses back two years to generate a refund of previously paid tax. Treasury expects this to benefit up to 85,000 companies a year and to cost $2.3 billion across the forward estimates (BP1, p. 168). When the temporary version of this measure ran during the pandemic, more than 70,000 companies used it, with an average benefit of roughly $50,000 each (BP1, Box 4.6, p. 169). Loss carry-back is the cleanest mechanism the corporate tax system has for blunting the asymmetric penalty on risk: profits taxed today, losses worthless for years.

Loss refundability for start-ups. From 1 July 2028, small start-up companies in their first two years can convert losses into a cash refund capped at the value of employment-related taxes paid. Treasury costs the measure at $410 million over five years and estimates up to 25,000 new small companies will be eligible annually, half of which will be able to fully offset their losses (BP1, p. 169).

The mechanism is unusually well-designed: it pays the company precisely for hiring Australians, when the company most needs cash, in the window the e61 Institute identifies as decisive for long-term firm trajectory (e61, The Young and the Restless, 2025, cited at BP1 p. 169).

R&DTI uplift. The core-activity offset rates rise by 4.5 percentage points across every band, which translates to a 25 to 50 per cent uplift in the effective subsidy per dollar of qualifying R&D. The refundable offset for young firms with turnover below $50 million rises from 18.5 per cent to 23 per cent.

The intensity threshold for the premium non-refundable offset drops from 2 per cent to 1.5 per cent. Maximum eligible R&D expenditure rises from $150 million to $200 million (BP1, Table 4.2, p. 174).

These rate increases are paid for by removing eligibility for "supporting activities" (literature reviews, equipment maintenance) which Treasury commissioned analysis found generated no additional R&D (BP1, p. 173). The trade-off is more reward for experimentation, less reward for structuring.

ESVCLP and VCLP cap uplifts. VCLP eligibility expands to investee businesses with assets up to $480 million (from $250 million). ESVCLP eligibility expands to $80 million (from $50 million), with the maximum committed capital rising to $270 million (from $200 million) and the asset cap at which investors retain full incentives lifting to $420 million (from $250 million) (BP1, p. 171). These are the first inflation adjustments to caps last set in 2002 and 2007. The superannuation performance test will also be reviewed to remove unintended barriers to venture-capital allocation across Australia's $4.5 trillion superannuation pool (BP1, p. 172).

Permanent $20,000 IAWO. From 1 July 2026, the $20,000 instant asset write-off is permanent for businesses with up to $10 million in turnover, covering 4.1 million businesses and costing $890 million across the forward estimates (BP1, p. 177). In 2023–24, 300,000 businesses claimed an average of $14,200 each under the temporary version (BP1, p. 177). Permanence is the variable that matters; it converts the IAWO from an ad-hoc gesture into a planning input.

This is the package the discount-anger has dragged out of the public conversation. It is, taken as a whole, a substantial reweighting of corporate tax toward early-stage risk capital and intangible investment, paid for by withdrawing a 1999 instrument whose dominant beneficiaries were leveraged property investors and pre-retirees timing share sales.

Why are the other changes missed?

Return now to the founders. The cost-base problem is real and the Forbes-circulated doubling of effective rates is not arithmetic anyone disputes. The point is that the instrument the discount-defenders are defending is not the instrument that was solving their problem. The discount delivered roughly 25 per cent effective rates on founder exits because the broad 50 per cent halving applies to anyone holding any CGT asset for twelve months.

The same instrument delivered the same 25 per cent effective rate to a Sydney landlord selling a negatively-geared two-bedroom unit and to a pre-retiree timing an ASX 200 share sale in a low-marginal-rate year. The country was paying roughly $40 billion across a decade for a broad property-and-timing subsidy and a small fraction of it reached founders. Defending the discount on founder grounds is asking the country to keep paying the full bill to preserve the by-product.

The Government has explicitly recognised the founder limit case. The Treasury factsheet released with the Budget contains the line: "Given the unique characteristics of the tech and start up sector the Government will consult on the interaction of the capital gains tax reforms and incentives for investment in early-stage and start-up businesses" (Treasury factsheet, 12 May 2026). The consultation is the policy choice the country actually has to make.

The Editor’s 2c:

The pre-existing scrip-for-scrip and small-business CGT rollover mechanisms already do most of the work. The 15-year exemption and the 50 per cent active asset reduction inside the small business CGT concessions already deliver substantially more relief than the discount for genuine operating founders below the $6 million net asset value threshold (Treasury factsheet, 12 May 2026; PwC, 13 May 2026).

The genuinely unaddressed segment is founders crossing the $6 million NAV threshold and exiting at $20 million to $1 billion, which is where the Atlassian, Canva and SafetyCulture cohorts sit. The right instrument is a founder-cohort rollover into qualifying productive assets: re-investment vehicles, ESVCLP units, qualifying R&D-active companies. The cost-base problem disappears because the gain is not crystallised.

Capital is then redirected into the ecosystem, not extracted.

— The Editor