What the Budget Actually Did to R&D: What Founders Need to Do Before 30 June

The 2026-27 Federal Budget was handed down on 12 May. Treasurer Chalmers delivered the first formal government response to the Ambitious Australia SERD report, and the first major structural changes to the R&D Tax Incentive since the 2020-21 reforms took effect.

The headline measures are real. A lifted refundable threshold, higher offset rates, a bigger expenditure cap, new loss refundability for early-stage companies. Directionally right. They reward scale and mature R&D programs.

Underneath them, three changes narrow access at the early end of the program, the end where Australia's next generation of innovative companies are built. Supporting R&D activities are being scrapped. The minimum expenditure floor is more than doubling. Refundability is being capped at ten years.

"This Budget reshaped the R&D program around companies already winning, and quietly narrowed the door for the companies still trying to." — Alex Knight, Founder and CEO, Advanced

The Wins: What They Actually Mean

The refundable threshold lifted from $20M to $50M. This is the most useful structural change in the package. The $20 million cliff has distorted scaleup behaviour for years. Companies manage to the threshold instead of growing through it. Raising it to $50 million removes that distortion and meaningfully expands the band of scaleups that can access the refundable offset. A clean win.

Start-up loss refundability from 2028-29. Start-ups in their first two years of operation will be able to claim a refund on tax losses up to the value of FBT and PAYG withholding paid on employee wages. For a pre-revenue founder with payroll, that's real cash on real activity, mechanically separate from the RDTI and stacking alongside it. It doesn't require the AusIndustry registration process. It lands on early-stage cashflow from a different angle. Treasury got this one right.

Offset rates increased by 4.5 percentage points. Treasury announced this as a 25 to 50 per cent increase. The more honest description is a flat 4.5 percentage point lift applied to every existing rate. The headline range is wide because the same flat lift produces different relative increases depending on which rate a firm is already sitting on. The biggest relative winners are large firms on low-intensity R&D rates. Early-stage founders already on the highest refundable rate see the smallest relative gain.

The expenditure cap raised to $200 million. Pre-Budget speculation had this going to $250-300 million. The SERD report recommended removing it entirely. It moved from $150 million to $200 million. A step forward. Not the step the sector was expecting.

The Three Changes That Narrow the Door

Scrapping supporting R&D activities

This is the change that does the most damage to early-stage founders, and Treasury has presented it as simplification. It isn't.

Under current law, supporting R&D activities are defined as activities directly related to core R&D activities. You cannot have a supporting activity without a core one. The two are bound together in the legislation. Supporting activities are the scaffolding around the experimental work: data collection, prototype assembly, environment setup, protocol design, user testing. They're not the science. They're what makes the science possible.

Removing eligibility for supporting activities does three things, none of which are simplification.

It forces founders into harder, more contested self-assessment on every dollar. Strip out the supporting category and every cost has to be argued as core experimental work. That's the most heavily scrutinised category in the program. AusIndustry's reviewer attention sits there. Founders will spend more time defending fewer eligible dollars.

It shrinks the claim of every genuinely innovative early-stage company. We can provide further details of examples from innovation businesses in our portfolio. Reach out directly if you'd like to understand how this plays out in practice for a company similar to yours.

And it doesn't remove the categorisation work. It shifts it. Today the line sits between core, supporting, and ineligible. Tomorrow it sits between core and not-core. The administrative burden doesn't disappear. It relocates into the most contested part of the program, where the cost of getting it wrong is highest.

The minimum spend floor more than doubling

Pair the loss of supporting activities with a higher minimum spend threshold, and the early-stage narrowing becomes unmistakable. A founder doing legitimate experimental work in year one, but below the new floor, is told the program isn't for them. A founder doing modest work where part of it would historically have been supporting activity loses both the supporting portion and the safety margin in the same stroke.

These reforms don't operate in isolation. They compound. And they compound at the end of the program where Australia's most innovative early-stage companies are formed.

The ten-year refundability sunset

Refundability is being limited to companies operating for less than ten years. The framing is reasonable. Older companies should graduate to the non-refundable offset. The execution misses the cohort that matters most.

Many companies may not be fully commercial within ten years. Think of a medtech navigating a clinical pathway and regulatory approval, an agritech running extended field trials, or a climate hardware business working through pilot deployments. These companies are pre-revenue or low-revenue by the nature of their science and commercialisation, not by inattention.

The Budget already accepts this principle elsewhere. The higher offset for core R&D is explicitly framed around backing patient, longer-development work. The ten-year sunset contradicts that intent. A carve-out for longer-development industries, tied to industry classification using the same logic as the National Reconstruction Fund and Future Made in Australia priorities, would fix this without reopening the broader settings.

The CGT Contradiction

The Budget asked founders to do harder, longer work, and to keep less of the upside when it pays off. Those two signals don't add up.

The 50% CGT discount is being replaced with CPI indexation plus a 30% minimum tax on real gains, effective from 1 July 2027. For a founder who builds for a decade and exits in 2030, that's a material change to what they take home. Founder equity is taxed harder. Angel and early-stage investor returns are taxed harder. The economics of backing the harder, slower work got worse at the exact moment the R&D package told everyone to do more of it.

The government has acknowledged the tension. The Budget papers commit to consultation on how the CGT reforms interact with incentives for early-stage and startup investment. That consultation hasn't happened yet. Which means founders, angels, and early-stage investors are building cap tables right now without knowing what the exit will cost them.

"This Budget taxes a founder's equity, an angel's stake, and a passive investor's housing portfolio at the same higher rate. That isn't innovation policy. It's a blunt instrument." — Alex Knight

The small business CGT concessions remain in place. That's a meaningful protection for owners exiting operating businesses. But it doesn't resolve the question for angels and early-stage investors backing the harder, longer work.

For founders navigating this now, the decision between dilutive and non-dilutive capital looks different in this environment. Here's the framework.

What Needs to Happen Before the Legislation Lands

Three changes would close the gap between what this Budget says about innovation and what it actually does for the companies trying to build it.

Keep supporting R&D activities in the program. If the concern is overclaiming, fix it with better self-assessment tools, clearer guidance on the dominant-purpose test, and properly calibrated audit thresholds. Removing the category shrinks every legitimate claim and pushes complexity into the most contested part of the program.

Replace the ten-year refundability sunset with a carve-out for longer-development industries. Tie it to industry classification, using the same logic already in place for the National Reconstruction Fund and Future Made in Australia priorities. If a business is operating in a designated longer-development vertical, it retains refundability beyond year ten subject to continued R&D intensity. Small administrative change. Protects the cohort that matters most.

Preserve targeted CGT concessions for innovation assets before July 2027. A founder selling equity in a long-horizon innovation business should not pay the same higher rate as a passive investor selling an established housing portfolio. Preferential CGT treatment for gains on genuine innovation assets: founder equity, qualifying angel and early-stage investor holdings, and shares in qualifying businesses. This would close the contradiction the Budget created between its own chapters.

What Hasn't Changed: Your Current Year Claim

The Budget changes take effect from 1 July 2026. Your current year claim covers R&D spend from 1 July 2025 to 30 June 2026 and is governed by the existing rules in full. Your refundable offset rate is 43.5% if your aggregated turnover is under $20 million. Nothing about that changed last night.

The 30 April 2026 registration deadline for the year ended 30 June 2025 has passed. That window is closed regardless of the Budget. Your 30 June 2026 deadline for this financial year's eligible spend is eight weeks away.

The 30 June Window: What to Do Right Now

Your R&D spend from 1 July 2025 to 30 June 2026 determines your refund for this financial year. The more eligible spend before 30 June, the larger the refund in the pipeline. That refund arrives typically in October to November after your claim is lodged and the ATO processes it.

There's a second effect worth understanding. Access your anticipated refund early before 30 June and put that capital back into eligible R&D before the end of the financial year, that additional spend generates its own refundable offset. At 43.5% on $1 million of eligible R&D spend, early access of $348,000 could generate up to an additional $151,000 in offset, before financing costs and subject to eligibility.

On top of that, Payday Super starts 1 July. From that date, super must reach employee funds on every payday, not quarterly. July and August will be tighter than most businesses have modelled.

Eight weeks is enough time to act on this.

If you want to understand where your business sits before making any decisions, start with the runway calculation. Your R&D refund is a committed receivable that belongs in that number.

How to calculate your startup runway (and what to do when it's too short) →

‍

‍