Who Initiated the R&D? The Question Reshaping Manufacturer Tax Claims

For a meaningful slice of mid-market UK manufacturers, the annual R&D tax relief claim has, for the better part of a decade, been treated as something close toa routine cashflow item. The adviser visits, files are gathered, an interview or two is held, a figure is arrived at, and a credit lands. It has been one of the more reliable lines in the management accounts. In 2026, that comfortable rhythm is breaking.

A series of changes that have been quietly working their way through the system, the merged scheme, the new contracted-out rules, the overseas restriction, and a markedly tougher HMRC compliance stance, are about to land together in the first wave of claims being filed under the new regime. For manufacturers who have grown used to treating the relief as background income, the 2026 cycle is likely to be uncomfortable.

‍

Why is 2026 the first real test?

The merged R&D scheme came into force for accounting periods beginning on or after 1 April 2024. For a March year-end manufacturer, that means the year to 31 March 2025 was the first affected period, and the claim for that period is being prepared and filed during 2026. The architecture of the new scheme has been on the table for some time, but, in the way of these things, most businesses have not yet had a real claim cycle to test it against. The 2026 filings are the first time the new rules will actually meet the messy reality of how mid-market manufacturers run their innovation programmes.

The new framework is, on its surface, a simplification. The old split between the more generous SME scheme and the leaner RDEC scheme has gone, replaced by a single above-the-line credit. A separate, enhanced regime continues for loss-making, R&D-intensive SMEs. So far, so administrative. The real substance, and the part that is going to catch out a lot of claimants, lies in how the new rules define who is entitled to claim in the first place.

‍

So who actually initiated the R&D?

The single most consequential change for the manufacturing supply chain is the reversal of the contracted-out rule. Under the old SME regime, when a supplier carried outwork that involved genuine R&D, the supplier was generally able to claim the relief, even if the work had been commissioned by a customer. That fitted the way much of the British manufacturing supply base actually develops technology. An OEM asks for a lighter seat, a more efficient gearbox, a quieter pump, and the supplier does the engineering work to deliver it.

Under the merged scheme, the position is reversed. The party that initiated the R&D project, in other words the party that decided the work was necessary, is now generally entitled to claim, regardless of who carries out the actual development. For the example above, the OEM commissioning the lighter seat would, on the face of it, be the claimant, and the seating specialist who does the engineering would not.

The implications for tier one and tier two manufacturers are considerable. A specialist supplier whose annual R&D claim has been a meaningful contributor to its margin may find, in 2026, that its entitlement is significantly narrower than it was. The work has not changed. The cleverness of the engineering has not changed. The cashflow consequence is real.

What follows from this is largely a commercial and contractual question, not a tax one. The first defensive move for a specialist supplier is to look hard at how its customer contracts characterise the work. Where the supplier is genuinely the originator of the technology, who has identified a market need and developed a solution on its own initiative, that needs to be visible on the page. Where the customer initiates, but the supplier shoulders the technical risk and uncertainty, contracts can be drawn to share or transfer claim entitlement explicitly. None of this happens by default, and it is rarely a conversation that procurement teams have been used to having.

‍

The overseas dimension

Sitting alongside the contracted-out reversal is a quieter but equally important change: the restriction on overseas subcontracted expenditure. Subcontracted R&D costs incurred outside the UK no longer qualify, save for narrowly defined cases where it would be wholly unreasonable to do the work domestically. For a meaningful number of mid-market manufacturers who have, for understandable cost reasons, used design houses, prototyping facilities, or analytical labs in continental Europe, India, or elsewhere, the qualifying expenditure pool just shrank.

The right response is not to repatriate work that genuinely needs to be done overseas for technical reasons. It is to recognise, when budgeting and planning future development work, that the tax economics of overseas versus domestic subcontracting have shifted, and to take that into account before commitments are made rather than after.

‍

The HMRC reality

All of this is taking place against an HMRC posture that is markedly more sceptical than it was five years ago. After a long period in which incorrect or speculative R&D claims went largely unchallenged, HMRC has been running a sustained compliance push, with enquiry rates on R&D claims now running at levels that would have been unimaginable a decade ago. Advance notification of claims is now mandatory for many claimants, supporting documentation requirements have tightened, and the practical bar for what counts as a qualifying R&D activity has, in the way HMRC interprets the rules, risen.

For a manufacturer whose claim has been prepared on the basis of a fairly loose conversation with an external specialist each year, this is the moment to apply a more disciplined approach. The expectation now is that the claim will be supported by a paper trail that ties qualifying activity to the statutory definition, demonstrates the scientific or technological uncertainty being addressed, and explains why the work goes beyond what a competent professional could have routinely deduced. Claims that were waved through in 2020 will be tested in 2026.

‍

Is this actually a bad thing?

It is worth offering the counter-argument. The reforms of the last few years can be read, not unreasonably, as an attempt to direct R&D relief towards the work it was always meant to support, and away from the somewhat opportunistic claims that grew up around the old SME scheme. The clarification of who can claim contracted-out work is, in principle, a recognition that the party bearing the commercial risk of innovation, not the party executing the work, is the one whose decision-making the relief is meant to influence. The overseas restriction is a reasonable expression of an industrial policy preference. The compliance push corrects a market that, by HMRC's own assessment, was producing several billion pounds of incorrect claims a year.

A manufacturer whose innovation programme is genuine, well-evidenced, and structured around UK capability has, on balance, less to fear from the new regime than the way it has been talked about might suggest. The question is whether the way the business has historically prepared its claim reflects that reality, or whether it has been getting through on the relative leniency of the old regime.

‍

The work that needs doing now

The manufacturers who will come through the 2026 cycle best are the ones treating the claim, this year, as a project rather than a process. That means looking at live and forthcoming customer contracts to establish where claim rights actually sit, and, where they sit awkwardly, negotiating clarity before disputes can arise. It means mapping the active development projects in the business against the new definition of qualifying activity, with documentation built in from the start rather than retrofitted at year-end. It means accepting that the relationship with the external R&D adviser needs to be a more rigorous one, and that the days of a brief annual conversation producing a defensible claim are largely behind us.

Above all, it means recognising that an R&D claim that has historically been treated as a quiet finance team item is now a board-level question, because both its size and its risk profile have changed. The credit may still land. It is much less likely to land by accident.

‍

This article has been prepared for information purposes only. Formal professional advice is strongly recommended before making decisions on the topics discussed in this release. No responsibility for any loss to any person acting, or not acting, as a result of this release can be accepted by us, or any person affiliated with us.

‍