R&D tax relief •11 min read

Beer Express Ltd v HMRC

MJ

Matthew Jones ACA CTA

Last reviewed September 2026 · Editorial standards

Tribunal
First-tier Tribunal (Tax Chamber)
Decided
6 May 2026
Judge
Judge Jennifer Dean and Member Mohammed Farooq

HMRC won. A drinks wholesaler had claimed R&D tax credits of £301,721.51 and £190,857.70 across two years. The appeal was dismissed because the only witness put forward could not speak for the technical work, and the reports written by the company’s adviser were given no weight at all. The director was found truthful and was expressly not criticised. By the time HMRC asked for the money back, the director told the tribunal he had been unable to make contact with the firm that prepared the claims.

What was at issue

Beer Express Ltd was incorporated on 18 June 1998 and wholesales wine, beer, spirits and other alcoholic drinks. It owns one pub, the Penny Gill, and six small wholesale depots across the north of England (paragraph 63).

It claimed SME R&D relief on four projects across the accounting periods ended 30 June 2020 and 30 June 2021. They were a level-loading system using AI-enabled demand forecasting; a pale lager recipe; the rebranding and repackaging of two beer products; and a business-to-business web platform with logistics and stock-monitoring software. HMRC’s closure notices under paragraph 32 of Schedule 18 to the Finance Act 1998 assessed tax of £5,768.78 and £9,618.56 and rejected credits of £301,721.51 and £190,857.70 (paragraphs 1 and 2).

Both representatives confirmed the sole issue was whether the company had discharged the burden of proving it met the Guidelines tests (paragraph 3). HMRC did not dispute that the company was an SME or that it traded; it disputed Condition D in section 1044 — whether there was qualifying expenditure at all (paragraph 18). Quantum was severed by agreement shortly before the hearing, so the tribunal decided R&D in principle and left the costs to be determined later if necessary (paragraph 13).

How the dispute got to a hearing

Both claims were made by amendment, not on the original returns. HMRC received the amended return for the year ended 30 June 2020 on 18 November 2021 and the 2021 amendment on 8 May 2022. It opened compliance checks into both periods on 26 October 2022 (paragraph 6).

A pre-decision letter on 15 March 2023 said the projects did not meet the Guidelines. The agent responded on 18 April 2023; the closure explanation letter followed on 7 June 2023 and the closure notices on 8 June 2023, which the company accepted were validly issued (paragraphs 7 and 8). The company appealed to HMRC on 7 July 2023, and HMRC issued its view of the matter on 22 December 2023. The company accepted an independent review on 11 January 2024, and the review conclusion upholding both notices came on 12 September 2024 (paragraphs 9 and 10).

The notice of appeal to the tribunal was dated 26 September 2024. The appeal was heard at Alexandra House, Manchester on 8 April 2026 and released on 6 May 2026 — four and a half years after the first amended return.

What each side argued

For the company, it was argued that HMRC’s case leaned too heavily on formal identification of a competent professional, which the Guidelines do not require (paragraph 73). They demand no formal designation, no particular qualification and no expert witness; the question is whether the difficulties encountered would have been readily deducible to a competent professional working in the field (paragraph 75). The work had been done by people with practical experience in software systems, logistics integration and brewing (paragraph 76). The absence of an off-the-shelf solution was not determinative on its own, the company accepted, but showed challenges routine application of existing knowledge could not resolve (paragraph 77).

For HMRC, the argument was about evidence rather than science. The company had not met the burden because there was no cogent evidence from a competent professional and no clarity about what was achieved or sought (paragraph 79). On the level-loading work the evidence showed no technical difficulties and no explanation of how it went beyond existing technology (paragraph 80). On the lager, HMRC characterised the work as “trial and error in experimenting with new beer and processed of bottling” and said a third party’s testing kit did not meet the Guidelines (paragraph 82) — HMRC’s characterisation, and not one we adopt. On two projects it said the largest obstacle appeared to be “the bespoke nature sought by the Appellant”, which is not enough (paragraphs 81 and 84).

The provision the case turns on

Relief under the old SME scheme required qualifying Chapter 2 expenditure under section 1044(5) CTA 2009. As it stood for these periods, section 1052 made Condition B that the expenditure “is attributable to relevant research and development undertaken by the company itself”. That phrase runs through a chain. Section 1041 CTA 2009 takes the meaning from section 1138 CTA 2010, where research and development “means activities that fall to be treated as research and development in accordance with generally accepted accounting practice”. That is subject to regulations the Treasury may make under section 1006 of the Income Tax Act 2007, which section 1138 imports. Those regulations are SI 2004/712, and they point to the Guidelines.

The tribunal recorded at paragraph 30 that there was no dispute the Guidelines carry the force of law as tertiary legislation. They are not commentary to be weighed against the statute; they are where the test is written. The Guidelines use the term competent professional at paragraphs 13, 20 and 23 and define it nowhere (paragraph 45).

What the tribunal decided

The appeal was dismissed (paragraph 109) in four steps.

First, the witness. The tribunal found the company’s owner and founder “to be a truthful and credible witness who did his best to assist the Tribunal” (paragraph 89). But he “did not have any in-depth knowledge of the projects, and although he may have overseen the projects or had a vision for the Appellant’s future” (paragraph 89). He “had not been responsible for the design, creation, planning and implementation such that he satisfied the Guidelines as a competent professional” (paragraph 89). Overseeing work is not doing it.

Second, the standard. The tribunal agreed that “the Guidelines do not mandate that a competent professional must hold specific qualifications or expertise (and we did not understand the Respondent to argue otherwise)” (paragraph 91). It then adopted the formulation from Flame Tree Publishing — appropriate qualifications, experience and up-to-date knowledge of the relevant principles — together with AHK Recruitment on the need for evidence from someone contemporaneously involved (paragraphs 90 to 92). The requirement is not a certificate. It is a person who can say what the field knew and why the work went past it.

Third, the reports. The company’s case rested on two applications prepared by its adviser. Their author gave no evidence and could not speak to their contents, and the tribunal found them “vague and loosely termed” (paragraph 95). Its conclusion followed: “we concluded that the reports were no more than bald assertions unsupported by evidence from a competent professional with contemporaneous involvement in the projects or documentary evidence in support” (paragraph 99). That evidence was needed “to demonstrate what technology was available in the relevant periods and why the projects advanced that knowledge and technology and accordingly we attached no weight to the reports” (paragraph 99). The expenditure reports were in the name of an individual about whom there was “no clear evidence as to who this was, their level of involvement or whether they were a competent professional” (paragraph 101). The one person named as having relevant software expertise was a former shareholder who had since retired; he did not give evidence (paragraphs 70 and 96).

Fourth, the consequence. There was “no cogent evidence setting out the baseline of scientific or technological knowledge for any of the projects, the uncertainties faced or advances sought as against that baseline” (paragraph 102). The claims described product development without establishing why it went beyond standard development in the trade. Applying Hadee Engineering at paragraphs 214 and 217, the tribunal also found no record of a plan and no explanation of how the activities were designed to resolve the difficulties (paragraphs 107 and 108).

Underneath all of it is the adviser. The applications were submitted by RDT Active (paragraph 46). The refund was paid to the advisers and passed on after deduction of their fees. HMRC rejected the claims in June 2023 and asked for repayment; on his own evidence, the director tried to contact the firm and could not (paragraph 46). The tribunal recorded that account in its own hedged words — “the company appeared to have disappeared” (paragraph 67) — and made no finding that the firm had ceased to exist or done anything wrong. The tribunal made no criticism of the director (paragraphs 89 and 98). The liability stayed with the company.

What it changes for a claim being prepared now

  • Identify the person who can give evidence before the claim is filed. The test is contemporaneous responsibility for the work in the relevant field. Ask of each project: who designed it, are they still contactable, and could they describe the state of knowledge in their field at the time? If the answer is nobody, the claim is not ready.
  • Write the baseline down while the work runs. What sank this claim was the absence of any statement of what was publicly known or achievable. One dated paragraph per project, naming the existing approaches and their documented limitations, answers it; a report reconstructed years later by someone nobody can produce does not.
  • Treat a report as evidence, not as the claim. A document signed by a name with no established involvement carries no weight. Where an adviser drafts the technical account, the competent professional should review it, correct it, and be recorded as its source.
  • Consider what happens if the adviser is gone. A fee taken out of the credit leaves the company facing repayment of the gross credit out of the net sum it received. Choosing an R&D tax adviser sets out what to ask before appointing one; who counts as a competent professional takes the first point further.

This is an old-scheme decision — the periods ended in June 2020 and June 2021, so Chapter 2 of Part 13 CTA 2009 applied as it then stood. For accounting periods beginning on or after 1 April 2024 the current schemes apply instead: the merged scheme in Part 13 Chapter 1A, or Enhanced R&D Intensive Support in Chapter 2, with which one applies decided by the company’s R&D intensity. The definition of R&D does not change. It still runs through the Guidelines, and the competent-professional requirement in them is identical.

Where it sits against the other decisions

It is the clearest statement yet that the competent-professional test has two halves: the field the person is qualified in (Flame Tree Publishing), and the contemporaneous involvement that lets them speak to the work (AHK Recruitment). It applies both alongside Hadee Engineering on the need for a recorded plan. Judge Dean also decided Hadee, and the passages she quotes from it are the ones this register has cited since.

It sits against Get Onbord, where a competent professional without formal qualifications was accepted. The two are consistent: there, the person who built the system gave evidence about it; here nobody did.

M-Sport Wheels Ltd v HMRC [2026] UKFTT 977 (TC) decides no R&D law — it is a costs decision — but shows the same exposure. The company’s intended witness had changed employer and his new employer would not release a statement; the replacement statement, from a witness abroad, was never signed, and the appeal was withdrawn (M-Sport at paragraphs 8 to 15). HMRC’s application for costs was refused (M-Sport at paragraph 47).

Has it been appealed?

No onward appeal has been reported, and no Upper Tribunal decision naming Beer Express appears on the published record. Applications for permission to appeal are not consistently published, so the absence of a reported appeal is not proof that none was sought; the 56-day window ran from release on 6 May 2026.

One matter is expressly outstanding: quantum was severed by agreement at paragraph 13, so a further decision on the costs claimed may follow. Nothing further has been published as at this review.

The decision is First-tier Tribunal, so it binds only the parties and sets no precedent. Its weight is practical: it is the most recent statement of what a competent professional has to be able to do, and HMRC officers argue from it. HMRC has not changed its published guidance as a result — nothing in the CIRD change log or the published Research and Development Communication Forum minutes refers to it. The other decisions sit in the register of R&D case law.

Sources

  • Beer Express Ltd v HMRC — the decision, cited as [2026] UKFTT 672 (TC), case number TC 09873, appeal reference TC/2024/05221, heard 8 April 2026 and released 6 May 2026. Every paragraph number above refers to it.
  • Section 1044 as at 30 June 2021 — Condition D, the only condition in dispute; and section 1052 as at the same date, Condition B in the point-in-time text.
  • Section 1006, Income Tax Act 2007 — the definition under which the Guidelines are made.
  • CIRD81300 and CIRD80560 — HMRC on the definition of R&D and on the records it expects.
  • Flame Tree Publishing Ltd v HMRC — [2024] UKFTT 349 (TC), the formulation adopted at paragraph 91.
  • M-Sport Wheels Ltd v HMRC — [2026] UKFTT 977 (TC), a costs decision determined on the papers on 22 June 2026 and released 30 June 2026; the withdrawal chronology at paragraphs 8 to 15 and the refusal of costs at paragraph 47.
  • Gripple Ltd v HMRC [2010] EWHC 1609 (Ch) at [12] — the narrow construction. The decision prints this at paragraph 3 as “[2010] EWCH 1906 (Ch)”, wrong in both the court abbreviation and the number.

Every CIRD paragraph cited above is indexed on our CIRD reference index, with HMRC’s own title for each, a line on what it says and the date HMRC last revised it.

First-tier Tribunal decisions bind only the parties to them and set no precedent. This entry describes the decision as it stood at the review date above, as general information rather than advice on your circumstances — see our terms. For an answer on your own facts, talk to us.