Wales faces a structural deficit of between £14bn and £16bn each year, according to the latest official data from the Office for National Statistics. The gap between the £22bn to £24bn the Welsh economy generates in tax revenue and the £38bn to £40bn its public services require is not a temporary shortfall – it is a permanent, self-reinforcing constraint on the country’s ability to fund its own priorities. No adjustment to the Barnett formula, no grant programme, no restructuring of existing spending can close it. The only credible route, argues Frank Holmes, founding partner of Gambit Corporate Finance and chair of the Cardiff Capital Region’s Economic Growth Partnership, is to build the high-value, IP-led businesses that generate the tax revenues Welsh public services depend on.
The raw intellectual capacity for that shift already exists. Cardiff University, Swansea University and the Compound Semiconductor Applications Catapult produce world-class patents and breakthroughs. Welsh researchers file patents, publish in leading journals and turn out graduates of genuine quality. The problem, Holmes insists, is not the research – it is the translation. R&D productivity has been declining for decades not because discoveries are running out, but because the institutional infrastructure that converts research into commercial value is chronically underdeveloped. IP valuation, patient finance, commercialisation expertise and market access – the basic scaffolding of a modern innovation economy – are either absent or fragmented in Wales. The country invents things. It does not, systematically, build businesses from them.
This matters more now than at any point in economic history. As Jonathan Haskel and Stian Westlake document in Capitalism without Capital, investment in intangible assets – patents, algorithms, software, brands – has overtaken physical investment as the primary driver of prosperity in advanced economies. In the United Kingdom, intangible investment overtook tangible investment around the year 2000, and by 2022 it stood at £199.7bn compared with £167.3bn for tangible assets. Software is the largest capitalised intangible; firm-specific training the largest uncapitalised. Wales has largely missed this shift, and the fiscal consequences are now visible in every underperforming school and every NHS waiting list. Intangible assets have a characteristic physical ones lack: they can be used simultaneously and at scale without being depleted. A patent can generate royalty income from fifty licensees at once. A photonics architecture licensed globally creates wealth in Cardiff without requiring a factory in every market. But the same characteristics – scalability, spillovers, the fact that social returns exceed private returns – mean markets systematically under-invest in them. Private capital alone will not create the ecosystem Wales needs.
The scale of what is at stake is not theoretical. A joint study by the European Patent Office and the European Union Intellectual Property Office covering 2021–2023 found that IPR-intensive industries generated nearly half of EU GDP – approximately €7.7 trillion per year – while accounting for over 30% of all EU jobs, paying wages nearly 41% higher than the rest of the economy and representing over 78% of all EU exports. Over 88% of all private equity and venture capital invested in EU start-ups flowed into these industries alone. IP is not a niche policy concern. It is the primary structural determinant of whether an economy generates high wages or low ones.
Estonia proves the point. With a population of 1.3 million – less than half of Wales – it entered the 1990s with a Soviet-era agricultural economy and near-zero private sector. The government made a deliberate institutional decision to invest in digital infrastructure and technology education before the private sector had any reason to believe it was credible. GDP grew from $5.7bn in 2000 to $36.3bn in 2021. Estonia now produces more unicorns per capita than any other country in Europe – including Skype, Wise and Bolt. The model is not laissez-faire. It is deliberate, mission-oriented public investment at stages of risk that private capital will not absorb.
Mariana Mazzucato’s The Entrepreneurial State makes the case that the state has always been the most important risk-taker in innovation economies. The internet, GPS, mRNA vaccines and touchscreens – the technologies underpinning the modern economy – were developed through patient, mission-oriented public investment. “The state does not just fix markets,” Mazzucato writes. “It creates them. And when it takes on the risk of early-stage investment in transformative technologies, it should expect to share in the rewards.”
For Wales, Holmes argues, this means a genuine reorientation: not grants, which socialise cost and privatise benefit, but co-investment structured to generate returns. He proposes that the Development Bank of Wales take minority equity stakes in Welsh IP businesses alongside private capital on fully commercial terms, with a put option requiring companies that relocate outside Wales to buy back that stake at a premium – creating a genuine commercial disincentive to leave. Loan guarantees would make IP bankable in the Welsh market for the first time, removing the risk that has historically prevented commercial lenders from lending against intangible assets. Subsidised independent IP valuations would break the chicken-and-egg problem facing every early-stage Welsh technology company: they cannot afford the £40,000 assessment that would unlock the £3m they need to grow. An IP insurance pooling facility, operated through the Development Bank of Wales, would reduce the cost of insuring patent portfolios from prohibitive commercial rates to something accessible to smaller Welsh businesses, making IP-backed lending viable across the ecosystem rather than only for its largest members.
Enhanced Welsh R&D tax incentives, layered above the UK’s existing Patent Box and Enterprise Investment Scheme regimes, would make Welsh IP investment measurably more attractive than comparable investment elsewhere in the United Kingdom. A Welsh IP Quality Mark, issued by a Welsh Centre for IP Excellence and marketed at investor conferences in Tokyo, Singapore and San Francisco, would signal credibility to international capital that would otherwise look past Wales entirely. And strategic public procurement – the Welsh public sector spends over £8bn a year – could provide the first customer relationships that transform pre-revenue businesses into investable ones. Not subsidies, but commercial contracts for services the public sector actually needs.
Institutions matter as much as policies. Growth is a fragile achievement that depends on the credibility of the frameworks within which private capital operates. Private investors will not commit to Welsh IP businesses at scale if policy changes with every administration. Holmes argues that the Welsh Government’s role is to deploy capital and to provide the stable, legally-grounded institutional environment that makes private capital rational. That means embedding the framework in statute, not policy – making it durable across political cycles, not dependent on the survival of any single administration.
The cost of running the entire programme is estimated at £8–14m per year, including loan defaults, IP valuations and the insurance reserve. That is less than a medium-sized road improvement scheme and approximately 0.025% of the Welsh Government’s total budget. The potential return, even in a conservative scenario of scaling 5,000 Welsh businesses to £2m EBITDA over a decade, is approximately £4bn in additional annual tax across the UK economy. That would reduce Wales’s fiscal deficit in the national accounts and strengthen the case for a reformed fiscal settlement with Westminster – where the Barnett formula, which determines how UK government funding to devolved administrations is updated, has been criticised for not accounting for relative needs, with calls for a “Barnett floor” to prevent underfunding.
Alternative proposals are already on the table. Plaid Cymru has outlined plans to create more than 35,000 jobs by increasing public procurement from Welsh businesses, establishing a National Development Agency for Wales, reforming the Development Bank of Wales, promoting “Brand Wales” to attract investment, devolving the Crown Estate’s profits to Wales and reforming community benefit funds for renewable energy projects. These proposals incorporate principles of community wealth building, aiming to keep wealth circulating within Welsh communities.
Meanwhile, existing Welsh IP-led companies are already demonstrating the potential. In April 2026, IQE, a leading Welsh tech firm, secured an £81m investment package – including £30m from US semiconductor manufacturer MACOM Technology Solutions – to support its growth plans, repay debt and invest in core technologies such as Indium Phosphide and Gallium Nitride. The Compound Semiconductor Applications Catapult, based in Wales, is a global centre of excellence.
Frank Holmes, who has been voted “Insider Wales Dealmaker of the Year” multiple times and has secured over £2.3bn in finance for clients including Wales’s largest management buy-out at £110m, argues that the Welsh IP-led companies already exist – in photonics, compound semiconductors, life sciences and fintech. They are not waiting for grants. They are waiting for the valuation infrastructure, the patient capital, the insurance pooling, the procurement relationships and the commercialisation expertise that would make building and scaling in Wales the rational commercial choice. The gap between Wales’s research output and its commercial return remains the central barrier to closing the fiscal deficit.
