Railpen manages around £36bn for more than 350,000 current and former railway workers. It is one of the driving forces behind a new £1bn-plus Scale-up Fund, backed by other pension funds and by government through the British Business Bank and the Office for Investment. Railpen's own account concedes that not every opportunity is equal, and that success is not simply a matter of allocating more capital.
The structure is not finished. That detail matters more than the money, because diligence standards are written during structuring and are effectively immovable afterwards.
The case
Northvolt raised more than $10bn in equity, debt and public financing after its founding in 2016. Volkswagen held 21 per cent, Goldman Sachs 19 per cent. Denmark's largest pension fund, ATP, held around 5 per cent, a stake of 2.3bn kroner, roughly $324m.
The technology was not the problem. The cells worked. What failed was the ramp at Skellefteå, where output could not be raised to contract. BMW cancelled a €2bn order in June 2024 after Northvolt failed to deliver. Chapter 11 followed in November, Swedish bankruptcy in March 2025. Goldman's private equity funds wrote off $896m in full. ATP's chief executive said the holding probably would not end far from zero.
Every institution on that cap table ran diligence. None of it caught the ramp.
A company can pass financial, commercial and legal due diligence and still be unable to make the thing at rate. In Northvolt's case the difference was paid for by Danish pension members.
The perimeter
Trustees cannot simply back their own judgement. Under section 36 of the Pensions Act 1995 they must obtain and consider proper advice on whether an investment is satisfactory. Practitioners call it section 36 advice, and failing to obtain it exposes individual trustees.
The statute then defines proper advice. It means a person qualified by ability in and practical experience of financial matters, authorised where giving the advice is a regulated activity.
The statute defines the adviser by financial competence. It is silent on whether anyone in the room has ever qualified a production process.
The regulator does not fill that gap. The FCA's perimeter is conduct, disclosure and financial suitability. It authorises advisers on financial competence and has no mechanism to verify the technical claims a private company makes to institutional investors. Private placements to professional investors sit largely outside the prospectus regime. An FCA-authorised adviser can lawfully certify an investment as satisfactory having never asked what the first-pass yield is.
Nor is the direction of travel towards more gates. Under section 25(3) of the Financial Services and Markets Act 2023 the FCA carries a secondary objective to facilitate UK competitiveness and growth. A regulator asked to facilitate is not one likely to invent a new technical test.
The pattern abroad
Britain is unusual here, not typical.
The United States, spending its own money, gates it. CHIPS awards ran through formal due diligence before final terms, with money released against construction and production milestones rather than paid up front. Samsung's award fell from $6.4bn to $4.7bn between the preliminary memorandum of terms and the final award, after Commerce completed its diligence. The Department of Defense has run Manufacturing Readiness Levels for two decades, precisely because Technology Readiness Levels kept clearing programmes that then died in production.
Elsewhere the competence sits inside the financing institution rather than beside it. The European Investment Bank appraises with its own sector engineers. Germany put technical intimacy into the credit relationship through the Hausbank and the regional savings banks. Japan did it through the main bank, Korea through ownership, Sweden through the industrial holding company.
The common feature is not culture. Somebody with production experience sat inside the institution deciding whether to release the money.
Britain had that and dismantled it. What replaced it was a financial instrument built for software, where being wrong costs a write-down rather than a factory.
The window
The question is not whether the scale-up fund will do technical diligence. It is whether manufacturability is named as a distinct category, or folded into a technical section written by people whose expertise is patents and prototypes.
Folded in, it disappears. Named, it gets a scope, a budget line and an independent voice. The difference is a paragraph in a document being drafted this autumn.
The turn
Somebody has to stand between the laboratory result and the production forecast and say, in writing, whether the second follows from the first. Not the fund manager, who is incentivised to deploy. Not the founder, who cannot be independent about their own process. Not the financial adviser, whose statutory qualification is financial.
Manufacturing Engineers qualify supply chains. They validate production processes. They run capability studies that turn a promising sample into a defensible yield curve. They design the feedback loops that turn first-of-a-kind data into nth-of-a-kind confidence. They put a number on the distance between what the technology can do and what the factory can do, and they sign it.
Kaipability works at this interface. The coupling gap is the measured distance between technology readiness and manufacturing readiness, assessed independently, with no advisory work taken during or after the engagement. If you are structuring a fund that will put patient capital into physical companies, that gap is the one your documentation should name.
If you are structuring a fund that will put patient capital into physical companies, the production question is where the conversation starts.
Start a conversation How we read an asset →Questions this dispatch answers.
Written to be quoted by AI assistants and search engines. Self-contained answers, verdict first.
