Every so often a set of accounts arrives that reads less like a scorecard and more like a confession. This one covers a UK-headquartered group that did not exist three years ago, turned over more than half a billion pounds in its first reported period, and still managed to end the year worth less than it started. Nothing in it is improper. Everything in it is instructive. We have kept the name out of this piece — the lesson travels better without it.
A group built by stapling, not by growing
The holding company was incorporated in late 2023 and, within eight weeks, swallowed two separate business families through share-for-share exchanges. Because the same shareholders sat on both sides, the transaction was treated as a merger rather than a purchase, which means the accounts are presented as though the group had always existed. It is a perfectly legitimate accounting route, and it has one practical consequence worth holding on to: the prior-year comparatives are unaudited. The auditors said so plainly in a separate paragraph of their report. You are comparing an audited year against a year nobody signed off. What sits underneath is six operating businesses running semi-independently — heavy engineering and valve services, fluid transfer and hydrogen systems, an EPC and consultancy arm, a voltage optimisation and battery storage manufacturer, and a flow measurement software house. The strategic story is decarbonisation and diversification away from upstream oil and gas. The financial story is more mixed, and considerably more interesting. Growth is real. The margin is thin.
Turnover came in at £553.8m against £500.1m, up 10.7%. Strip out the fact that this was a 53-week period against a 52-week comparative and underlying growth is closer to 8.6% — still respectable in an industrial services market.
Below that, the picture tightens fast. Gross margin slipped from 27.0% to 26.5%. Operating profit was £25.8m, an operating margin of 4.7%. EBITDA reached £42.2m, or 7.6% of sales. For a group with 3,226 employees, revenue per head is around £172,000 and staff costs alone absorb 32% of turnover. This is a labour-heavy, low-margin machine that converts scale into cash but leaves very little room between a good year and a bad one. Two percentage points of gross margin is the difference between comfortable and uncomfortable.
There is also a geographic mix shift that deserves a raised eyebrow. UK revenue jumped from £287.4m to £391.8m — up 36% — while rest-of-world revenue fell from £164.3m to £119.9m, a 27% decline. Either the group has rapidly re-domesticated its earnings base, or the way revenue is attributed has changed. Both explanations matter, and neither is disclosed.
The tax line is the tell
Profit before tax was £18.6m. The tax charge was £10.2m. That is an effective rate of 54.8% against a headline UK rate of 25%.
Profit after tax landed at £8.4m, down from £13.4m. The reported earnings actually attributable to the owners were £9.5m, because minority shareholders absorbed a £1.1m loss. Eighty-five per cent of the net worth is goodwill
Now to the balance sheet, where the real story lives.
Net assets stood at £120.5m. Of that, intangible assets account for £112.9m, and goodwill alone is £102.8m. Deduct every intangible and the group's net tangible asset base is roughly £7.5m — on half a billion pounds of turnover. This is what a business assembled through seven acquisitions looks like from the inside: the value is in the premium paid for other people's companies, amortised over twenty years, and tested annually against a forecast.
That test bit this year. A £4.0m impairment was booked against one subsidiary's goodwill, taking its carrying value down to £5.7m. The disclosure around it is refreshingly candid and quietly alarming: if unsecured contracts were stripped out of that unit's cash-flow forecast, the remaining goodwill would be written off entirely, and a mere half-point rise in the discount rate would cost another £0.9m. In other words, the value of that asset rests on work not yet won.
The sequencing makes it sharper still. In October 2024 the group issued £5.5m of shares to buy out the 32% minority in that same business. Weeks later, at year end, it impaired £4.0m of its goodwill. Paying up for full ownership of an asset you are simultaneously writing down is a decision that deserves an explanation the accounts do not give.
Elsewhere on the balance sheet, stock of £77.6m is stated net of provisions of £29.5m — meaning more than a quarter of gross inventory is carried at nil or reduced value. Debtor days sit around 56 and stock days around 71, both reasonable. Cash of £45.8m is flattered by £11.9m of overdrafts sitting on the other side.
A £40m dividend out of an £8.4m profit
The group generated £36.8m of net cash from operations. It then paid a £40.0m dividend. Distributions exceeded both the year's profit and the year's operating cash flow. The profit and loss reserve fell from £55.0m to £14.7m, and total net assets dropped by £31.1m. The dividend was funded, in substance, by the refinancing completed the same month. Look at how the operating cash was generated and it becomes more delicate again. Debtors rose by £26.8m and stock by £7.4m — a £34.2m drag — offset almost exactly by a £34.3m increase in creditors. The cash conversion is genuine, but it is being carried by the payables line. Suppliers are, quietly, part of the funding structure.
The price of the money
In October 2024 the group refinanced into a £100m package: a £20m amortising term loan at SONIA plus 4.00%, an £80m bullet loan repayable in full in October 2030 at SONIA plus 4.50%, and a £50m revolving facility, undrawn at year end. Arrangement and professional fees came to £6.3m. Every share in the company and its material trading entities has been pledged as security.
Margins of 400 to 450 basis points over SONIA are not investment-grade pricing. They are sponsor pricing, and the accounts confirm why: the ultimate controlling party is a Delaware limited partnership — a private equity fund. Net debt closed at £60.0m, roughly 1.4 times EBITDA, with interest cover around 4.2 times. Those headline ratios look manageable. They also exclude the property commitments below, and covenants are tested quarterly on cash-flow cover and adjusted leverage.
The lease that does not appear as debt
In March 2023 two subsidiaries sold properties with a fair value of £15.2m for £40.3m and leased them straight back for 25 years, with options to extend a further 15. The sale price was 2.65 times fair value. The £25.1m surplus is being released to profit over 25 years, with £24.4m still parked in creditors.
Nobody pays 2.65 times fair value out of generosity. The premium was recovered through rent — annual reviews indexed to inflation with a floor of 2% and a cap of 4%, so the rent can never fall. Because the arrangement was classified as an operating lease under UK GAAP, none of it appears as a liability. Total non-cancellable lease commitments are £147.6m, of which £109.7m falls beyond five years. Capitalise that and the leverage picture changes materially. This is the single most important number in the accounts that does not appear on the balance sheet.
Everything points to an exit
Read the small print and the direction of travel is unmistakable. Share options — 4.29 million of them at a weighted average exercise price of £3.72 — become exercisable only on a sale or listing of the group. Management is paid to reach a transaction, not a dividend. Acquisitions accelerated after the year end: a US controls business for $13.25m in April 2025, an Australian group in June 2025 funded by shares with AUD 32.0m of its debt repaid using cash and the revolving facility, plus two smaller trade-and-asset deals. The revolver that was undrawn at year end is now working. Meanwhile the accounts themselves were signed nine months after the period closed and filed nearly twelve months after it — late, for a group of this size with a Big Four auditor.
Add the board churn, the buy-out of minorities, the recapitalisation, and the £40m return of capital, and the shape becomes clear: this is a platform being tidied, levered and scaled for sale.
What we would want to know next
Six questions decide the answer, and none of them can be settled from the filing alone: which sub-group actually earns the £42.2m of EBITDA and which ones dilute it; whether the geographic revenue swing is commercial or presentational; what the covenant headroom really is once property commitments are capitalised; whether the impaired unit's pipeline converts; how sustainable the payables-funded working capital position is; and what the bullet repayment in October 2030 assumes about exit timing.
That is the difference between reading accounts and analysing them. A group can grow revenue by 10%, generate £42m of EBITDA, distribute £40m to shareholders, and still be worth £31m less at the end of the year than at the start — with 85% of what remains resting on goodwill and a forecast.
Thank you.