Financial Shenanigans

The Forensic Verdict

Kainos's FY2025 numbers are clean by the standards of the UK mid-cap tech sector, but FY25 was the year management quietly stretched the non-GAAP toolkit. Reported PBT fell 25% to £48.6m; the new "adjusted PBT" was £65.6m, a 35% premium versus 19% the year before, with the gap driven entirely by a first-time £8.4m restructuring add-back of which only £3.0m was actually paid in cash. Cash conversion of 112% therefore embeds a £5.4m unpaid restructuring accrual that management itself acknowledges would knock the metric to 103%. The balance sheet remains clean (£128m net cash, no debt, KPMG audit report unqualified with no emphasis of matter), receivables are not stretched, and CFO has covered net income by an average of 1.5x for three years — so the risk grade is Watch (25/100), not Elevated. The single data point that would change this grade is sustained reappearance of "non-recurring" restructuring or post-combination compensation in FY26 results: if it returns, the adjustment is no longer non-recurring and the underlying earnings trajectory is materially worse than the headline.

Shenanigans Scorecard

A category-by-category read of the 13-test framework. Severity is the strongest position the evidence supports, not a worst-case interpretation.

Breeding Ground

The conditions for accounting strain are mixed: governance is mostly textbook, but the executive chair is now reoccupied by the founder-CEO who built the company. That changes the incentive geometry.

Brendan Mooney founded Kainos, was CEO for 22 years to September 2023, stepped off the executive seat, then returned in December 2024 after Russell Sloan's eight-month tenure as acting CEO. The same FY in which the founder returned was the first in which "restructuring costs" appeared inside the adjusted-PBT bridge — not because management lied about them, but because the timing creates the textbook pattern of clearing the deck for a fresh start. The restructuring charge is small (£8.4m on £367m revenue) and management has quantified expected savings of £19m. The shenanigans risk is not the charge itself; it is the precedent of treating cyclical cost actions as non-recurring.

The breeding ground does not amplify the FY25 yellow flags into anything more severe. Audit, governance, and incentive structures are within UK FTSE 250 norms; what tilts the risk dial up is the combination of CEO transition with first-time restructuring add-back at a moment when the LTIP and bonus formulas were about to face headwinds.

Earnings Quality

Reported earnings reconcile to the cash-flow statement and to the balance sheet without strain. The yellow flag is non-GAAP, not GAAP.

Receivables ended FY25 at £38.5m, lower than FY24 and effectively the same as FY22 despite revenue having grown 21% over the same window. DSO at 39.9 days is within the 38-46 day range Kainos has run for four years. Revenue is not being pulled forward through receivables.

The adjusted-PBT bridge changed materially in FY25. SBC, acquisition amortisation, and acquisition fees are stable add-backs, but two new lines reshape the picture: the £8.4m restructuring add-back appears for the first time, and acquired-intangible amortisation drops from £4.2m to £0.8m as Blackline-related intangibles roll off. Net of these changes, the adjusted-to-GAAP gap widened from £12.4m (19% of GAAP PBT) to £17.0m (35% of GAAP PBT) in a year when GAAP PBT actually fell. The risk is taxonomic — once "restructuring" is in the adjusted definition, future cyclical cost actions become eligible candidates too.

Capex/D&A has run below 1.0x for three years (0.41x, 0.69x, 0.61x). For an asset-light services business this is not necessarily under-investment, but One Bankmore — the new Belfast HQ — is set to add £8m of capex in H2 FY26 and ~£20m in FY27, which will reverse the pattern. Software development is fully expensed (per CFO transcript H1 FY26: "we continue to expense product development to the P&L"). That is a clean choice; if Workday Products' contribution margin falls under sustained product investment, no capitalisation reservoir is hiding the impact.

Cash Flow Quality

CFO has covered net income comfortably for four years, but FY25's cash conversion is partially borrowed from FY26.

CFO/NI 3-yr average is 1.51x; FCF/NI 3-yr average is 1.41x; FCF after acquisitions to NI 3-yr is 1.23x. These ratios are robust, especially because Kainos has structurally negative working capital — Workday Products' SaaS-style billings are paid in advance, creating a deferred-income tailwind that repeats every year (FY24 deferred income added £8m in working capital).

The £11.8m working-capital tailwind in FY25 includes the £5.4m unpaid restructuring accrual. Strip it out and CFO falls to ~£53.4m, FCF/NI moves from 1.56x to ~1.41x, and the 3-yr FCF/NI to ~1.36x — still strong. The mechanism is disclosed and quantified by the CFO ("103% ex-restructuring"), but the H1 FY26 transcript confirms the give-back: H1 cash conversion fell to 48% versus a 75% target, "13% of which" management attributes to paying the restructuring accrual in April-May 2025.

Investing cash flow flipped positive in FY25 (+£7.9m) primarily because of the £6.2m investment-property disposal at fair value (no gain or loss recognised) and the absence of an acquisition. There is no boomerang capacity, no operating cost capitalised as investing outflow, and no factoring or securitisation disclosure. The CFO/CFI/CFF taxonomy is clean.

Metric Hygiene

The metrics management asks investors to focus on are mostly defensible, but the FY25 redefinition of adjusted PBT is the cleanest example of definition drift in the file.

The dividend payout ratio is the under-discussed signal in the FY25 file. On a GAAP basis, dividends declared exceeded earnings (100.5% payout), funded by the cash pile rather than current-year profit. On an adjusted basis the payout looks like 73%. Investors choosing between the two should know which one is asking the cash bucket to fund the gap.

What to Underwrite Next

The diligence list is short, specific, and quarterly-trackable.

Track in FY26 results

• Restructuring add-back recurrence. If management again classifies any cost-action charge as adjusting in FY26, the £8.4m FY25 charge stops being non-recurring. Watch the FY26 adjusted-PBT bridge line by line.
• Working-capital normalisation. H1 FY26 cash conversion of 48% confirms the FY25 timing benefit is unwinding. The full-year FY26 cash conversion print versus the 90% management target is the single cleanest forensic data point.
• Goodwill impairment in Workday Services. Workday Services revenue fell 12% in FY25 and Blackline procurement consulting was wound down. Goodwill at £37.3m (14% of assets) was not impaired. Monitor FY26 Workday Services CGU disclosure for any trigger event.
• One Bankmore capex through-line. Belfast HQ build adds £8m H2 FY26 and roughly £20m FY27. Watch the FCF after-capex versus management's "90% cash conversion" framing — FY27 will be the first year where capex/D&A meaningfully exceeds 1.0x.

Signals that would downgrade the grade

• A second consecutive year of "restructuring" inside adjusted PBT.
• DSO drift beyond 50 days, or receivables growing faster than revenue.
• A goodwill impairment that lands as an adjusting item rather than a one-time GAAP charge.
• Auditor change, partner change before the 5-year cycle, or any qualified or emphasis-of-matter audit opinion.

Signals that would upgrade the grade

• FY26 results in which restructuring drops out of the bridge entirely.
• Cash conversion at or above 90% on a normalised (ex-accrual) basis.
• Continued absence of related-party, factoring, or supplier-finance disclosure.

Position-sizing implication. This is a Watch-grade forensic profile, not a thesis breaker. The accounting risk does not justify a valuation haircut on its own — the balance sheet is genuinely fortress, the cash conversion is genuinely strong, and the auditor sign-off is genuinely clean. What it does justify is treating the FY25 adjusted PBT of £65.6m as the optimistic case, not the central case. A central-case earning power of roughly £55-58m PBT (GAAP plus the recurring elements of the bridge — SBC, acquisition fees, post-combination comp, but without restructuring) is the number to underwrite. The dividend payout already exceeds GAAP earnings; if FY26 earnings do not recover, the buyback or the dividend will need to give. That is the practical investor consequence of the year's accounting choices, and it is a footnote-level risk rather than a thesis-level one.