Celebrus Technologies plc (CLBS) — Investment Research Note

Executive summary

Celebrus is a UK-listed (AIM) software vendor that captures, contextualises and activates first-party digital data — predominantly for financial-services customers — for marketing personalisation, digital-identity and fraud-prevention use cases. Across the period covered, the group has transitioned from being a hybrid hardware/services reseller into a more focused subscription-software company, but the headline trajectory has deteriorated sharply in FY26 (year ending 31 March 2026), with total revenue collapsing to $23.3m (FY25: $38.7m) and an adjusted loss of $0.2m vs prior-year profit of $8.7m as the company simultaneously absorbed an IFRS 15 accounting change to straight-line license recognition and missed on new business. The single most important point for valuation today is the balance sheet: with $32.0m of net cash against a £33m market cap, the market is paying close to zero for ~$15m of Celebrus ARR and a 93% gross-margin software platform that grew 10.3% even in a poor year.

Fair value estimate

• Methodology: sum-of-parts on EV/ARR plus net cash. The income statement is too distorted by the FY25→FY26 accounting transition to anchor on a P/E approach, and the cash mountain dominates the equity bridge.
• Inputs: FY26 closing cash $32.0m ≈ £25.6m (at $1.25/£). Celebrus ARR at FY26 close $15.0m ≈ £12.0m. Assumed EV/ARR range 1.5x (bear: continued execution failure, decelerating growth) to 4.0x (bull: pipeline converts, FY25 profitability restored) — reflecting that the underlying software business has 93% software gross margin and was profitable on a stable basis pre-accounting change 2025-07 final results; 2026-04 trading update.
• Fair value range: 108p – 184p per share, equivalent market cap range £43m – £74m.
• Mid-point: ~146p / ~£58m mcap.
• Versus current £33m mcap (~83p): absolute upside of ~76% at the midpoint, with a downside band that still sits above the current price thanks to the net-cash floor.

Sector context

• ICB classification confirmed: Technology / Software.
• Quality profile vs. peers: gross margin (93% software) is in line with high-quality vertical SaaS; growth (10% ARR) is below peers; balance sheet (net cash > 75% of mcap) is materially stronger than typical AIM software. Operating losses currently are below sector, but reflect a transition year rather than structural unprofitability.
• Listed peers (closest comparables): GB Group (GBG) on digital identity, Eckoh / Netcall in voice/CX adjacencies, Quantexa (private) as a partner-cum-competitor in fraud. Salesforce, Adobe and Pegasystems are listed competitors — Celebrus generally lands where they are deployed as a complementary first-party data layer.

Investment thesis

• Net cash dominates the equity bridge — you are paying ~£7m EV for the operating business. With $32m of cash, no debt, and a market cap of £33m, the enterprise value is roughly £7-8m, against $15m of Celebrus software ARR running at 93% gross margin. Even a modest re-rating to peer EV/ARR multiples implies meaningful upside 2026-04 trading update; 2025-12 half-year.
• Software economics are genuinely high-quality once the transition noise clears. Celebrus Software gross margin was 93.1% in H1 FY26, ARR grew 14.7% in H1 FY26 then 10.3% for the full year, NRR was 97.6%, and the company added wins including a US fintech and a European bank that displaced Adobe — evidence the platform is competitive 2025-12 half-year; 2025-07 final results.
• Operational leverage is real if pipeline converts. Operating expenses were $22.9m in FY25 against gross profit of $30.5m, and management has shown cost discipline (FY26 opex came down despite weaker revenue). If new-business execution improves, incremental ARR should drop disproportionately to profit given the fixed engineering/G&A cost base 2025-12 half-year; 2025-07 final results.

Key risks

• Sales execution failure has been chronic, not one-off. In April 2025 the company warned FY25 would miss consensus (revenue $43.4m → $38.6m delivered); in April 2026 it confirmed FY26 new business "under-delivered" with deals at contractual stage being lost or delayed at the line. That is the second consecutive year of new-business disappointment 2026-04 trading update; 2025-04 trading update.
• Customer concentration is material and renewal risk is visible. One channel partner accounted for $22.9m of $38.7m FY25 revenue (~59%). Two banking customers cut their Celebrus footprint at renewal in FY26 because of divestitures. Concentrated renewal cycles in a small-revenue base will continue to drive lumpy outcomes 2025-07 final results; 2026-04 trading update.
• Limited true AI-receiver exposure. Management says it has "embraced" AI but cannot point to an AI-driven revenue line or a measurably expanded TAM — most "AI" references are internal productivity and product features rather than picks-and-shovels capture of the AI buildout. Investors paying for AI exposure won't get it here 2025-12 half-year; 2025-07 final results.

Operating leverage

Celebrus has a classic high-fixed-cost software P&L: software gross margin 93.1% in H1 FY26, total operating expenses largely flat year-over-year ($22.9m FY25 vs ~$22m run-rate H1 FY26 annualised), and the cost base is dominated by R&D, customer success and a 151-person headcount that doesn't scale linearly with revenue. The implication: if Celebrus ARR were to grow by, say, $5m (one-third of current ARR), substantially all of that incremental revenue would drop to operating profit, plausibly more than doubling the FY25 adj. PBT of $8.7m. Conversely, recent commentary shows the leverage runs in reverse too — the FY26 revenue decline took adjusted PBT from $8.7m profit to a $0.2m loss with no meaningful cost reduction (only modest restructuring). So the operating leverage is genuine and high — the question is whether the company can generate the volume to capture it 2025-12 half-year; 2025-07 final results; 2026-04 trading update.

Value-trap signals

• Two consecutive years of new-business misses against management's own guidance.
• Net Revenue Retention slipping to 97.6% (below 100%) — existing-customer revenue is no longer growing organically.
• Largest customer is a channel partner representing ~60% of group revenue, with the underlying end-customers unknown to investors.
• Multiple revenue-recognition / segmentation / currency changes (USD reporting from FY25, straight-line license recognition from FY26, restated ARR definition) make trend analysis harder — not necessarily aggressive, but it does cloud the read.
• However: there is no rising-debt-with-no-growth pattern; cash has been broadly stable, dividends progressive, no going-concern flag. Not a textbook value trap, but value-trap risk is present.

Earnings vs. expectations

The track record over the period covered is mixed-to-poor at the headline but improving on profitability discipline. The April 2025 trading update flagged FY25 revenue at $38.6m vs. consensus $43.4m — a clear miss, though adjusted PBT came in above expectations at $8.7m vs. $8.5m on cost discipline. FY26 was broadly in line with the reset late-FY26 consensus ($23.3m revenue vs. consensus $24.1m; adj loss of $0.2m vs expected $0.3m loss) — but only because expectations had already been ratcheted down through the year. The pattern across recent results is: revenue misses against earlier-year guidance, with cost control delivering profit close to or slightly above the (lowered) bar. That is a low-quality "beat" pattern.

Conviction

Conviction: 3 (moderate).

• Anchors: net cash represents most of the market cap so the valuation downside is bounded; the software gross margin and ARR disclosure are clean enough to make a sum-of-parts call; the IFRS 15 accounting change creates a one-off P&L distortion that should normalise over a 3-year transition.
• Limits: the underlying ARR growth rate is decelerating and FY26 actually showed slight H2 contraction in Celebrus ARR ($15.6m H1 → $15.0m year-end); the path back to FY25 profitability ($8.7m adj PBT) is not visible in the disclosed pipeline; and the heavy customer/partner concentration means a single contract change can swing the thesis. A different methodology (e.g. forward P/E on recovered earnings) would give a wider range with weaker support.

Driver scoring

Overall score: 460/1000

A partial fit. The valuation is genuinely cheap (the entire operating business is being given away for free relative to net cash), the balance sheet is fortress-grade, and the operating leverage profile fits the investor's brief well. But the AI-receiver angle is thin (this is a marketing/fraud platform that uses AI as a feature, not a beneficiary of the AI buildout), and the company has missed new-business expectations two years in a row with no visible inflection in the FY27 commentary. The score reflects "cheap balance sheet + latent operating leverage" offsetting "weak AI angle + decelerating growth + execution doubts".