CERILLION PLC (CER) — Investment Research Note

Executive summary

Cerillion is a UK-listed vertical software vendor providing product-centric billing, charging and CRM (BSS/OSS) suites to ~70 telco customers across 45 countries, competing against far larger bespoke-implementation rivals like Amdocs and CSG. FY2020–FY2025 saw revenue compound from £20.8m to £45.4m and adjusted PBT more than quintuple from £3.7m to £21.8m at a record 50.9% EBITDA margin 2025-11 final results; however, FY2026 is very second-half-weighted, with H1 revenue down 14% and PBT down 41% as high-margin licence recognition slipped to H2, triggering a sharp de-rating from 1,700p in January to 1,002p today. The critical question for valuation is whether the £82.1m record back-order book and £42.5m Omantel win convert into the H2 licence revenue management is guiding to; if they do, FY26 should broadly hit consensus and the current price is undemanding.

Fair value estimate

• Methodology: Forward P/E cross-checked with EV/EBITDA and a light sum-of-parts (net cash + operating business).
• Key assumptions:
  • FY26 adjusted PBT of ~£20-23m (broadly consistent with the FY25 £21.8m base and management's confirmation of consensus expectations in the H1 statement).
  • Post-tax earnings of ~£16-18m → ~55-61p EPS.
  • Apply a 18-22x forward P/E, reflecting a high-margin recurring/subscription-heavy vertical SaaS with net cash, discounted for AIM small-cap illiquidity and the recent guidance-timing wobble.
  • Add ~110p per share of net cash (£32.5m / 29.5m shares).
• Fair value range: 1,100p – 1,400p per share (mid ~1,250p).
• Implied market cap range: £325m – £413m (mid ~£369m).
• Vs. current £292.5m mcap (1,002.5p): ~25% upside to the mid.
• Cross-check EV/EBITDA: EV of £260m / FY25 EBITDA £23.1m = 11.3x; well within global vertical-SaaS norms and cheap versus quality of the recurring base.

Sector context

• Sector: Technology / vertical enterprise software (BSS/OSS for telecoms).
• Quality: Above sector average on gross margin (~80%) and cash conversion; below average on scale.
• Peers: Amdocs (NASDAQ), CSG Systems (NASDAQ), Comarch (WSE). Cerillion trades at a discount to Amdocs on EV/EBITDA despite higher margins, reflecting scale, AIM listing and lumpiness.

Investment thesis

• Large-deal breakthrough validates the product-centric model. The £42.5m Omantel win (Jan 2026) is 68% larger than the prior record £25.3m deal and gives a strong Tier-1 reference in the Middle East, alongside a £271m pipeline at a record high 2026-06 interim.
• Deep operating leverage on H2 licence recognition. ~£72.6m of contracted orders sit in the back-order book with ~39% due to unwind in the next 12 months, and gross margin on incremental licence sits at ~85%+, meaning H2 profit should more than triple sequentially if the phasing plays out 2026-06 interim; 2025-11 final results.
• Fortress balance sheet with growing dividend. £32.5m net cash, no bank debt, and 5.5p H1 dividend (+15%) — plenty of headroom to invest through the cycle without dilution 2026-06 interim.

Key risks

• H2 concentration risk. FY26 depends heavily on H2 licence deals from existing customers closing on time; a slip creates a material earnings miss and would trigger further de-rating 2026-06 interim.
• Customer concentration. Top customer generated £13.1m of FY25 revenue (~29%) and top three would be materially higher; a large customer exit or dispute would hit both revenue and support renewals 2025-11 final results.
• CEO placing signal. Founder-CEO Louis Hall sold 1.33m shares at 1500p in June 2025 (~30% of his holding), just below what turned out to be near-peak pricing; while lock-ins are in place, this reduces skin-in-the-game and coincided with subsequent share-price weakness 2025-06 placing.

Operating leverage

Cerillion has one of the stronger operating-leverage profiles among small-cap software. Cost of sales is largely people-based at ~£8.4m FY25 (19% of revenue) with gross margin 81.5%. Operating expenses of ~£16.7m FY25 (37% of revenue) are dominated by staff costs (£10.2m) that scale with headcount, not revenue. As a result, the incremental gross margin on new licence revenue is >90%, and the incremental adjusted EBITDA drop-through on high-mix licence deals is 70-80%. This is visible in the FY22→FY25 sequence: revenue rose 39% while adjusted EBITDA rose 67%, lifting the EBITDA margin from 42.0% to 50.9%. The reverse dynamic explains the H1 2026 pain: a 14% revenue decline (with almost no licence recognition) produced a 38% EBITDA decline. A 10-20% revenue beat with licence-heavy mix in FY27 could plausibly add £5-8m to operating profit versus base case (i.e., 25-40% profit uplift on a 10-20% top-line beat). AI angle is real but modest — Cerillion 26.1 embeds A2A AI-agent capability that plausibly grows value-per-customer over time, but is not itself an AI-driven revenue line yet.

Value-trap signals

None identified. Revenue is growing (albeit lumpily), back-order book at record high, dividend consistently raised, net cash rising, no accounting warning signs, audit unqualified, and management stable.

Earnings vs. expectations

• FY23: Guided ~£14.3m adjusted PBT (Oct 2023 update noted "meaningfully ahead"); delivered £16.8m — clear beat.
• FY24: Consensus ~£17.9m; delivered £19.8m — comfortable beat.
• FY25: Consensus ~£20.5m per the Oct 2025 pre-close; delivered £21.8m — slight beat.
• H1 26: Trading update in April 2026 flagged the H1 shortfall (£18m revenue / £6.2m EBITDA) and reiterated the FY26 consensus was still achievable; results delivered in line with the trading update.
• Pattern: Consistent history of small-to-material beats FY23-25, then FY26 is a stress test of whether management can deliver on their explicit second-half loading claim. The share price has already priced in the risk.

Conviction

3 — moderate. Anchors: (a) clean, well-disclosed accounts and a consistent multi-year beat pattern; (b) the record back-order book gives near-term visibility; (c) net cash removes any leverage risk from the range. Limits: (a) FY26 is uniquely H2-loaded and the model is inherently lumpy — a single delayed licence renewal can move PBT £2-3m; (b) small-cap AIM name with limited sell-side coverage means fair value has a wider plausible range than for a larger comp.

Driver scoring rationale

• AI beneficiary (45): Vertical SaaS with embedded AI agents; unlikely to be displaced (deterministic BSS use case) but not a pure AI picks-and-shovels beneficiary either. Value-per-seat uplift is possible but unproven.
• Operating leverage (80): 80%+ gross margin licence business with predominantly fixed cost base; upside surprises should convert to multiples of profit.
• Earnings surprise trend (72): Consistent multi-year beats FY23-25; H1 26 was in line with an explicit downgrade.
• Cyclicality (28): Telco software with long-term contracts; some capex-cycle sensitivity but customers rarely rip-and-replace.
• Moat (70): Very high switching costs, 26-year track record, standards-based (TM Forum) integration; but competes against larger vendors with deeper pockets.
• Leverage (5): Net cash £32.5m, no borrowings.
• Earnings quality (65): Clean; small SBC add-back; contract assets (accrued income) of £25m warrant monitoring but reflect legitimate long-term SaaS/licence contracts.
• Management quality (65): Founder-CEO with strong long-term record; recent large personal share sale near the top is a mild negative.
• Growth momentum (58): Long-term compounding intact; near-term optics distorted by H1/H2 phasing; £42.5m Omantel win is a step-change indicator.

Overall score rationale

Score reflects: strong operating leverage, fair (arguably attractive) valuation post the recent sell-off, fortress balance sheet, and moderate but genuine AI-augmentation angle. Held back from a higher band by the fact that AI exposure is enhancement-of-existing-product rather than a direct AI infrastructure play, and by the H2-loaded FY26 risk. Fits the "right idea at a fair price" bucket for this investor.