Software Circle PLC (SFT) — Investment Research Note

Executive summary

Software Circle is a UK/Ireland-focused serial acquirer of Vertical Market Software (VMS) businesses that runs its portfolio on a decentralised basis and reinvests cash into further acquisitions, in the style of Constellation Software / Judges Scientific / Halma. Across FY23–FY26, revenue has compounded from £11.7m to £22.3m (24% CAGR), recurring revenue from £4.1m to £16.8m (60% CAGR, now 76% of the mix), and Adjusted EBITDA margin has stepped up from 3% to 26%, with FY26 delivering +22% revenue and +83% aEBITDA growth 2026-07 final results. The single most important valuation point is that the shares have de-rated ~45% from a 30p peak to 16.9p even though FY26 beat management's own 25% aEBITDA target and the group has just refinanced onto a £25m Santander RCF — leaving it trading at roughly 10–11x run-rate aEBITDA, a discount to serial-acquirer peers.

Fair value estimate

Methodology: EV / run-rate Adjusted EBITDA, benchmarked against listed UK/EU serial acquirers of software / niche assets.

Key inputs (all from the 2026-07 final results and 2026-05 trading update):

• Run-rate revenue: £25.0m at 27% aEBITDA margin → run-rate aEBITDA ≈ £6.8m
• Net debt at year-end: £11.4m (leverage 2.0x reported, ~1.8x adjusted for contingent consideration and part-year contribution)
• Shares in issue: 390.1m

Applied EV/run-rate aEBITDA range 12x–18x (Judges Scientific trades at ~18–22x and Constellation-style compounders at >20x; SFT deserves a discount for scale, AIM listing, and shorter track record):

Fair value range: 18p – 28p per share (implied market cap £70m – £111m; mid £91m) vs. current £61.4m market cap / 16.9p → upside of ~7% (low) to ~66% (high), ~37% at midpoint.

Cross-check: management's "Gate 5" scenario (£15m aEBITDA in ~3–4 years) at 15x and typical £15m net debt discounts back at 12% to roughly £145–170m equity, or 37–44p — supportive of the range without needing heroic assumptions.

Sector context

Correctly classified as Technology (application software; vertical SaaS). Compared to typical UK Tech peers: quality is in line (76% recurring revenue, 34% oEBITDA margin at operating level, 43% at acquired-portfolio level, low customer concentration), growth is above (mid-20s% CAGR), leverage is above (2x net debt vs. many peers with net cash) but manageable. Closest listed comparables:

• Judges Scientific plc (JDG) — buy-and-build compounder; premium multiple
• Volution Group (FAN) or discoverIE (DSCV) — capital-allocator archetype
• Constellation Software / Topicus / Lumine — the aspirational Canadian VMS compounders that management explicitly cites

Investment thesis

1. Genuine compounder flywheel, now self-funding. FY26 deployed £12.1m across two acquisitions (AIF, BIS forming an Irish fintech cluster) at disciplined multiples, delivered Operating Return on Capital Deployed of 25% and Operating Cash Flow Per Share doubled to 1.1p; the new £25m Santander RCF plus £10m accordion (May 2026) means the acquisition programme is now debt- and cash-flow-funded, removing dilution risk 2026-07 final results, 2026-05 RCF announcement.
2. Valuation has de-rated despite operational beats. Shares are ~44% off the FY25 26p high on no bad news — FY26 beat the 25% aEBITDA margin target, hit Gate 4 (£5m aEBITDA), and current trading is "in line with internal expectations" — leaving SFT at ~10–11x run-rate aEBITDA vs. Judges Scientific at ~20x 2026-07 final results, market data.
3. Owner-aligned, quality-anchored governance. Board additions Marc Maurer (ex-Constellation, ex-Revalize; representative of largest shareholder CHAPTERS Group AG) and Brad Ormsby (CFO of Judges Scientific through its 9x market-cap growth) bring genuine serial-acquirer DNA; the Investment Committee has formal guardrails and remuneration is tied to Operating Cash Flow Per Share, not deal count 2024-09 board changes; 2025-07 final results.

Key risks

1. Amortisation still masks true profitability and depresses reported EPS — FY26 statutory loss of £0.8m despite £5.8m aEBITDA due to £4.8m acquisition-related amortisation; if M&A pace slows or a deal disappoints, the market will fixate on the reported loss rather than the cash economics 2026-07 final results.
2. Nettl Systems (Graphics & Ecommerce) remains a drag — organic revenue declined 11.5% in FY26 and this legacy business continues to shrink; it still represents 34% of revenue but only 13% of oEBITDA, so mix is improving, but ongoing decline offsets acquired-portfolio organic growth (Group organic was -1% in FY26 vs. +7% for the acquired book) 2026-07 final results.
3. AI disruption risk to vertical SaaS is real if the "own the agent" strategy fails. Management has laid out an articulate defence (proprietary data + regulatory embed + transaction embed) but this is prospective — if a horizontal AI agent commoditises workflows in one of their verticals (adoption, care, financial broker, driving instructor, marketing compliance), the "mission-critical" moat erodes and multiples compress 2026-07 final results; inferred.

Operating leverage

The acquired portfolio operates at 43% Operating EBITDA margin with 76% recurring revenue, and central costs of £1.8m are largely fixed and shrinking as a % of oEBITDA (24% in FY26 vs. 33% in FY25) 2026-07 final results. On a per-head basis, recurring revenue has doubled from £45k to over £100k on a run-rate basis since FY23 while headcount grew 2.0x and recurring revenue 4.1x — a clear operating-leverage signature. Direct costs are only 21% of revenue (gross margin 79%). This means that where FY26 revenue at £22.3m produced aEBITDA of £5.8m, a 10–20% revenue upside — plausible from a single accretive acquisition — would drop through at incremental margins closer to 40–50%, adding £0.9m–£2.2m to aEBITDA, i.e. 15–40% aEBITDA uplift on 10–20% revenue uplift. This is textbook software operating leverage, though the compounding requires ongoing capital deployment rather than pure organic scaling. Fixed R&D is capitalised (£1.8m in FY26) and does not scale with revenue 2026-07 final results.

Value-trap signals

• Nettl legacy business in structural decline — 11.5% revenue decline in FY26, extending a multi-year trend; management's own words are "revenue to stabilise at around this year's level" but that has been said before 2026-07 final results.
• Rising leverage — net debt of £11.4m (from net cash of £6.9m two years ago) as M&A accelerates; adjusted leverage 1.8x is comfortable but rising, and the new facility permits 4x — worth watching.
• Persistent statutory losses — despite growing aEBITDA, the group has reported a small statutory loss every year since 2022. Not a trap on its own (M&A amortisation), but retail investors and quant screens may keep the multiple depressed until statutory profitability arrives.
• Nothing in the value-trap category rises to "cheap for a reason permanently"; the setup looks more mispricing than terminal.

Earnings vs. expectations

The filings mainly reference internal management targets rather than sell-side consensus. Track record vs. management guidance and targets:

• FY25 final results (Jul 2025): Delivered aEBITDA margin 17%, meeting the stated aim to "exceed 15%" 2025-07 final results.
• FY26 pre-close (May 2026): Delivered 26% aEBITDA margin, beating the previously communicated 25% target; 22% revenue growth 2026-05 trading update.
• FY26 final (Jul 2026): Confirmed above numbers; hit Gate 4 (£5m annualised aEBITDA) 2026-07 final results.
• Gate progression: Gate 1–4 hit on plan since FY22; Gate 5 (£15m aEBITDA) targeted 3–4 years out.

Pattern: consistent meet-or-beat versus management's own stated targets; no visible profit warnings across the five years of filings. No formal analyst consensus disclosed in filings.

Conviction

Conviction: 3 (moderate)

Anchors (support confidence): (a) clean, well-structured APM framework with reconciliations to statutory numbers and multi-year KPI trend tables; (b) clear disclosure of run-rate metrics that anchor the valuation; (c) established peer comparables (Judges Scientific, Constellation) validate the methodology.

Caveats (limit confidence): (a) the compounder multiple I've applied is a judgment call — a 12x vs. 18x range materially changes the fair value; (b) forward acquisition-driven growth depends on continued execution and pipeline conversion — this is not a pure organic-growth story, and one bad deal could dent the model.