Pennant International Group plc (AIM: PEN) — Research Note

Executive summary

Pennant is a UK micro-cap providing Integrated Product Support ("IPS") software (Auxilium), technical services, and training-system hardware to defence, aerospace and rail customers. The period covered shows a multi-year strategic pivot from project-based training hardware to higher-margin recurring software, executed against meaningful operating turbulence — revenue fell from £16.0m (2021) → £13.8m (2024) → £9.7m (2025) as a restructuring of Training Systems coincided with delayed MOD contract awards, and Auxilium ARR grew to £2.4m (FY25, +26% YoY). The single most important valuation question is whether the FY26 break-even adjusted PBT guidance and the three-year plan to grow ARR >£4m and hit 20% EBITDA margin are credible — because the audit report carries a material uncertainty about going concern and the equity is effectively a call option on Auxilium execution.

Fair value estimate

Fair value range: 20p – 35p per share (implied market cap £9.5m – £16.7m). Methodology: sum-of-parts cross-checked against a forward EBITDA multiple.

Key assumptions:

• Auxilium software at FY26e ARR of £3.0m 2026-03 final results, applying 3–4x ARR (modest for niche defence software with high churn risk and small absolute scale) → £9–12m.
• Technical Services run-rate £5–6m at 0.5–0.7x sales → £3–4m.
• Training Systems £2m revenue at 0.3–0.5x → £0.6–1m.
• Less net debt £0.5m and shareholder loan £0.3m due April 2026.
• Cross-check: management 2028 targets imply ~£15–18m revenue at 10% PBT (£1.5–1.8m). Applying 10–15x P/E and discounting back at 15% gives £11–20m equity value.

Current market cap: £12.4m. Midpoint fair value 28p (£13m) implies the stock is fair-valued with c.+8% upside to mid, +35% to high, –23% to low.

Sector context

ICB classification "Technology" is technically correct but misleading: ~95% of revenue is to defence/aerospace/rail end-markets, so peer comparison should mix small-cap UK defence (e.g. Cohort, MS International), defence software (QinetiQ at a different scale) and niche industrial software. Pennant's quality is below typical listed peers: smaller scale, weaker balance sheet, lower gross margins than pure-play software peers (49% vs >70% for software comparables), and a going-concern qualification. Closest listed analogues by mix are tiny — Cohort plc is the most relevant UK defence-services peer but trades on much stronger metrics.

Investment thesis

• Auxilium ARR scaling rapidly off a small base. £1.9m → £2.4m → guided >£3.0m (FY26), with 26% YoY growth, the Siemens Digital Industries global OEM partnership signed in Q3 2025, and new distributors in South Korea/Japan/India 2026-03 final results. A successful integrated Auxilium launch in April 2026 could materially re-rate the software piece 2026-03 final results.
• Restructured cost base means meaningful operating leverage on revenue recovery. The 2024–25 restructuring delivered ~£2m of annualised savings and reduced headcount from 140 to 121; with FY24 having delivered £1.7m adj. EBITDA on £13.8m revenue, a return to £13–14m revenue at maintained 49% gross margin should drop a high share to EBITDA 2025-04 final results; 2026-03 final results.
• Strengthened order book underpins FY26 visibility. £23.3m three-year contracted order book (vs £15.9m FY24), of which £9.7m is for FY26 delivery — combined with £18m FY25 order intake, this gives the highest revenue visibility entering a year that the company has had recently 2026-03 final results.

Key risks

• Going concern material uncertainty. The FY25 audit report contains an emphasis of matter on going concern: cash forecasts rely on uncertain pipeline conversion timing and the £1m HSBC overdraft is the only committed facility 2026-03 final results, Going Concern note. The £0.3m Brett Gordon shareholder loan at 9.75% due April 2026 is a near-term liquidity event.
• Customer concentration & MOD procurement risk. A single customer (Boeing Defence UK / Apache) contributed £3.5m in 2024 (25% of revenue). The 2025 revenue collapse was driven by MOD procurement delays around the Strategic Defence Review and the GenFly contract delay 2025-09 interim; 2025-08 trading update. Recurrence of timing slippage would re-trigger the going-concern issue.
• Sub-scale software business with thin moat. Auxilium ARR of £2.4m is a tiny base; £1.5m/year of capitalised R&D against £2.4m ARR is high relative to incremental revenue. Heavy intangible asset balance (£4.9m) carries impairment risk if execution slips (FY24 saw £0.8m impairment of Training intangibles) 2025-04 final results.

Operating leverage

The cost base is dominated by fixed people and central costs: FY25 admin costs of £6.6m (adjusted) sit against £9.7m revenue, with internal management/licence-fee allocation of £3.0m reflecting central R&D and corporate costs that are essentially fixed 2026-03 final results, segment note. Gross margin has been remarkably stable at 49–50% through revenue swings from £15.5m to £9.7m, which is the clearest evidence of operating leverage in the model — the missing revenue dropped almost entirely to the bottom line, taking the group from £1.7m EBITDA to (£0.4m). The Software & Services CGU within FY25 generated £1.1m segment profit on £7.6m revenue (14.5% margin); incremental software revenue should carry contribution margins well above 70% as platform costs are sunk. Management's own bridge — from FY25 (£0.4m) EBITDA to FY28 target of 20% EBITDA margin on ~£18m revenue (£3.6m EBITDA) — implies roughly £4m of incremental EBITDA on £8m of incremental revenue, i.e. ~50% incremental EBITDA margin. A 10–20% revenue surprise above the current FY26 path of ~£13m would plausibly add 60–100% to EBITDA from the FY24 baseline.

Value-trap signals

• Going-concern emphasis of matter in both FY24 and FY25 audit reports. Repeated, not a one-off.
• Multi-year statutory losses (FY23: £0.9m loss; FY24: £2.6m; FY25: £2.3m) despite "transformation" narrative.
• Repeated equity raises at discounts (May 2024 placing at 25p; September 2025 subscription at 21.5p; FY25 placing of ~£0.9m). Dilution of ~7m shares (~17%) since 2023.
• Related-party shareholder loan at 9.75% — sign of constrained access to commercial credit.
• Capitalised development costs of £1.5m/year vs total intangibles of £4.9m — aggressive capitalisation; cash R&D burden is real and persistent.
• AGM 2025 disapproved 10% placing authority; AGM 2026 disapproved pre-emption disapplications — shareholder pushback on dilution 2026-05 AGM results; 2025-06 AGM results.

Earnings vs. expectations

• FY23 (reported June 2024): in line with market expectations; EBITA £1.4m (statutory operating profit £0.1m) — met.
• FY24 (reported April 2025): adjusted EBITA £1.2m vs prior guidance — met expectations but with material £2.3m exceptional restructuring costs.
• H1 2025 (reported Sept 2025): preceded by an 8 Aug trading update that lowered FY25 guidance below market consensus due to GenFly delay; H1 revenue £4.5m vs £7.4m H1 2024.
• FY25 (reported March 2026): revenues £9.7m, adjusted loss £1.9m — in line with the reduced expectations set by the August trading update, but a clear miss vs. expectations entering the year. Pattern: management hit results after re-basing guidance lower; original-year guidance has consistently been missed when MOD contract timing slipped.

Conviction

Conviction: 2 (low–moderate).

Anchors: (i) the financial disclosure is reasonably clean and the segment / ARR breakdown allows a defensible sum-of-parts; (ii) the recurring software piece (ARR £2.4m, growing) is the most tangible value driver and the methodology to value it is unambiguous; (iii) net debt is now small.

Limits: (i) the FY26 break-even guidance depends on pipeline conversion timing that has slipped before, and the going-concern emphasis means tail-risk is non-trivial; (ii) sub-scale software businesses with concentrated defence customers are hard to multiple — a 3x vs 5x ARR judgement swings my fair value by 25p+; (iii) management has missed its initial-year guidance two years running.

Driver scoring & overall score

Overall score: 230 / 1000. Pennant is a partial-fit-at-best for this strategy. The AI angle is very thin (defence IPS software with no demonstrable AI-driven revenue line), and while operating leverage and balance-sheet quality are acceptable, downside protection is weak (going-concern emphasis, near-term shareholder loan refinancing, dependency on MOD procurement timing). Valuation is roughly fair rather than cheap, so the "fair price for the right idea" criterion is not met because the idea itself isn't an AI-receiver thesis.