CALNEX SOLUTIONS PLC (CLX) — Investment Research Note

Executive summary

Calnex designs test-and-measurement instrumentation validating high-performance network synchronisation and network emulation for telecoms operators, hyperscalers, and government/defence customers (blue-chip roster including Nvidia, Meta, Ericsson, Nokia, Intel, NTT). Over the 5-year window the business rode a 5G capex peak in FY23 (£27.4m rev, £7.2m PBT), suffered a sharp telecoms-driven downturn in FY24 (£16.3m rev, small loss), and has since re-based around a more diversified end-market mix — telecoms fell to 30% of orders in FY26 while digital infrastructure (hyperscalers/data centres) reached 49% and government & defence 21% 2026-05 FY26 final. The single most important valuation point today: the stock trades at £53m for a company with £11m net cash, 76% gross margins, and increasing exposure to AI-driven hyperscaler timing/network emulation demand — but FY27 is an explicit investment year with the growth inflection deliberately pushed into FY28.

Fair value estimate

Range: 60p – 90p per share (£53m – £79m market cap), mid ~75p (£66m).

Methodology: blended multiple-of-forward-earnings + EV/sales cross-check, anchored to a normalised through-cycle earnings estimate rather than depressed FY26 numbers.

FY26 revenue £21.9m, Underlying EBITDA £1.8m (adjusted for R&D amortisation) 2026-05 FY26 final. Peak FY23 was £27.4m revenue / £8.0m Underlying EBITDA.
Assuming FY28 recovery to £26-30m revenue with EBITDA margins re-inflating to 15-25% (mid-cycle historical range), gives normalised EBITDA of £4-6m. Applying 10-14x EV/EBITDA + £11m net cash = £51m-£95m EV range → market cap £62m-£106m.
Cross-check on EV/Sales: current EV ~£42m / FY26 revenue £21.9m = ~1.9x. Sector peers (Spirent pre-acquisition, Viavi) typically trade 2-3x through cycle. Applied to FY28E ~£25m revenue → EV £50-75m.

Vs. current £52.8m cap: midpoint implies ~24% upside; downside ~1% (view: fair-to-slightly undervalued).

Sector context

Confirmed classification: Telecommunications / Test & Measurement equipment. Calnex's quality profile (net cash, 76% gross margins, 79% repeat orders, capital-light, no direct debt) is above typical telecoms-hardware peers, though its scale (£22m rev) is small. Growth profile currently sub-par vs. software peers but has structural AI/data-centre tailwind emerging. Listed peers: Viavi Solutions (partner, larger US-listed comp), Keysight Technologies (much larger US comp), Spirent Communications (recently acquired — its distribution relationship with Calnex was terminated in 2024, effectively freeing Calnex to build its own channel).

Investment thesis (3 bullets)

Direct hyperscaler / AI-infrastructure exposure is building beyond marketing language. FY26 saw a "significant repeat Sentry order from a leading hyperscaler" for data-centre sync monitoring; digital infrastructure is now 49% of orders (FY25: 49%, FY23: 34%). Post-period Calnex demonstrated a 400G network emulation for AI congestion testing, and the Sentry, SNE emulation, and 1.6Tb/s sync roadmap are aimed squarely at AI-training network validation 2026-05 FY26 final; 2026-09 CMD.
High operating leverage against a fixed cost base offers asymmetric upside if FY28 growth materialises. Revenue swung from £27m→£16m→£22m over three years while operating profit went £7.2m→(£0.6m)→£1.0m — a 65-70% incremental drop-through. If FY28 acceleration delivers £25-30m revenue, PBT plausibly returns to £4-6m 2026-05 FY26 final.
Fortress balance sheet + strong customer economics reduce downside. £9.3m period-end cash (£11.2m weeks later), no debt, dividends maintained through the downturn, 20-year operating history, 79% repeat-order rate, and customer base including Nvidia, Meta, Intel, BT — this is not a fragile business 2026-05 FY26 final.

Key risks (3 bullets)

Customer concentration is meaningful and increasing. In FY26, one customer accounted for 22% of revenue and 25% of orders (FY25: 17%); top-10 rose to 55% of orders on 3-yr average (FY25: 45%) 2026-05 FY26 final. A single hyperscaler pushing out orders would materially move the P&L.
FY27 is an investment year with deliberately deferred payoff. Management is explicitly increasing R&D and go-to-market spend in FY27 while targeting the step-change in FY28. If growth doesn't materialise, margins remain depressed on a higher cost base 2026-05 FY26 final; 2026-09 AGM.
Capitalised R&D is large relative to profits, blurring earnings quality. FY26 capitalised R&D was £6.3m (vs £4.6m amortisation, £1.2m PBT). The gap means underlying cash burn on R&D exceeds the P&L charge, and reported profits are flattered by capitalisation policy 2026-05 FY26 final.

Operating leverage

Calnex has high operating leverage — the historical financials give a clean illustration. From FY23→FY24 revenue fell £11.2m and underlying EBITDA fell £7.9m, implying a ~70% incremental margin. Reverse: FY24→FY26 revenue rose £5.6m and underlying EBITDA rose £1.7m (~30% — lower because of planned investment in sales/marketing headcount). Gross margins are structurally 75-76% 2026-05 FY26 final. Fixed cost base includes ~163 headcount (80 in R&D), capitalised R&D pipeline, and lease overheads — none scales linearly with revenue. Consequence: if FY28 revenue grows 15-20% above FY27 (implicit management guide), operating profit plausibly doubles or triples off the low FY26 base. Not a doubling of the whole business, but of profit. The user-relevant "long-tail" characteristic is genuinely present.

Value-trap signals

None identified. The FY24 drawdown was a well-documented telecoms industry cycle rather than a company-specific problem, revenue and orders have recovered, dividend maintained, no debt, customer base intact and diversifying, no restatements or auditor emphasis-of-matter. The only genuine watch item is capitalised R&D vs cash earnings, which is normal for a hardware-development business and not out of line with peers.

Earnings vs. expectations

Across the 5-year window Calnex has generally delivered in line with or slightly ahead of consensus, with one material miss: in October 2023 the company issued a trading update warning FY24 revenues would be 20-30% below market expectations due to telecoms customer spending delays — a clear miss driving the FY24 loss 2023-10 trading update. Otherwise: FY22 slightly ahead, FY23 in line, FY25 in line with post-downgrade expectations, FY26 "slightly ahead of market expectations" 2026-05 FY26 final, and FY27-to-date "in line" per the September 2026 AGM statement. Pattern: mostly in-line delivery, but with a demonstrated willingness to guide down when the cycle turns — not a serial disappointer, but not a stock that consistently exceeds either.

Conviction

3 — moderate. Anchors: (i) clean IFRS-compliant disclosure with detailed segmental and product-line breakdowns, (ii) simple business model that lends itself to normalised-earnings valuation, (iii) balance sheet strength eliminates financing risk. Caveats: (i) FY28 growth acceleration is the crux of the bull case and remains unproven — the range of outcomes at £22m vs £30m revenue is material for a small-cap, (ii) customer concentration means a single hyperscaler decision can move numbers meaningfully.

Driver scoring rationale

AI beneficiary (60): genuine hyperscaler traction — repeat Sentry order, 400G AI emulation capability, 1.6Tb/s sync in development, Nvidia/Meta customers — but still 30% of orders are legacy telecoms and AI-specific revenue is not yet dominant.
Operating leverage (72): high fixed cost base, 76% gross margins, small scale — recent history demonstrates ~30-70% incremental margins in either direction.
Earnings surprise trend (55): mostly in-line with one FY24 miss/guidance cut.
Cyclicality (55): telecoms testing is moderately cyclical (proven in FY24); digital infrastructure diversification reduces but does not eliminate.
Moat (55): specialised technical IP, deep long-term customer relationships (10-yr average with top 10), high switching costs once designed into test racks, but not dominant network effects.
Leverage (5): net cash of £11.2m, no debt, fortress balance sheet.
Earnings quality (58): clean disclosure but heavy capitalised R&D relative to profit; policy change on RDEC accounting in FY26 reduces comparability.
Management quality (65): founder-led (Tommy Cook since 2006), consistent execution, disciplined cost response in FY24, sensible small acquisition of iTrinegy 2022 that appears well-integrated.
Growth momentum (58): FY26 +19% revenue growth is healthy; FY27 explicitly a plateau; FY28 acceleration guided.

Overall score rationale

Score 560/1000: solid fit for the AI-receiver strategy with genuine (not press-release) hyperscaler exposure, real operating leverage against a fixed cost base, and strong downside protection from the balance sheet — but current market cap doesn't obviously undervalue the business, FY27 is a stated investment year, and customer concentration is a real risk for a small-cap. Not a screaming buy, but a legitimate portfolio candidate at current levels with meaningful upside if FY28 delivers.