STHREE PLC (STEM) — Investment Research Note

Executive summary

SThree is a UK-listed specialist STEM staffing company placing engineers, life sciences and technology contractors and permanent hires across 11 countries, with an 85% Contract net-fee weighting and a sector-leading contractor order book providing ~5 months of net-fee visibility. Trading has cycled sharply — FY23 PBT of £77.9m collapsed to £25.5m in FY25 and guidance for FY26 is just c.£10m 2025-09-16 Q3 update; 2026-01-27 FY25 results — driven by prolonged European softness in Technology and Life Sciences hiring, offset by USA/Japan growth linked (in part) to AI-driven power/data-centre buildouts. For valuation today, the single most important question is whether the reader is buying a cyclical trough with ~£43m net cash, a completed technology platform (TIP) and improving quarter-on-quarter momentum, or a structurally challenged staffing model at risk of agentic-AI disintermediation.

Fair value estimate

• Fair value range: 260p – 400p per share (mid ~330p)
• Implied market cap: £310m – £480m (mid £395m)
• Vs. current £256.4m mcap / 215p share price: +21% to +86% upside (mid +54%)

Methodology — mid-cycle earnings x cyclical multiple:

• Mid-cycle net fees: £340m (blend of FY23 £419m peak, FY24 £369m, FY25 £323m, H1 FY26 annualised ~£295m). Note company has structurally reduced headcount ~15%, so like-for-like mid-cycle capacity may be lower than history.
• Mid-cycle operating profit conversion ratio: 12–15% (below the 18% FY23 peak, above the 8% FY25 and 2% H1 FY26 trough; reflects TIP cost efficiencies of c.£6.5m annualised achieved to date 2026-01-27 FY25 results).
• Mid-cycle operating profit: £41–£51m → PBT ~£40–£50m → PAT (25% tax) ~£30–£38m → EPS ~25–31p.
• Applying 10–13x mid-cycle P/E (staffing sector range): 250p–400p.
• Add back trapped net cash of £43m (~35p/share): 285p–435p.
• Central fair value ~330p.

Sensitivity: if FY26 truly is trough, and net fees recover to £370m at 14% conversion by FY28 (PBT ~£50m), the stock could re-rate materially. If Europe stays depressed and AI disrupts staffing structurally, £180–£220m of net fees at 8% conversion delivers PBT ~£15m and fair value closer to 150–180p.

Sector context

• ICB classification confirmed: Industrial Goods and Services (staffing/professional services sub-sector).
• Quality/growth/leverage vs. peers: Balance sheet is materially cleaner than typical UK-listed peers (net cash £43m, no debt). Contract mix and order-book visibility are structurally better than transactional generalist recruiters. Growth profile is currently worse than peers due to European concentration (DACH + Netherlands 50% of net fees), but STEM specialism is above sector average in quality.
• Listed peers: PageGroup (PAGE), Hays (HAS), Robert Walters (RWA). SThree's STEM/Contract mix distinguishes it from the more Permanent/generalist orientation of PageGroup and Robert Walters.

Investment thesis (3 bullets)

1. Cyclical trough with visible sequential improvement. New-business activity was stable YoY in H1 FY26 with a return to growth in the contractor order book (+3% YoY, first growth since Q1 FY23) and 6 of 11 Contract countries in growth 2026-07-21 H1 FY26 results. USA net fees grew 12% YoY in H1 driven by demand tied to grid hardening, data-centre construction and AI-linked energy infrastructure — a genuine AI-adjacent revenue stream, not a marketing claim.
2. Completed £32m technology platform (TIP) delivering measurable productivity gains and £6.5m annualised cost savings. Placements-per-consultant +6% Group-wide, +18% in USA (first market to go live), time-to-placement down 22% in USA vs FY23 2026-01-27 FY25 results. The platform provides operating leverage on any revenue recovery and enables agentic-AI integration on top of a unified data lake.
3. Fortress balance sheet supports capital return through the trough. £43m net cash, £98m accessible liquidity, dividend maintained at 14.3p (6.7% yield at 215p), £20m buyback launched Feb 2026 (£8.8m completed by 20 July) 2026-07-21 H1 FY26 results. Board has departed twice from its stated 2.5-3.0x dividend cover policy to sustain the payout — signals confidence but is also using retained earnings the business may need if the cycle extends.

Key risks (3 bullets)

1. Structural AI disruption of recruitment itself. Agentic AI candidate-sourcing, screening and matching directly threatens the value-add of contingent staffing consultants. Management position SThree as an "AI-enabled" firm but the disruption risk to their own model is not disclosed as a principal risk — noted only obliquely under "Industry innovation" 2026-07-21 principal risks. Inferred but material.
2. German fiscal stimulus deferred and Europe remains weak. DACH (32% of net fees) declined 15% YoY in H1 FY26; the €500bn fiscal programme has spent only €14bn in 2025 with meaningful flow expected only from 2027 2026-07-21 H1 FY26 results. If European recovery slips a further year, FY27 guidance will disappoint.
3. Dividend cover of 1.0x in FY25 is unsustainable if the trough extends. FY25 profit £17.7m vs dividends £18.4m 2026-01-27 FY25 results. Combined with the £20m buyback, the Group is returning c.£38m against ~£3m of retained earnings potential in FY26 on current guidance. A further cycle-year would force a policy-consistent dividend cut, damaging equity story.

Operating leverage

SThree has meaningful but not extreme operating leverage. Cost base is dominated by staff (£222m of £297m opex in FY25 = 75%) — largely fee-earning consultants whose numbers management flex (average headcount -15% H1 FY26 vs H1 FY25). The fixed component is central costs, the newly-consolidated Glasgow shared service centre for Candidate Operations, the TIP platform (D&A now £8.6m/half rising), and property (£49m lease liabilities). The Contract net-fee margin has been remarkably stable at 21.7% for multiple years 2026-01-27 FY25 results, meaning incremental revenue essentially drops through gross profit at ~22%.

Quantitatively: from FY25 £322.7m net fees / £26.1m op profit (8.1% conversion) to H1 FY26 £147.7m / £3.4m (2.3% conversion), each £10m net-fee swing moves op profit by roughly £5–8m at current headcount. A 10-15% net-fee recovery from FY25 base (to ~£360m) with the improved cost base and TIP efficiencies (£6.5m annualised) should return conversion to 12-14% — implying op profit of £43-£50m, roughly 2x FY25 and 5x current guidance. The key inflection point is the Glasgow shared-service-centre decoupling of net-fee growth from support-cost growth 2026-07-21 H1 FY26. Score: material but not asymmetric — this is not a software business.

Value-trap signals

• Repeated guidance cuts: FY26 PBT guidance cut from consensus £30.5m to £10m in September 2025 (-67%) 2025-09-16 Q3 update.
• Deteriorating dividend cover: 2.6x FY24 → 1.0x FY25, with Board formally departing from stated policy twice.
• Technical breach of Companies Act on distributions in FY25 requiring shareholder remediation resolution — Company had insufficient distributable reserves at the parent when paying interim/final dividends and completing the buyback 2026-01-27 FY25 results, Note 11. No cash impact but a corporate-governance red flag.
• Guarantor cover covenant on RCF breached at prior testing date (76% vs 80% required) requiring lender waiver until 13 October 2026 2026-07-21 H1 FY26 note 10.
• European macro headwinds: DACH down 15%, Netherlands down 24% in H1 FY26; German fiscal stimulus deferred to 2027+.
• AI disintermediation risk to core business model — not disclosed as principal risk.

Earnings vs. expectations

• FY23: Delivered £77.9m PBT vs consensus £71.3m → beat.
• FY24: Guided c.£67m in Dec 2023, delivered £67.6m → in line.
• FY25: Guided c.£25m in Dec 2024, delivered £25.5m → in line (but was itself a downgrade from prior consensus).
• FY26: Consensus was £30.5m in Sept 2025; management guided to c.£10m — material cut. H1 delivered £2.7m PBT (£9.8m underlying ex non-recurring), guidance reiterated in H1 results.

Pattern: Two years of in-line/beat, then a large single guidance cut in Sept 2025 flagging that the cycle would persist longer than expected. Since the cut, management has consistently reiterated the reset guidance.

Conviction

Conviction: 3 (moderate).

Anchoring the call: (1) exceptionally clean disclosure with detailed geographic/vertical/service breakdowns; (2) contractor order book gives real forward visibility for the next 4-5 months; (3) balance sheet is unambiguously strong (net cash + undrawn facilities).

Limiting the call: (1) staffing is a highly cyclical industry and mid-cycle earnings are inherently uncertain, especially with unresolved question of AI disruption to the staffing model itself; (2) recovery timing depends on European hiring cycles which have already surprised negatively; (3) fair value range is wide (260–400p) reflecting genuine bimodal outcome — cycle recovery vs structural decline.

Driver scoring