WINKING STUDIOS LIMITED (WKS) — Investment Research Note
Executive summary
Winking Studios is a Singapore-headquartered, Cayman-incorporated global AAA game art outsourcing group with c.1,400 employees across 14 studios in Asia and Quebec, serving 22 of the top-25 global publishers. Reported revenue has grown strongly on the back of the 2025 Mineloader and 2026 Ampera acquisitions (1H2026 revenue +21.1% to US$23.5m, organic +8.9%), but profitability has inflected the wrong way: 1H2026 Adjusted EBITDA fell 49% to US$1.2m, the Group posted a US$2.5m statutory loss, and the Board now guides to a "modest Adjusted EBITDA loss" for FY2026 as it pushes discretionary spend into Ampera and AI-enabled game development. The single most important point for valuation today is that WKS is a labour-based services business in transition — the current price does not embed the AI narrative, but the operating model is not one that captures the AI-receiver thesis with much leverage.
Fair value estimate
- Fair value range: 12p – 18p per share ≈ £53m – £80m market cap.
- Methodology: Blended EV/revenue and EV/normalised-EBITDA cross-check.
- Balance sheet: US$23.2m cash + US$1.4m bond investments – US$4.0m short-term borrowings = c.US$20.6m net cash ≈ £16m 2026-08 half-year.
- FY2026e revenue c.US$50-55m (1H US$23.5m + guidance that 2H > 1H).
- Applying 1.0x–1.4x EV/sales (in line with games-services peers post de-rating) gives EV c.US$50-77m, plus net cash = mcap c.US$70-97m ≈ £53-73m.
- Cross-check: 8-10x normalised Adj-EBITDA of c.US$5-6m (broadly the FY2025 level, ex-Ampera/AI investment) = EV US$40-60m + net cash = US$60-80m ≈ £45-60m. Applying a small premium for optionality on Ampera/AI conversion pushes the top end to c.£80m.
- Vs. disclosed market cap of £63.0m: shares are trading roughly at the midpoint of my range — modest downside of c.-6% to modest upside of c.+26%, mid-point implied c.15.5p vs. spot 14.25p, i.e. c.+9% upside.
Sector context
ICB classification (Industrial Goods & Services) is technically correct but misleading — economically WKS is a games / creative services outsourcer, closer to Keywords Studios (delisted 2024, taken private by EQT), Team17 Group and Sumo Group (delisted 2022) than to industrials. Quality profile is in line with peers: comparable customer concentration (top-25 publishers), similar gross margin band (28-32%), and similar cyclical exposure to publisher spending. Balance sheet (net cash, no debt of note) is better than peer average. Growth is comparable to (or slightly ahead of) surviving listed peers.
Investment thesis
- Consolidation play in a fragmented, structurally-growing niche. Outsourced game development is forecast to reach US$16bn by 2030 (+US$5bn on 2025), and publishers are consolidating around fewer, larger integrated partners; WKS already serves 22/25 top publishers and has 24-month indicative bookings of US$51.6m, up on both June 2025 (US$49.4m) and December 2025 (US$48.6m) 2026-08 half-year, 2026-07 trading update.
- Fortress balance sheet gives optionality. US$24.6m of cash/bonds and only US$4m of newly-drawn credit-facility borrowings against a £63m market cap; roughly a quarter of the market cap is net cash, so the enterprise is valued at c.£47m for a business doing c.US$50m of revenue 2026-08 half-year.
- Ampera + AI investments create asymmetric optionality — if they work. Ampera cost only c.US$0.4m upfront cash + deferred stock and puts founder Claude Bordeleau (ex-Keywords senior leader) in as CRO with pipeline visibility on multiple large Western contracts expected to award in 2027; AI-enabled game development is being sold as a way to add engineering/full-game capability without proportional headcount 2026-08 half-year, 2026-03 acquisition announcement. If either lands, revenue mix upgrades — but neither is in current numbers.
Key risks
- Guidance cut and profit inflection — Board now expects an Adjusted EBITDA loss for FY2026 vs. the +7-13% growth previously guided at the FY2025 trading update, driven by simultaneous Ampera and AI investment 2026-08 half-year, 2026-01 FY25 trading update. Track record of "invest through" narratives payingoff quickly in this industry is mixed.
- AI is a two-edged sword. Management is candid that AI art assets are <1% of revenue today, and the core art-production business argues human creativity is needed — but if AI coding/asset tools improve materially, the labour-arbitrage moat compresses and pricing pressure follows (not disclosed but inferred).
- Customer/geographic concentration & China exposure. Mainland China & HK combined is 40.6% of 1H2026 revenue; a small handful of publishers (Sony, Microsoft, EA, 2K, Tencent evergreen titles) drive a large share of revenue 2026-08 half-year, 2025-08 half-year. Related-party history with Acer (now divested) shows the ownership structure has been in flux.
Operating leverage
This is the weakest pillar for this investor. WKS is fundamentally a labour-based services business: >1,400 employees, gross margin of c.30% underlying (24% headline in 1H2026 after AI/Ampera drag), and cost of sales scales close to 1:1 with revenue — 1H2026 revenue +21% and cost of sales +32%. Distribution and admin expenses jumped 52% and 40% respectively on the same revenue base, showing that fixed-cost absorption is actually going backwards during the investment phase. On a like-for-like basis excluding Mineloader Q1, Ampera and AI, underlying gross margin was 30.1% vs. 30.2% prior year — i.e. essentially flat, no operating leverage visible 2026-08 half-year. A 10-20% revenue beat above plan at this cost base would likely add only 20-40% to operating profit at best; there is no software gross margin, no meaningful spare capacity in a talent-bound business, and no discernible network effect. The theoretical operating leverage would only emerge if the AI-enabled game-development pitch is real — i.e. if incremental revenue can be delivered with disproportionately less headcount. That is not visible in the numbers yet.
Value-trap signals
- Guidance was cut sharply between January 2026 (Adj EBITDA +7-13% for FY25 on a growth trajectory) and August 2026 (Adj EBITDA loss for FY26).
- Adjusted-to-statutory gap is widening: 1H2026 Adjusted expenses of US$2.3m (+396% YoY) are a growing share of the reported profit, dominated by share-based comp (US$0.7m), acquisition costs (US$0.6m) and FX (US$0.6m).
- Related-party arrangements with Acer / Acer Gaming existed for years; Acer divested in 2H2025 — a "flushing" of the shareholder base without a natural successor anchor.
- Cash outflow from operations of US$5.3m in 1H2026 (vs. +US$0.6m 1H2025). Cash has fallen from US$39.8m end-2024 to US$23.2m at 30 Jun 2026.
- Persistent modest dilution: performance-share plans (34m+ shares in issue-flight) and Ampera/Mineloader stock-based incentive pools.
Earnings vs. expectations
- FY2024 → FY2025: Delivered — Jan-26 trading update flagged revenue at least +40% (vs. broker consensus US$43.6m); delivered on M&A-driven growth path. Adj EBITDA guided +7-13% and broadly met.
- FY2025 → 1H2026: Missed on profits. July-26 trading update flagged Adj EBITDA of US$1.0-1.3m (down from US$2.4m 1H25) — delivered US$1.2m (middle of the guide) but this itself was materially below the trajectory implied by January's FY25 guidance.
- FY2026 outlook: Board now guides to an Adj EBITDA LOSS — a clear downgrade to the growth story broker consensus was building. Pattern: revenue delivery is credible; profit delivery is not, particularly when management chooses to invest.
Conviction
Conviction: 3 (moderate).
- Anchors: (i) disclosure is clean and reasonably granular, with segment, geographic and organic-vs-M&A splits; (ii) balance sheet is transparent — the net cash cushion is unambiguous; (iii) two independent valuation approaches (EV/sales, normalised EV/EBITDA) land in overlapping ranges around the current price.
- Limits: (i) the business is in a self-declared investment phase with the endpoint unclear — Ampera large-contract awards not expected until 2027, AI monetisation "modest at this stage"; (ii) profit trajectory has just inflected, so the "normalised" EBITDA I'm valuing off is not what the company will actually earn in the next 12-18 months.