Capital Ltd 2025 net profit surges 318% on investment gains

Net profit after tax rose 318% to $71.0M in fiscal 2025, driven mainly by investment-portfolio gains, while revenue slipped 0.6% and operational EPS fell 11.6%. Adjusted EBITDA increased 1.1%, EBIT was $46.6M and capex totaled $47.1M. Management forecast fiscal 2026 revenue of $410M–$440M (≈23% growth at midpoint) and capex of $55M–$65M; MSALABS revenue is guided to $85M–$95M. Results are mixed: a strong headline profit offset by softer core operations, though guidance is growth-oriented.

Analysis

Nvidia is the clear second-order beneficiary of large-scale, repeat orders from verticals like Tesla and SpaceX: recurring, high-ASP GPU demand crystallizes a multi-year upgrade cadence for vehicle and edge compute that is much stickier than one-off datacenter refreshes. That stickiness amplifies pricing power for Nvidia and its packaging/foundry partners (TSMC/OSATs) because incremental demand is not just capacity but system-level validation that raises switching costs for customers.

Supply-side constraints are the immediate gating factor — foundry and advanced packaging lead times create a window (months to ~18 months) where Nvidia can sustain premium pricing before competitors/verticals scale internal solutions or alternate silicon appear. Export controls, TSMC capacity allocation decisions, or a sudden acceleration in verticals building bespoke in‑house accelerators (Tesla Dojo-like iterations) are the binary catalysts that would materially re-rate expectations.

For Tesla the tradeoff is pragmatic: buying Nvidia chips buys short-term performance and reduces Dojo execution risk but increases per-vehicle BOM exposure and ties EV gross margins to an external supplier whose economics are rapidly improving. That creates an asymmetric opportunity to capture Nvidia upside while hedging Tesla execution — the market is underpricing cross-company exposures and supply-chain concentration risk.

Contrarian vector: consensus narrative treats Nvidia GPU demand as nearly perpetual and margin-accretive; it underweights the probability that capex-led capacity additions (foundry + HBM) and faster-than-expected custom silicon adoption by hyperscalers/vehicle OEMs could compress ASPs within 24–36 months. Position sizing should therefore assume a 30–40% drawdown scenario if commoditization or supply normalization occurs.

Key Decisions for Investors

• Long NVDA directional but limited-risk: buy a 6–12 month call spread (e.g., buy NVDA Sep/Oct 2026 mid-OTM call, sell a higher strike ~20–30% above) sized 2–4% portfolio. Rationale: capture sustained GPU ASPs while capping premium. Target 2.5–4x payoff if NVDA sustains premium; cut if spread premium declines 50% or if foundry allocation headlines turn negative.
• Relative-value pair: long NVDA equity (3% portfolio) / short TSLA equity (1.5% portfolio) for 3–9 months to isolate GPU demand vs EV execution risk. Rationale: monetizes Nvidia’s structural pricing power while hedging auto-market cyclicality. Stop-loss 15% on either leg; trim NVDA exposure if Tesla announces an internal-AI silicon production ramp within 6–12 months.
• Supply-chain long: accumulate TSM (or ASML where accessible) on 12–24 month view, 2% portfolio, targeting ~20–30% upside as foundry and lithography capex stays tight. Risk: allocation shifts by TSMC or a weaker-than-expected fab utilization cycle — stop at 20% drawdown.