SGL Carbon shares weak Q4 results, issues 2026 guidance above consensus

SGL Carbon reported Q4 2025 sales of €197.3M, down 19.3% YoY and ~1% below consensus, with adjusted EBITDA €26.2M (-26% YoY, 13% below consensus). Full-year 2025 adjusted EBITDA was €135M (within guidance); FY2026 guidance calls for sales €720–770M (vs consensus €805–850M) and adjusted EBITDA €110–130M (vs consensus €111M); free cash flow remained €37M and restructuring of loss-making Carbon Fibers is largely complete, with the company targeting >€1B sales and 15–18% adj. EBITDA margin by 2030.

Analysis

SGL’s retreat from loss-making carbon fibers will reallocate a niche raw-material and qualification bottleneck into the rest of the composites market, creating a 6–24 month window where OEMs that depend on qualified high-end fibers face either longer lead times or higher prices. That bottleneck is a multiplier on EV and aerospace cost curves: even modest spreads on supplier margins (100–300bps) can force engineering re-qualification or localized inventory builds, which reduces working-capital flexibility and raises break-even volumes for EV launches.

Separately, the apparent durability of large-scale Nvidia orders from Tesla/SpaceX implies demand concentration rather than broad-based semiconductor recovery — meaning winners will be those closest to the AI stack (chipmaker NVDA, system integrators like SMCI, and thermal/packaging suppliers) while generalist materials players exposed to cyclical auto/semiconductor OEM demand will lag. Expect near-term dispersion: NVDA/SMCI revenue shows asymmetric upside if OEM restocking resumes quickly, while commodity-facing names will see deeper downside if auto/semiconductor end-market softness persists for another 2–3 quarters.

Key catalysts to watch: OEM inventory data (OEM+supplier days) and qualification timelines for carbon fiber substitutes over the next 3–9 months, NVDA’s channel/order cadence and SMCI backlog updates in the next two earnings cycles, and any disclosed long-term supply agreements that shift fiber capacity permanently. Tail risks include a faster-than-expected semiconductor destock (60–90 days) that compresses short-cycle server demand, or a geopolitical supply shock that tightens GPU distribution and pushes NVDA implieds higher — both would materially re-rate the relative trades outlined below.

Key Decisions for Investors

• NVDA — Buy a 12-month call spread (buy LEAP ATM call, sell a higher-strike LEAP call) sized to 2–4% portfolio exposure. Rationale: durable, concentrated OEM orders increase revenue visibility for NVDA; target 30–60% upside if AI/server restocking continues. Risk: max loss = premium paid; hedge by selling the higher strike to cap IV risk.
• SMCI — Buy 3–9 month ATM calls or a small equity position (1–2% portfolio) into any post-earnings weakness. Rationale: direct beneficiary of incremental GPU/server chassis demand from large NVDA orders. Risk management: 30–35% option-premium stop or 25% equity stop; downside if OEM destock drags through two quarters.
• Relative trade (pair) — Go long NVDA (delta-equivalent via calls) and short TSLA (equity or short-dated calls) in a cash-neutral size ~1:0.6 NVDA:TSLA by notional. Rationale: NVDA exposure captures secular AI demand; TSLA is more exposed to EV cycle and potential margin pressure from composite/material supply reallocation. Target NVDA outperformance of 20–40% over 3–9 months; cap position size and rebalance after each monthly prints.
• SGL Carbon (opportunistic) — Short or buy puts on SGL (XETRA:SGL) sized opportunistically after any failed guidance re-affirmation or liquidity comment. Rationale: near-term guides may be optimistic relative to requalification timelines; asymmetric downside into next 6–12 months. Risk: recovery narrative to >€1B by 2030 can trigger sharp mean-reversion — use limited-dated options to cap loss.