Linde plc

▲ Bull Case

• ◆Electronics + AI structural tailwind extends through 2028+. TSMC's $100B+ US fab investment, Intel 18A and 20A nodes, and Samsung's DRAM buildout all require multi-year specialty gas supply. Linde's electronics segment grows at 12-15%/yr (vs. consensus 6-8%), driving incremental EPS of $0.50-$1.00/share by FY2028 above consensus.
• ◆Margin expansion to 32%+ by FY2028. Linde has expanded adj. operating margin from 27.6% (FY2023) to 30.0% (Q1 2026) — a 240bps gain in two years. With energy passthrough mechanics insulating margins from input volatility, pricing discipline continuing, and electronics mix accretive, margin can reach 32%. Each 100bps of additional margin = ~$1.10/share of EPS.
• ◆Capital return engine: $7-8B/yr. Dividends + buybacks total $7.4B in FY2025 with shares declining ~2%/yr. Combined with ~6-7% EPS growth, total return profile compounds 8-10% annually before any multiple movement — a defensive growth profile rare in global industrials.

▼ Bear Case

• ◆You're paying for perfection at 28.8x forward P/E. The stock's 10-year average forward P/E is ~23-25x. Even at 10% EPS CAGR, if multiple contracts from 28.8x to 24x, the 3-year price appreciation is near zero. The bear case is multiple compression risk in a market that may stop paying for quality scarcity at the same premium.
• ◆Backlog execution risk is being underpriced post-ATR/TNS delay. The Q1 2026 ATR/TNS startup slipped from Q4 2026 to Q1 2027 due to subcontractor challenges. APD's NEOM experience shows mega-project risk is real and punishing. If 2-3 additional projects slip in FY2027, sector sentiment turns and multiple compresses.
• ◆EMEA structural deindustrialization is real and persistent. EMEA is ~25% of revenue. Management has explicitly guided EMEA volumes negative in H2 2025 and has not signaled recovery in 2026. European chemicals, steel, and metals demand declining is structural — could persist 3-5+ years. If EMEA stays a drag and electronics fails to deliver, base-case EPS growth slips to 5-6%.

“The Street agrees on Linde's quality. The debate is about multiple persistence + algorithm extension. Bull bookend (PT $560-620): Linde deserves 28-30x forward P/E indefinitely because the wide moat is widening as ROCE rises, electronics provides an AI-derivative growth angle without semi-cyclicality, and the 30%+ margin is durable. Bear bookend (PT $440-490): Linde deserves 23-25x at most because it's still a 3% revenue grower at heart, margin expansion above 30% is mathematically constrained (gap to peers is already 8-10pp), and clean hydrogen backlog conversion is uncertain on policy (IRA risk). At $515.68, the market is implying ~9% adj. EPS CAGR for 10 years at constant 28x exit multiple — between our base (7.6%) and bull (variant) cases.”

• ◆Q2 2026 earnings (July 2026) — backlog update, electronics momentum, EMEA trajectory
• ◆FY2026 guidance raise potential — current guide $17.60-17.90; likely raise to $17.90-18.10 at Q2
• ◆Electronics segment growth print — sustained +10%+ confirms AI thesis
• ◆ATR/TNS Gulf Coast startup (Q1 2027) — first major clean H2 project online; execution proof point
• ◆Margin algorithm acceleration — sustained crossing of 30.5% in 2026, 31% in 2027
• ◆AI fab buildout scaling through 2028 — TSMC 2nm, Intel 18A, Samsung DRAM all require gas supply

• ◆EMEA structural decay — ongoing; ~25% of revenue exposed; could persist 3-5+ years
• ◆Industrial recession / GDP <2% growth — ~50% of revenue is GDP-sensitive merchant/packaged
• ◆FX headwinds — USD strength reduces reported revenue (60% of revenue is non-USD)
• ◆IRA / 45Q rollback (20% prob over 5Y) — would compress ~$2-3B project NPV; impairs ~2/3 of $10B backlog
• ◆Backlog project delays — ATR/TNS slip Q4 2026 → Q1 2027; if pattern repeats, sector sentiment turns
• ◆Multiple compression — 28x → 23-25x = -15-20% repricing even with flat EPS