Apple May Break a 10-Year Chip Strategy

Apple is exploring moving some lower-end iPhone SoC production away from its decade-long TSMC relationship, with reports and analysts (including GF Securities' Jeff Pu and Ming-Chi Kuo) suggesting Intel could begin fabricating select lower-end A-series and M-series chips as early as 2027–2028 — limited to manufacturing rather than chip design. The shift is driven by TSMC reallocating capacity to Nvidia and AI customers, while rising memory prices have had a minimal recent impact but are expected to modestly pressure gross margin next quarter; Apple reported revenue of $143.8 billion (up 16% YoY) last quarter and is guiding for 13%–16% YoY revenue growth with a 48%–49% gross margin this quarter.

Analysis

Market structure: A partial Apple move to Intel (target window 2027–2028) benefits INTC (foundry revenue upside) and NVDA (TSMC reallocates capacity to AI); TSM (TSMC) is the obvious near-term loser on Apple SoC volumes and negotiating leverage. If Intel captures 10–20% of A‑series wafer volume by 2028, TSMC’s Apple-related revenue could fall materially (single‑digit % of TSM revenue), while memory suppliers (Samsung, SK Hynix) retain pricing power forcing Apple to absorb margin pressure near term (Apple guides 48–49% gross margin this quarter).

Risk assessment: Tail risks include a failed Intel fab ramp (yield shortfalls), a public Apple‑Intel MOU that triggers a selloff in TSM, or geopolitical disruption in Taiwan — each could move prices >10% in a day. Time horizons split: days–weeks (earnings/guidance shocks, memory price headlines), months (contract negotiations, capacity reallocation), and multi‑year (actual wafer supply transitions 2027–2029). Hidden dependencies: packaging, node parity (18A claims), and ODM integration; a source shift that’s fabrication‑only still requires yield parity to avoid product cycles disruptions.

Trade implications: Favor tactical exposure to INTC via LEAPS or call spreads for a 2027‑expiry window to capture a supply deal (size 1–2% portfolio). Run a relative pair: long INTC / short TSM (ratio 1:1) size 0.5–1% to express shifting fab share while limiting broad semiconductor beta. For AAPL, maintain or add on pullbacks >5% driven by memory cost headlines; hedge downside 3–6 months with OTM puts if gross margin guidance disappoints.

Contrarian angles: Market may overstate permanent damage to TSMC — historically Apple split sourcing (A9 era) without lasting share collapse; shorting TSM on headline risk could be crowded and temporary. Conversely, the consensus underprices execution risk at Intel: absent firm contractual milestones by mid‑2026, scale expectations to ~0–5% probability of meaningful 2027 supply. Action threshold: if Apple/Intel announce sampling or MOU by Q4 2026, increase long INTC and trim TSM exposure; absent that, keep positions small and event‑driven.

Key Decisions for Investors

• Establish a 1–2% portfolio position LONG INTC via 2027 LEAPS (buy 2027 $40–$60 call spread depending on basis) to capture potential foundry revenue from Apple; scale in after any official MOU or sample‑shipment announcement (target entry within 30 days of announcement), trim if INTC misses ramp milestones for two consecutive quarters.
• Initiate a 0.5–1% PAIR TRADE: LONG INTC / SHORT TSM (TSM) equal notional to express expected fab-share shift; use stops at 15% adverse move and reassess if TSM reports Apple volume decline >5% YoY in a quarterly release.
• Keep AAPL core holding steady; add on pullbacks >5% tied to memory‑price headlines. Protect 3–6 month downside by buying 3–5% notional OTM puts (e.g., 5–10% OTM) ahead of Apple’s next quarterly guidance if management flags gross‑margin risk.
• Avoid outright large short on NVDA despite positive read‑through; instead use 3–6 month call spreads around earnings/catalysts if seeking leverage (limit to 0.5% portfolio) because NVDA’s AI demand likely offsets Apple‑TSMC reshuffle.