Morgan Stanley China Summit reveals split views on AI transceivers and smartphones

Most investors at the Morgan Stanley China Summit view co‑packaged optics as the base case and expect strong demand for AI transceivers over the next 2–3 years, though views diverge on which suppliers will benefit. Smartphone sentiment has turned bearish since early 2026 as rising memory costs are expected to pressure shipments and margins; suppliers with Apple exposure are viewed more favorably while Android‑focused suppliers face weak confidence.

Analysis

Investor positioning is bifurcating: co‑packaged optics is moving from theoretical to base case and will compress the runway uncertainty for data‑center hardware winners. Expect 20–30% ASP premium on first‑wave AI transceivers through 2026 as hyperscalers trade higher port density for lower switch power; that premium flows disproportionately to chassis/ODM partners with early integration expertise rather than commodity optics vendors.

Smartphone DNA is shifting defensively toward share‑stable, high‑margin OEMs — Apple benefits from pricing power and services insulation — while Android suppliers face a two‑way squeeze from higher memory costs and OEM inventory digestion. If memory spot prices remain elevated beyond the next two quarterly cycles (60–90 days per cycle), expect at least a 5–10% hit to aggregate smartphone OEM EBITDA in FY26 from margin compression and modest shipment downgrades.

Second‑order winners include server ODMs and sub‑system integrators that can re‑engineer BOMs to accept co‑packaged optics (faster win rate than legacy module makers), and test/validation tool vendors who sell into initial qualification flows — these revenues spike early and are lumpy but high‑margin. Key catalysts: hyperscaler capex guides and Apple’s Sep‑26 product/price moves; reversals include a sharp memory price decline (within 2–3 quarters) or a slower-than‑expected ASIC/node roadmap that slows co‑packaged uptake.

Key Decisions for Investors

• Long SMCI (size 2–4% portfolio) via 6–12 month call spread to capture transceiver integration wins; target +100–150% if hyperscaler guides accelerate rev recognition, stop at -30% of premium to limit tech‑capex re‑risk.
• Buy AAPL stock (size 3–5%) as defensive smartphone exposure for 3–6 months, funded by selling 6–8 week covered calls to harvest premium; R/R: expect ~1–2% outperformance vs peers if Apple preserves ASPs, downside capped by stop‑loss at -8% from entry.
• Tactical long APP on any >12% pullback (size 1–2%), 3–6 month horizon — ad monetization is sticky and should outlast a one‑quarter shipment wobble; target +40–70%, stop -25%.
• Event hedge: if memory price rollback occurs (signaled by Micron/DRAM headline), rotate 50% of SMCI position into cyclical Android‑OEM suppliers (or an Android supplier ETF) within 2 weeks, locking 30–50% gains — this protects against demand re‑acceleration that favors low‑cost suppliers.