Deutsche Bank raises STMicroelectronics price target on AI growth

STMicroelectronics reported Q4 FY2025 EPS of $0.11 vs $0.28 expected (a -60.71% surprise) while revenue beat slightly at $3.33B vs $3.29B. Deutsche Bank raised its price target to EUR32 from EUR28 (Buy), while Mizuho and BofA lifted targets to $32 and $34 respectively (Neutral), following a multi‑billion, five‑year AWS commercial agreement. Management/analysts see Cloud AI revenue >$500M in 2026 and ramping to >$1B in 2027, and the stock trades at $33.53 (≈4% below the 52‑week high) after a 31% six‑month gain.

Analysis

STM is the clear marginal beneficiary of incremental AI datacenter silicon spending, but the more interesting second-order winner is the supply chain nodes that support mixed-signal and analog-heavy AI I/O (mature CMOS fabs, power-management die, substrate & testing providers). That creates a multi-year content growth vector that is stickier than a single SoC win: once validated in a cloud OEM deployment, qualification cycles and system-level optimization favor incumbents and raise switching costs over 12–36 months.

Main risks are execution and margin mix rather than top-line visibility: growing datacenter revenue from tens to hundreds of millions pressures product mix (higher NRE, system-level cost) and could compress near-term gross margins if pricing/packaging is aggressive to win share. A cloud capex pullback or a large, lower-margin design win could flip consensus growth into a 6–12 month sentiment unwind; conversely, a second large hyperscaler win would be non-linear upside.

From a positioning perspective, the trade is hingeing on 2026–2027 milestones (initial >$500M run-rate in first half of 2026 vs sustainable >$1B in 2027). That makes 12–24 month option structures and pair trades more attractive than outright levered equity; you want optionality around the 2026/2027 validation cliff while protecting against a binary guidances miss.

Contrarian read: the market narrative emphasizes cloud logos (AWS) but understates that STM’s path to scale requires repeatable design wins across multiple hyperscalers and steady ASPs; the stock’s move is probably underpricing downside from a single-quarter EPS miss or a customer concentration scare, while overpricing the probability of rapid margin expansion. Treat current optimism as conditional, not permanent.

Key Decisions for Investors

• Directional options play (12–18 month): Buy a STM calendar/vertical call spread to capture the 2026→2027 revenue ramp while capping premium. Example sizing: buy Jan-2027 1:1 call spread ~OTM (delta ~30) and sell a higher strike to fund ~50–70% of premium. Risk/reward: limited known loss (premium) vs asymmetric upside if STM hits >$1B revenue in 2027 (target 2–4x payoff), stop-add discipline on quarter-to-quarter miss.
• Paired equity hedge (6–12 months): Long STM (small core position) funded by a sized short in SMCI to neutralize cyclicality of hyperscaler server demand. Rationale: STM captures silicon content gains, SMCI exposes system-level demand and valuation multiple mean-reversion. Target: aim for net delta ~0.25–0.5; expect 30–50% relative outperformance if STM executes, downside capped to ~20% outright on execution failure.
• Tactical buy-the-pullback (weeks→months): Accumulate STM on any >10% post-earnings pullback into guidance reprints, with initial stop at 12–15% below entry and scale-out on positive datapoints (additional hyperscaler wins or margin guidance upgrade). Risk/reward: asymmetric — modest initial allocation with ability to double on validated 2026 bookings.
• Event hedge: enter a protective collar around existing STM exposure into major news/call periods (buy put ~10–15% OTM, sell near-term calls) to finance protection through the next 2 quarterly results. This preserves upside participation while limiting drawdown risk from a surprise EPS miss.