Morgan Stanley reiterates Overweight on Arm stock, $135 target By Investing.com

Arm unveiled its ARM AGI CPU — a 3nm dual-chiplet design targeting x86 rack-scale AI cloud workloads — with high-volume manufacturing readiness by end-2026, sampling underway, production ramp expected end of fiscal 2027 and initial sales/ramp in fiscal 2028 with roughly $1B revenue guidance for fiscal 2028. Shares trade at $156.52 (up 23% YTD); Morgan Stanley reiterated an Overweight with a $135 price target while Evercore ISI, RBC and Guggenheim raised targets and 19 analysts revised earnings higher, supporting bullish sentiment despite InvestingPro flagging the stock as overvalued versus its fair value estimate.

Analysis

This strategic pivot from a pure-IP model toward full-chip involvement creates a bifurcated market: incumbents that monetize via scale silicon (hyperscalers, bespoke cloud players) versus traditional licensees who may see margin pressure or conflict. Expect the fastest second-order profit capture to accrue to wafer/packaging capacity owners and EDA/IP ecosystems as design complexity rises and customers outsource more integration risk.

Key risks are execution and node scarcity. The board-level and supply-chain choreography needed to get leading-node volumes online is measured in quarters-to-years, so near-term multiple expansion is hostage to sampling cadence, compiler/stack maturity and customer-specific validation timelines; any slip will compress forward EBITDA expectations materially.

For market structure, a modest share of cloud inference deployments switching away from incumbent accelerators would compress ASPs in that subsegment and reallocate incremental gross profits toward chipmakers and fabs rather than GPU suppliers. Watch cadence of hyperscaler validation, leading-node capacity allocations and EDA billings as early, high-signal indicators of adoption versus marketing noise.

Key Decisions for Investors

• Go long ARM via 12–18 month LEAP calls (allocate 1.5–3% of equity risk): asymmetric upside if design wins convert into multi-hundred‑million revenue runs; downside limited to premium if HVM slips. Monitor quarterly sampling/validation commentary to scale in.
• Buy TSM (TSM) exposure via 6–12 month call spreads (buy calls, sell further OTM) sized 2% notional: benefits from node demand and ASP mix; hedge by selling 0.3x notional NVDA exposure to partially offset cyclical fab‑cycle risk.
• Long Synopsys (SNPS) or Cadence equivalents for 6–12 months (buy stock or 9–12 month calls): EDA revenue is a leading indicator of complex SoC ramp; treat as a barometer trade—tighten stops if billings don’t accelerate within two reporting seasons.
• Pair trade for defensiveness: Long META (0.5–1% allocation) / Short NVDA (0.5% allocation) over 6–12 months — expresses faster CPU/inference adoption at hyperscalers while protecting against a broad‑based GPU re-rating. Close or rebalance on evidence of persistent GPU ASP resilience or slower-than-expected software stack adoption.