Raymond James raises Arm Holdings stock price target on CPU growth outlook

Raymond James raised its Arm (ARM) price target to $641 from $565 (Outperform), citing an upside from ~$238.78 and improved earnings outlook, with 25 analysts increasing estimates for the upcoming period. The fabless CPU ramp is modeled to reach ~$1.4B in FY2028 (vs ~$1B management forecast) as royalties also grow toward ~$5.4B in a bull case, though other firms cut targets on valuation/margin pressure. Net takeaway: analyst sentiment is mixed but skewing constructive, especially around AI/data-center strength and Arm’s IP-led valuation.

Analysis

The near-term winner is not necessarily the IP owner; it is the manufacturing toll collector. If ARM-based server silicon scales, TSM captures the first dollars of incremental value through wafer starts, packaging, and capacity tightness, while ARM’s economics remain a longer-duration revenue share story. That matters because the market is pricing a multi-year adoption curve today, but the cash flows are still mostly back-end loaded and dependent on external capacity.

The bigger second-order risk for ARM is channel conflict. Once ARM becomes a chip vendor, hyperscalers may diversify architectures to preserve bargaining power, which could slow the very royalty acceleration bulls are extrapolating. That creates a classic duration mismatch: the stock can keep levitating on analyst revisions, but any disappointment in 2025-2026 data-center mix will hit a multiple that is already assuming unusually perfect execution.

For AMZN and GOOGL, the upside is better silicon economics, not a direct revenue line. The structural benefit is lower cost per workload and improved capex efficiency if ARM cores gain share in clusters, but that is a margin tailwind that takes quarters to show up in reported numbers. Contrarian view: the market may be overpaying for ARM optionality and underpaying for the cleaner, higher-confidence benefit at TSM.

Key falsifiers: a lack of confirmed hyperscaler design wins over the next 2 earnings cycles, any delay in manufacturing ramp at TSMC, or evidence that ARM’s fabless margins are materially below the implied model. If those show up, today’s bullish revisions will likely look like peak optimism rather than the start of a durable rerating.

Key Decisions for Investors

• Long TSM / short ARM as a 3-6 month relative-value pair: the thesis is that capacity scarcity monetizes faster than ARM’s 2027-2028 revenue optionality; target 15-25% pair alpha if ARM’s multiple compresses and TSM captures the volume.
• Do not chase ARM outright after analyst target resets; if expressing upside, use a call spread 12-18 months out to cap valuation risk while keeping exposure to a genuine AI-server adoption inflection.
• Add AMZN and GOOGL only on pullbacks and only as a 6-18 month margin story: the trade is better silicon economics, not immediate earnings upside; risk/reward improves if next capex commentary confirms custom-CPU share gains.
• Set a watch item on TSM capacity commentary and lead times: if advanced-node bottlenecks ease faster than expected, TSM loses some relative benefit and the long TSM / short ARM pair becomes less attractive.
• If ARM prints two consecutive quarters without accelerating data-center mix, consider a tactical short via put spreads; valuation is the vulnerability, and the move would likely be driven by multiple compression rather than fundamental collapse.