AlphaLens
Research
财报Equal-weightTP $191.00005月6日 · Morgan Stanley

Arm Holdings Marginal Beat on Estimates, Licensing Strong but Royalties Weak

中文EN⚠ quality lint: see notes

Arm Holdings Marginal Beat on Estimates, Licensing Strong but Royalties Weak

Core Conclusion

Arm’s FY4Q26 results delivered a modest beat on both revenue and EPS, but the composition reveals a structural tension: licensing surged to $819M (7% above consensus of $775M), while royalties undershot at $671M (4% below consensus of $700M). The company’s newly raised FY27-28 sales outlook to $2B+ (doubled from previous guidance) reflects strong AI custom-chip design momentum, yet royalty weakness in a period when cloud AI royalties should have been robust suggests either conversion delays or share loss. Management clipped opex growth ($760M Q1 guide vs $800M+ expected), supporting EPS guidance 11% ahead of consensus ($0.40 vs $0.36). At $208.84, the stock trades 9% above the $191 sum-of-the-parts target price. The Equal-weight rating implies limited near-term upside and embeds the view that current pricing already reflects AI enthusiasm without confirming sustained royalty acceleration.

What the Market May Be Underpricing

Two disconnects stand out. First, royalty softness is not one-quarter noise: this quarter’s miss follows a pattern of below-expected royalty conversion from what should have been strong cloud AI deployments. If this reflects share loss to custom accelerators or longer time-to-royalty from new design wins, the near-term revenue mix could remain skewed toward licensing (lumpy, lower-margin on a per-sale basis) rather than recurring royalty streams. Second, the $2B+ FY27-28 sales target, while optically impressive, remains pre-commercial. The “Arm Everywhere” event produced design announcements but no committed revenue backlog. Market pricing for a 40x+ revenue multiple already assumes successful execution; any delay in commercial availability or competitive pushback (e.g., RISC-V, in-house silicon) would compress that multiple.

Evidence Chain

Licensing strength confirms AI design wave. Q4 licensing of $819M (QQ up) beat consensus by $44M. This aligns with the company’s disclosure that AI custom-chip projects (including AGI chip commercial availability on multiple platforms) are driving design activity. The FY27-28 revenue outlook doubling to $2B+ directly references these new chip-making engagements. Investment implication: licensing momentum validates Arm’s strategic pivot toward higher-value custom silicon, but it is a lead indicator, not a revenue guarantee.

Royalty miss challenges the AI monetization thesis. Royalties of $671M were $29M below consensus. Analysts expected cloud AI royalties to be robust in Q4 and again in FY27e, yet the number came in light. Possible explanations: (i) design wins not yet converting to volume shipments, (ii) royalty rate compression as customers shift to custom architectures, or (iii) market share erosion in server-class CPUs. Investment implication: until royalty growth re-accelerates consistently, the revenue quality debate will cap multiple expansion.

Opex discipline boosts short-term EPS. Q1 opex guide of $760M (vs $800M+ consensus) allowed management to guide EPS 11% above consensus. This demonstrates cost control optionality, but it also raises the question of whether under-investment could impair long-term growth. Investment implication: opex compression provides a near-term earnings cushion, but investors should monitor R&D intensity relative to peers.

New business outlook is ambitious but unproven. The doubling of FY27-28 sales to $2B+ is the most bullish signal from management. However, the company has not disclosed specific customer contracts or revenue contribution milestones. Commercial availability of AGI chips on several platforms is a step, but pre-revenue announcements do not guarantee linear ramp. Investment implication: the upside case hinges entirely on execution; any misstep will expose the premium valuation.

Key Risks

  • Competitive erosion: RISC-V and in-house silicon from hyperscalers (AWS Graviton, Google Axion) threaten Arm’s server and edge share.
  • Smartphone royalty concentration: If handset unit growth or royalty rates disappoint, overall royalties remain tethered to a mature market.
  • China JV uncertainty: Arm China’s revenue and legal structure introduce opaque risks.
  • Litigation: Outstanding disputes over licensing fees could alter the business model or create financial liabilities.
  • New chip business execution: Delays in commercial deployment of Arm Everywhere designs would make the $2B+ target unachievable.

Valuation and Trade Implications

Sum-of-the-parts values the IP business at 45x FY29 earnings (in line with historical IP comps) and the chip business at 20x (discount to global chip makers, reflecting early-stage risk). Discounting at 11% WACC yields a $191 target. Current price of $208.84 implies ~8.5% downside. Catalyst for upside: sustained royalty reacceleration, a major customer announcement for Arm Everywhere, or further opex outperformance. Until royalty trends show consistent improvement, the risk/reward is unfavorable. Equal-weight rating is appropriate; investors should wait for either a pullback toward $190 or confirmation of royalty inflection before adding positions.

Appendix: Actuals vs Consensus (FY4Q26)

MetricActualConsensusVariance
Revenue$1.49B~$1.45B+2.8%
Licensing$819M$775M+5.7%
Royalties$671M$700M-4.1%
EPS$0.60~$0.55+9.1%
Q1 Rev Guide (mid)$1.26B$1.24B+1.6%
Q1 EPS Guide$0.40$0.36+11.1%
Q1 Opex Guide$760M$800M+-5%+

Related (同 ticker)