Arista (ANET) Q3 2026 Earnings Call Transcript

Arista Networks reported Q3 2026 revenue of $3.66 billion, up 58.5% year over year and above its $3.3 billion guidance; non-GAAP diluted EPS rose 55.8% to $1.20, and operating margin was 49.1%. Management raised FY2026 revenue guidance for the fourth time to $13.3 billion and introduced FY2027 revenue outlook of approximately $18.5 billion, implying 38%–40% growth; AI fabric revenue is targeted at $7 billion in 2027, double the 2026 target. Management cited strong AI and enterprise demand, while warning that supply constraints persist and inventory fluctuations could affect near-term operating cash flow.

Analysis

The key upside is not simply more AI spend: accelerator diversity can weaken the advantage of vertically integrated networking stacks and make an open Ethernet fabric more relevant. That could expand Arista’s addressable role across deployments involving AMD, Microsoft, Arm, Qualcomm and others, while reducing dependence on any single accelerator ecosystem. But partner design work and customer trials are not equivalent to production revenue; scale-up appears to be a 2028 opportunity, not a material 2027 earnings driver.

Near term, the raised outlook supports estimate revisions and momentum in ANET. The more important 1–3 month test is whether shipments and supplier availability convert management’s demand confidence into revenue without sacrificing gross margin or cash conversion. Inventory and purchase commitments increase execution leverage in both directions: delayed customer deployments or component mismatches could leave working capital tied up and pressure cash flow. Multi-year supply agreements mitigate access risk but do not eliminate timing or obsolescence risk.

The structural upside is broader than hyperscaler AI. Enterprise security and campus refreshes may diversify demand, although long refresh cycles make this a gradual share-gain story. Contrarian risk: the headline 2027 growth outlook may encourage investors to capitalize early-stage scale-up and security monetization as if already proven. AVA pricing is not established, and management’s vulnerability comparison is a company claim, not independent evidence of customer willingness to pay. Watch gross margin, inventory turns, operating cash flow, and whether 2027 growth is supported by realized shipments rather than supply commitments alone.

Key Decisions for Investors

• Maintain a positive ANET bias, but add on post-event weakness rather than chase an earnings-gap move; the near-term catalyst is delivery against Q4 guidance and the 2027 outlook. Reassess if revenue growth misses the guided trajectory or gross margin falls below the stated 2027 range of 62%–64%.
• For a relative-value expression, consider long ANET versus Cisco Systems, sized modestly: Arista appears better positioned to capture open-Ethernet AI fabrics and software-led campus refreshes, but this is a share-gain thesis, not a claim that Cisco’s business is impaired. Exit or reduce if Arista’s enterprise momentum stalls or AI shipments fail to convert into revenue.
• Treat scale-up as a watch item, not a 2027 earnings underpinning. Revisit the thesis as 2027 trials disclose design wins, customer commitments, and Arista’s role in rack fulfillment; production timing or economics remain unverified.
• Track cash conversion alongside growth: rising inventory and commitments are sensible if deployments arrive on schedule, but deteriorating inventory turns or operating cash flow while revenue guidance remains intact would signal a growing timing mismatch and warrant reducing exposure.