Arista Networks Is Expensive, But Still Worth It

Analyst reiterates Arista Networks as a Strong Buy, setting a $216 price target versus $166 (about 30% upside). The thesis cites AI Fabrics, 800G switching, Campus expansion, software/services, deferred revenue conversion, and operating leverage supporting 2028 non-GAAP EPS of ~$5.03. The target is derived by applying a 43x forward non-GAAP P/E multiple to the 2028 EPS estimate.

Analysis

The core message for us is not that ANET is “cheap” or that this note changes 2025 numbers; it is that the stock is still being underwritten by a multi-year share-gain story in AI networking that can justify a premium only if execution remains near-perfect. In other words, the upside is less about one quarter of beat-and-raise and more about whether Arista keeps converting hyperscaler cluster buildouts into a durable mix shift toward software, support, and higher-speed Ethernet attach. That mix matters because it can make revenue less cyclical and margin expansion more durable than a pure box vendor.

The second-order risk is multiple compression: at a high-teens to low-20s growth profile, a 43x 2028 earnings framework is only defensible if investors believe AI fabrics, 800G, and campus can all compound without a digestion year. If hyperscaler capex pauses, or if Broadcom/custom silicon and Cisco pricing pressure narrow the share gains, the market can re-rate ANET before the fundamentals actually break. The key falsifier is not a modest beat/miss; it is any evidence that RPO conversion slows, gross margin stalls, or enterprise/campus offsets fail to absorb data-center lumpiness.

Near term, the stock likely trades on sentiment around AI infrastructure rather than on this reiteration itself. Over 1-3 months, the catalyst path is earnings/guidance and hyperscaler capex updates; over 6-18 months, it is whether 800G and software attach create a structurally higher revenue base. Contrarian view: the crowd may be overestimating how linear AI networking adoption will be, especially if power, optics, or integration bottlenecks delay cluster commissioning.

Key Decisions for Investors

• Stay long ANET only on pullbacks; treat this as a 6-18 month structural winner, but require confirmation in the next earnings/guidance cycle that revenue growth and margin expansion remain intact.
• Pair trade: long ANET / short CSCO for a 3-6 month window to express AI data-center share gain versus legacy enterprise networking pressure; thesis breaks if Cisco’s campus/data-center commentary improves materially or ANET growth decelerates.
• Use a downside alert, not a full short: if ANET loses ~10%-15% on an earnings gap with weakening deferred revenue or RPO conversion, de-risk because the market will likely de-rate the multiple before consensus EPS is cut.
• Watch hyperscaler capex commentary from MSFT, META, AMZN, and GOOGL over the next 1-2 quarters; if cluster spend pauses, ANET’s premium multiple is the first thing to compress.
• If seeking convexity, consider a modest call-spread structure into the next two earnings dates rather than outright shares, since the valuation already discounts a fair amount of long-term success.