That's an interesting idea — let me pull the relevant data on Arista Networks (ANET) and AMD to see if there's a case for pairing them. I'll look at price performance, valuation, and fundamentals.The pivot didn't come out cleanly. Let me redo this properly.Now let me also chart the price performance side-by-side.

Loading chart...Now let me compute a relative performance ratio to better assess the pair.

Loading chart...

Loading chart...Here's my assessment:

The Pair: Long ANET / Short AMD — A Compelling Idea, With Caveats

The Bull Case for the Pair

The fundamental divergence between these two names is striking and has widened materially over the past five years:

ANET is a margin machine — it earns roughly 3x the net income on 1/4 the revenue base. It spends far less on R&D as a percentage of sales, yet generates superior returns. ANET's moat in high-speed data center switching (competing with Cisco) is deep; it's a pick-and-shovel AI infrastructure play without the silicon execution risk.

AMD, despite strong top-line growth, has seen its operating margins collapse from 22% (FY21) to 10.7% (FY25). The gross margin hasn't broken above 50% despite the AI GPU cycle. AMD is caught between NVIDIA on the high end and the custom ASIC/ARM shift among hyperscalers. The structural concern: AMD's business model may never achieve the operating leverage that its revenue scale would suggest.

The Price Action Tells the Story

Since Jan 2020, ANET has returned roughly 8-9x vs. AMD's ~2.5x. The ratio chart (ANET/AMD) shows a powerful uptrend, especially from late 2022 onward — exactly when the AI infrastructure buildout accelerated and ANET's switching business took off, while AMD struggled to convert AI tailwinds into sustained margin expansion.

The Risks to the Pair

1. ANET valuation is rich. ANET trades at a significant premium. Any slowdown in hyperscaler CapEx (especially Microsoft/Meta) could compress the multiple sharply.

2. AMD is not broken — it's cheap. At 12.5% net margins and a mid-single-digit P/E on trough-ish earnings, AMD has optionality. If MI300X/MI400 gains traction against NVIDIA, or if the x86 server cycle turns, AMD could re-rate quickly. The short leg carries asymmetric risk.

3. Limited fundamental overlap. ANET (networking switches) and AMD (CPUs/GPUs) sit in different parts of the data center stack. This is less a classic pair trade and more a relative-value expression on AI infrastructure quality. Correlation exists through hyperscaler CapEx, but sector shocks could hit both legs simultaneously.

4. Crowding risk. Long ANET has been a consensus trade. If AI spending narratives shift, ANET could correct faster than AMD.

Verdict

As a medium-term fundamental pair, long ANET / short AMD makes sense — you're long the highest-quality AI networking franchise with pristine unit economics and short a structurally margin-challenged silicon name. But the entry point matters enormously after ANET's massive run. The pair works best as a spread trade on business model quality within AI infrastructure, not as a market-neutral hedge. Size accordingly.

Let me pull the last 12 months of price data and compute trailing performance, plus the latest quarterly trends.Now let me compute TTM figures for both companies as of the latest quarter.The last 12 months tell the exact opposite story. Let me summarize.

Last 12 Months (Aug 2025 – Aug 2026): AMD Trounces ANET

Price Performance

AMD has tripled ANET's return over the past 12 months — and by a wide margin: +177% vs. +53%. The long ANET / short AMD pair would have been a painful loser over this horizon.

What Changed: AMD's TTM Turnaround

The fundamental picture shifted dramatically. Look at AMD's TTM trajectory:

AMD's gross margins bottomed in mid-2025 and have re-accelerated above 50%, with its most recent quarter (Q2 2026) showing 54.3% gross margin — a huge inflection. Net income doubled over 12 months. The operating leverage AMD always promised is finally appearing.

Meanwhile, ANET continues executing beautifully but saw modest compression:

The Pair Has Worked in Reverse

The ANET/AMD ratio has compressed significantly. AMD's re-rating reflects the market pricing in a genuine AI GPU second-source story — MI300X shipments ramping, data center GPU revenue inflecting, and x86 server recovering. ANET has been steady but hasn't had the same "surprise" factor.

Bottom Line

Over the last 12 months, long ANET / short AMD would have been dead wrong. The pair works brilliantly as a 5-year structural quality trade, but the last year shows that AMD's cyclical recovery and AI GPU ramp can overwhelm that narrative in shorter windows. If you're going to put this pair on, you need conviction that AMD's margin improvement is temporary or that ANET is about to accelerate — because the near-term momentum is firmly with AMD.