Arista’s Very Good, Very Bad Year

The networking company is drowning in the best demand its CEO has ever seen, and that’s precisely the problem.

If you held Arista Networks (ANET) over the past year, you’re probably feeling pretty good. A stock that climbs 83.1% while the S&P 500 manages a mere 27.9% is the kind of thing that pays for a nice vacation. But before you pack your bags, look around the neighborhood. Peers like Dell and HPE posted returns of 277.9% and 183.9%, respectively. So what gives? Why the stellar, yet somehow lagging, performance?

The answer is a paradox. Arista is facing a problem most companies only dream of: demand is simply too good.

“The Best I’ve Ever Seen”

You don’t often hear a CEO with decades of experience make a statement like this, but Arista’s chief executive was unequivocal on a recent call: “Our demand is actually the best I’ve ever seen in my Arista tenure.” That’s not just talk. The company has been busy raising its forecasts, now aiming for $11.5 billion in revenue for 2026. The engine for this is the all-consuming buildout of AI infrastructure. Arista now expects to sell $3.5 billion worth of gear into AI projects this year alone, effectively doubling its AI sales annually.

This is the story that powered the stock’s ascent. Investors saw a company perfectly positioned to provide the high-speed plumbing for the AI revolution, and they bought in.

A Two-Year Traffic Jam

Here’s the catch. Arista can’t build its products fast enough. Management was blunt, stating, “Our demand is outstripping our supply this year.” This isn’t a minor hiccup; the CEO described the supply chain crunch as a “1- or 2-year phenomenon,” with shortages hitting everything from wafers to optics. The company is paying more to secure what parts it can, which it expects will create “gross margin pressure.”

The result of this logjam is a ballooning operational liability. Arista is now sitting on a total deferred revenue balance of $6.2 billion. That’s money for products that have largely been shipped, and in many cases are physically deployed, but cannot yet be recognized as revenue because hyperscale customers must formally sign off on complex AI network deployments before Arista books the sale, a process that can take as long as six to eight quarters.

That massive deferred revenue figure is the physical manifestation of Arista’s unique operational challenge. It’s a testament to incredible demand, but it’s also highlights a significant multi-quarter fulfillment cycle.

Which leaves just one question for anyone holding the stock: can Arista efficiently convert that massive backlog into recognized revenue before customers look to alternative suppliers?

Also see our other analysis, Dell Stock And The Best Kind of Problem

Where Does This Fit In Your Portfolio

Chasing single-name moves is its own kind of risk. The Trefis High Quality (HQ) Portfolio takes the other side of that bet: 30 quality names, sized and re-balanced with discipline, and a track record of outpacing the S&P 500, S&P Mid-cap, and Russell 2000.