Can Arm possibly justify its huge value bump?

In what is likely one of the biggest stories in tech for the week, Arm Holdings, a company that only recently returned to being a public company in September, saw a huge spike in its stock, nearly 60% at one point, after its quarterly earnings report. As recently as Halloween Arm was under $50, yet broke $120 on Thursday.

To most observers, the sudden valuation increase doesn’t match up to the earnings statements and forward projections. The company reported $824M of revenue for the quarter, up 14% YoY with two unique business models covering it all. The licensing business, where Arm gives access to its various IP for designing CPUs, graphics, AI engines and more, brought in $354M of that revenue, up 18% YoY, while the royalty business, where Arm earns a fee for every physical chip sold using its IP, generated $470M in revenue, up 11% YoY. Next quarter revenue pointed up $100M over previous guidance and full year guidance went from $3.0B to $3.2B in revenue.

In all, an impressive “beat and raise” that the tech field absolutely loves as it tends to be a leading indicator to a larger trend.

But after hours, and in the first handful of trading hours the next day, Arm was up more than 60% at one point, closing at shy of 48% above the previous day. How do you explain that? And is this something that points to a sustainable growth period for Arm?

Some of the worry about Arm’s long-term prospects were that it was risky and aligned too closely with the cyclical smart phone market across the world, and in particular in China. But the quarterly results from the royalty division show us that Arm isn’t dependent on the smart phone market to grow.

Of the total royalty revenue from this quarter, only 35% of it was sourced from smart phone chips. Back in 2016 that number was around 70%. Even when the global smart phone market is down, Arm had royalty revenue increase 11% thanks to the company’s expansion into servers, automotive, and even consumer laptops. And if the smartphone market has a resurgence thanks to AI and general upgrade cycles, as indicators from companies like Apple and Qualcomm indicate, then Arm has even more opportunity for this number to grow.

But to justify the huge spike in Arm’s stock price this week, we need something more than just “we see some evidence of diversification” don’t we? What is the bull case that shows Arm can keep this going?

The key to me is the royalty revenue stream and its growth over the next 1-3 years. The most advanced architecture IP that Arm offers is called “Armv9”. It represents just 15% of the royalty revenue this past quarter. This Armv9 architecture royalty rate is double that of the previous generation Armv8 based IP, cores, and other products. As more of the Arm product mix moves towards this more advanced, more capable, and more profitable Armv9 architecture, then royalty revenue for the company stands to increase at a higher rate than individual unit growth.

It's the growth areas for Arm that are most likely to use the Armv9 architecture as well. Because it offers the best performance of the Arm product stack, it is utilized in new server chips from Nvidia, Amazon, and even Microsoft as part of their custom silicon projects. It is also utilized by more of the automotive segment as that market continues to see need for higher performance and processing of on-vehicle data and AI. Even the growing PC laptop market, driven by new products like the Qualcomm Snapdragon X-Elite coming later this year, use Armv9.

Arm partners and customers ship north of 7 billion chips each quarter (!) but the average royalty per chip is measured in single digit cents when you do the math on $470M in royalty revenue for this past quarter. The vast majority of those shipped products are using older architectures, smaller designs, and are designed into IoT products around the world. But the royalty on a single high performance, custom server chip that uses the Armv9 architecture could be in the tens or hundreds of dollars of royalty. It’s easy to see how the transition to a higher mix of Armv9 products is the way to drive revenue.

Arm knows that to continue to get that royalty rate (or grow it further) it can’t stand still and needs to continue to improve its product capabilities. A recent analyst report referencing a project called “Blackhawk” that aims to be “the highest-performance CPU core for smartphones later this year” directly from Arm is an example of this strategy. More advanced technologies like that are part of the uniqueness of the company’s business model that strongly incentivizes R&D investment in new IP.

Because Arm is essentially the corporate personification of an entire ecosystem, many of them actually, the sudden meteoric rise in value has some interesting implications for the rest of tech. The custom silicon market, most notably huge cloud companies like Microsoft, Google, and Amazon, along with dozens of smaller design houses, are building custom AI accelerators to compete with Nvidia and AMD GPUs, driving the most profitable segment of Arm’s royalty business. Even Nvidia is part of that dynamic, with its Grace Hopper Superchip integrating both GPU cores and high-performance Arm cores.

A company like Qualcomm is an interesting case study as well, as its similarity to the path that Arm is on is interesting. As it tries for its own corporate transformation from a communications company to a compute company, growth in the smartphone, automotive, and laptop markets, for Arm are a sign of growth for Qualcomm. Despite some corporate tension, they seem tied at the hip in many ways.

There are other interesting dynamics I want to dive into at another time, like how we might be able to map the growth in Armv9 architecture success to the detriment of classical x86 computing from Intel or AMD.

For now, though surprising to most investors (clearly), it seems that Arm is getting the credit it deserves for the current, and future, state of the computing market.