Arm Holdings IPO is promising to be the biggest IPO in 2023. The Cambridge-based company is offering 95.5 million shares at an offering price range between $47 and $51. This puts the Arm IPO at a valuation between $48 and $52 billion. Arm expects its shares to trade on the Nasdaq under a ticker symbol “ARM”.

Softbank purchased a 25% Arm’s stake from the Vision fund with a $64 billion valuation previously. Intel, Nvidia and other big tech companies are lining up to be anchor investors in the Arm IPO as well.

But, why does the Arm IPO have such a high price per share? Is it worth investing in the Arm IPO at such a valuation? Let’s tackle these questions.

What Does Arm Holdings Do?

Modern chips in electronic devices must have a certain set of instructions. What do I mean by that? When you type a word in MS Word, there are thousands of operations going on in a split second behind the scenes.

First, the computer processes you pressing the buttons. Then, it recognizes which letter you pressed. Next, this letter must be stored somewhere in memory. After that, it processes it and shows what you are typing on your screen. As you see, all these operations must have rules to execute before a letter you typed shows up on your computer display. All these rules and instructions is what powers modern devices.

Arm is one of the designers of these set of rules that are called instruction set architecture (ISA). The other two are x86 architecture by Intel and RISC-V. RISC-V is a relatively new architecture since 2015. x86 has been around since 1978.

Unfortunately for Intel, x86 architecture is power-hungry and heats easily. Moreover, chips designed with x86 architecture can be expensive. Conversely, Arm architecture has low cost, low power consumption and low heat generation. This makes it perfect for light and portable devices, such as smartphones, laptops and tablets.

Arm Holdings: R&D-Focused Company

Note that Arm is not building actual CPUs. It designs instructions for chips, but does not actually produce them. Arm is a high-intensity research and development (R&D) and light-asset company.

Arm’s main assets are human resources, such as scientists and engineers. If you look at Arm’s cash flow and balance sheet, you will notice little capital expenditures. But, you will see massive R&D costs, which amount to over 40% of its revenues. You can think of R&D as Arm’s CapEx in a sense.

Table 1. Arm Holdings select financial metrics (in millions USD dollars)

Arm is very profitable by generating sales from recurring royalty and license revenues. Customers license Arm chip architecture and pay from 1% to 2% of the value of each chip sold.

Arm’s Monopoly in Handheld Devices

There are over 250 billion of chips shipped to date that use Arm architecture.

Growth turned exponential since 2010 with smartphones becoming mainstream. Amazon, Meta, Microsoft and others use Arm-based architecture in their cloud infrastructure, primarily for CPUs. Moreover, Apple has transitioned its CPUs to Arm architecture since 2020.

Arm is a monopoly when it comes to handheld devices. 99% of all smartphones use CPUs based on Arm architecture. Over 50% of Arm’s revenues come from smartphones and consumer electronics.

Back in 2012, Intel had to back off from developing chips for smartphones as it was getting nowhere. This attests to how much effort it takes to produce a commercially viable product. That’s a big competitive advantage for Arm.

Arm Holdings Financial Performance

Let’s take at Arm Holdings’ financials.

Table 2. Arm Holdings’ Financial Performance (in millions USD)

Arm revenue grew by about 7% for the past 8 years on average, while its operating profit stagnated. This is a red flag. Profitability declined as a result of growing R&D expenses.

There is this graph that Arm has in its IPO prospectus showing the cost of design for more and more advanced chip architecture.

The gist of it is this. It becomes more expensive for Arm to produce products the smaller the chip gets. If there is anything you can remember from this article is this fact. The market at some point starts caring about profits. And if Arm has nothing to show for its revenue growth, the valuation must inevitably decline.

Arm IPO Valuation

Arm has virtually no debt. Assuming a $50 billion valuation, here are Arm’s relative valuation metrics.

Table 3. Arm Holdings and Nvidia Relative Valuations

With a quick comparison to Nvidia, we see that Arm has a very rich valuation, especially based on EV/EBIT. Arm is on par with Nvidia, but with less impressive metrics to show for it. Even if past is not a predictor of future, we see that Arm does not even closely match Nvidia in terms of historic growth.

Arm IPO Risks

Nvidia attempted to buy Arm a few years back for $40 billion. But why has the valuation increased to almost $50 billion since then?

1. AI Hype

There is a significant hype surrounding anything related to AI. Arm positions itself as if it will benefit from the current AI trend. Recently, Nvidia built Grace Hopper. This is its first platform to handle massive AI workloads with Arm CPU architecture at its core. You may get the impression that Arm will be at the center of the AI trend. But will it really profit from it that much? Let’s take a closer look.

Consider the NVIDIA DGX H100 supercomputer, which is a state-of-the-art machine when it comes to training AI models. H100 costs around $270,000 according to Semi Analysis. Of this 270K, over 70% of the cost goes towards GPUs and associated baseboards. What about the CPU that potentially can use Arm designs? The answer is less than 10%. Arm gets 1-2% of the value of the CPUs used in the Grace Hopper. So, it is less than 0.1%.

Even if we assume that chips with Arm architecture show up elsewhere in the Grace Hopper, we see that it is Nvidia which profits from the AI trend, but not so much Arm. CPUs that use Arm architecture are good for making predictions based on trained AI models. But as far as heavy lifting in algorithm training, it is all done by Nvidia’s GPUs. No wonder that Nvidia is flying high these days.

2. Arm China Risk

Another red flag I noticed was Arm’s exposure to China with about 25% sales. Escalating trade wars do not bode well for Arm. Moreover, historically speaking, there were conflicts between Arm Holdings and its Chinese subsidiary called Arm China. For instance, it took Arm over two years to replace a CEO of its Chinese subsidiary.

What is surprising is that Arm Holdings does not control Arm China. And yet, 25% of its revenues depend on it? Moreover, the amount of royalties Arm earns from PRC depends on financial information provided by Arm China. Arm further mentions in its IPO prospectus that there were issues obtaining timely and accurate information from Arm China. It seems that there could be questionable things going there. This is a massive red flag for 25% of Arm revenues.

3. Consumer Electronics Cyclicality

As I mentioned, over 50% of Arm’s revenues are dependent on how consumers electronics industry performs in a given year. Sales of gadgets, especially laptops and phones, are notoriously cyclical. It is not clear where we stand now, but it could be a down cycle for all we know.

4. RISC-V Competition

Finally, there are signs that RISC-V chip architecture may gain traction in China and Europe. RISC-V emerged as Arm’s competitor in 2015 and it is an open-source architecture.

It is an open question if it will be able to compete with Arm though. But still, it could. China is very eager to shed its dependence on anything it cannot control, especially American or European technology. It is possible that at some point, RISC-V will replace Arm’s architecture in certain case uses.

Concluding Remarks

The final question is this: is it worth buying into the Arm IPO?

Looking at the big picture we see that the AI trend is strengthening in investors’ minds. But, Arm is not really benefiting from it in terms of generating strong profits. Its R&D cost devours a large chunk of its margins.

Do not get me wrong, it is a terrific company. But, is it worth paying such a price at around a $50 billion valuation? I would say that for me to even closely consider buying Arm’s stock, it has be to at most a $30 billion valuation. This means that the Arm IPO price per share should be no more than $30. This is what a prudent investor should pay to get a margin of safety.

“It’s far better to buy a wonderful company at a fair price than a fair company at a wonderful price”.

Berkshire Hathaway Letter to Shareholders