Arm FQ4 24 – Volatile Visibility.

The story will remain v9 for some time.

• Arm reported good results where it made good on its FY24 forecasts, but the shares fell as the guidance for the coming year was not quite up to expectations meaning that the shares will now remain a push and pull between what multiple should be paid for AI and its relatively consistent earnings.
• FQ4 revenues / EPS were $928m / $0.36 nicely ahead of forecasts of $866m / $0.36.
• This represented 47% YoY for FQ4 where the main story is the adoption of the v9 architecture within the company’s product lines which grew to 20% of revenues compared to 15% of revenues in FQ3 24.
• v9 is a more advanced architecture and, as such, it attracts a higher license price and higher royalties on the chips that use it.
• Hence, on top of the general AI growth trend, Arm is seeing an improvement in the product mix which if it continues at this rate will be a major growth driver for some time to come.
• On the AI front, Arm is seeing traction even if it will take a little time to translate into the largest revenue segment of the company.
• Nvidia, Google and many others are all using Arm cores in the control systems that make up their AI chips meaning that there is finally proper engagement in the cloud despite a number of stops and starts over the last decade.
• Ironically, the biggest potential delta on Arm’s short-term performance was glossed over which I think is the potential for Windows on Arm to finally work properly and for Arm to gain substantial share in laptops from Intel.
• There was one tiny mention in the shareholder letter and a single question on the conference call about this topic.
• I suspect that part of the reason for this is the current dispute that is ongoing between Qualcomm and Arm over the royalties that are payable on Qualcomm chips that use its new in-house Oryon processor.
• This is unfortunate as if Qualcomm and Arm were pushing this together, I think that there would be greater awareness and hence, fewer barriers to adoption.
• The initial signs and claims are looking very promising, but the experience will have to be seamless with no limitations being placed on the user from using Arm as opposed to Intel.
• The only thing the user should notice is that while the experience and performance are the same, the same battery lasts much longer.
• Apple has brilliantly managed this transition and now Qualcomm and Arm need to ensure that the more difficult Windows transition is just as seamless.
• I think that the dispute may well be getting in the way and the best interest of both companies would be to settle the dispute and for Qualcomm and Arm to jointly push this proposition.
• If there is going to be upside in Arm’s numbers for the coming fiscal year to March 2025, I suspect that this is where it is going to be found.
• Against a backdrop of a good set of results and strong medium-term drivers of revenue growth, the only problem with the numbers was the guidance.
• Here FQ1 25 revenues / EPS are expected to be $875m – $925m ($900m) / $0.32 – $0.36 ($0.34) which was in line with estimates of $897m / $0.34.
• FY 2025 revenues / EPS are expected to be $3.8bn – $4.1bn ($3.95bn) / $1.45 – $1.65 ($1.55) which is also in line although below the midpoint of estimates at $3.95bn / $1.53.
• Normally, this would be brushed off, but AI is supposed to be driving a consistent cadence of beats and raises, which was met with disappointment.
• Arm’s shares closed at a 2025 PER of 69.2x which is far higher than even Nvidia or Microsoft explaining the market’s skittishness which sent the shares down 8% in after-hours trading.
• The volatility of the share price is not helped by the fact that only 10% of the company is available to be traded as SoftBank is showing no signs of selling down its stake into the market.
• This means that liquidity is far from what it should be ensuring that the shares will react more violently to both and bad news.
• Even with this decline, the good news massively outweighs the bad as the shares are still up 90% from the IPO and Arm has taken its place in the AI pantheon which is something I thought would be difficult to do.
• This combined with the fact it has a surprising level of visibility in terms of its revenues and profits means that it is likely to hold a premium valuation for as long as the AI bubble lasts.
• There is no real sign of this bursting as the large cloud companies are all substantially increasing their investments in AI and valuations of the larger generative AI players remain at all-time highs.
• Hence, I think that the shares of Arm are going to remain volatile although in terms of the overall valuation, it is likely to hold up for a while yet.
• I don’t own Arm, preferring instead to own Qualcomm which trades at less than ¼ the multiple and is exposed to many of the same trends albeit not as profitably.