Arm vs. Qualcomm – Preamble

A situation that makes little sense.

• The dispute between Arm and Qualcomm is one of the more bizarre that I have seen in my time, but it is descending into the usual tit-for-tat which is likely to end in a settlement at some point.
• What is unusual in this case is that Arm opened hostilities right before its IPO (where valuation will be an issue) and the one thing markets do not like when it comes to valuation is large legal uncertainty.
• To be clear, I have not yet spoken to either company about the facts or merits of the dispute and this represents my view based on what I have seen so far and will almost certainly be revised over time as more facts come to light.
• I suspect that at the heart of this dispute lies SoftBank’s desire to achieve a good return on the investment that it made when it purchased Arm for $32bn in 2016.
• The plan then was to massively invest in IoT, dominate the market and then return Arm to the market at a much higher valuation (see here)
• Unfortunately, market events and the pandemic got in the way and now those investments have now been split off from Arm and, having failed to sell Arm to Nvidia, SoftBank is down to plan C which is to return Arm to public ownership.
• This will now almost certainly take place on Nasdaq where Arm is expected to fetch $30bn – $70bn where the size of the spread indicates that no one really has any idea what the market will pay for it.
• A sale to Nvidia would have been a great solution for SoftBank as the run-up in Nvidia’s share price would have resulted in SoftBank achieving a valuation in excess of $70bn and a very healthy return on investment.
• Unfortunately, the regulatory pressure made it too difficult for Nvidia to complete the acquisition which it abandoned n February 2022.
• Behind the pressure were Qualcomm, Intel and Microsoft and I suspect that SoftBank thinks that Qualcomm is largely to blame for the deal failing.
• This adds another dimension to the dispute which makes it even more difficult to evaluate as it involves the large personalities at the top of the technology industry.
• At a very high-level Arm has a case and Qualcomm has a pretty robust defence.
• The complaint essentially boils down to Nuvia which consisted of a series of highly talented engineers but no product for which Qualcomm paid $1.4bn in January 2021
• These engineers originated from Apple where they worked on its processor cores which Qualcomm will now use to close the gap on the M-series of processors.
• Nuvia had an Arm license but, because it was a new company, it would have ended up paying a higher royalty rate to Arm when it sold its chips than Qualcomm does for similar products.
• Qualcomm on the other hand appears to want to use its own license to ship products with Nuvia-designed cores, which will attract lower royalties.
• Put this in the context of Arm needing to maximize its profits and growth to generate a healthy return at or after the IPO and one can begin to see where this lawsuit is coming from.
• However, the response from Qualcomm has raised the possibility that Arm is seeking to completely change its business model and, while this could explain a motive behind the lawsuit, it is also logically inconsistent.
• Public filings led to an article by the FT (see here) which claimed that Arm intends to stop charging chip companies for its IP and instead move to charging device makers a royalty based on the wholesale price of the device.
• This is not new and is how developers of intellectual property (IP) used in handsets have earned revenues for years and it is a very well-established practice.
• This would massively increase Arm’s revenue given that the price of the chip and the price of the handset are radically different and would almost certainly provide the kind of lift in valuation that SoftBank is looking for.
• However, if chipmakers have already paid Arm for the IP then the concept of patent exhaustion (which is now better established in US law) would preclude Arm from also charging device makers (as the royalty has already been paid at the chip level).
• This means that Arm would need to stop charging chipmakers for royalties and instead move to device makers.
• The problem here is that Arm already has long-term contracts with chip makers meaning that these would need to be cancelled, potentially creating a motive for Arm to go after Qualcomm (i.e. to cancel the long-term contract).
• However, if it no longer intends to charge Qualcomm royalties, this lawsuit makes no sense as it appears to boil down how much Qualcomm will pay Arm for Nuvia cores which appears pointless in this context.
• The latest iteration of this is the idea that Arm intends to build and sell its own chips for device makers which puts it into direct competition with its customers and destroys its independence which was a key element in its success.
• Arm has built chips before for testing and development purposes and I suspect that this is more of same and that there is nothing to see here.
• I have real doubts about whether Arm intends to change its business model in this way and I think that at the end of the day, this boils down to a dispute over how much Qualcomm will pay Arm for its IP.
• The net result is that we have another bitter lawsuit on our hands which will make for a great spectator sport but, as ever in these situations, the only real winners are the lawyers.