Arm for sale?

Arm is perceived as one of the most unique, valuable technology companies on the planet, yet for some reason it is for sale, again. The Wall Street Journal reported on Tuesday that Japanese investment company Softbank is seeking to sell or IPO Arm. Softbank acquired Arm only in 2016, at the time hailing Arm as a big part of the company’s vision for the future. What has changed?

ARM makes ‘blueprints’ for some of the core math that runs most semiconductors used in electronics today. Chip companies license Arm’s intellectual property (IP) and then burn those code blocs into their own chips. This is a huge time saver for those chip companies as they do not have to reinvent the wheel for this fundamental but complex piece of the chip. More importantly, over the years everyone has optimized their product to run on Arm processors, building a huge software barrier to entry. This is not software in the sense of apps, but low-level software crucial to the running of many electronic devices including smartphones, PCs and increasingly in data center servers. Arm’s IP has enjoyed a duopoly as one of only two chip architectures in widescale use, the other being Intel’s x86. Put simply, every smartphone in the world, and plenty of other electronics, relies on Arm.

So why isn’t Arm the king of the world, and why would anyone went to sell that asset? There are a few wrinkles in the Arm story.

First, while Arm charges a royalty for every bit of IP used in a phone, those do not yield huge amounts. True, they have content in every one of the 1 billion-plus mobile phones sold, but often that is only a few cents per phone. While that adds up to a nice amount, when they were public, we would regularly speak to people who were ‘disappointed’ at how little the total was. Their near-monopoly status does not translate into eye-popping profits from an iron-clad pricing model.

Second, they have to spend a lot on R&D developing their core IP. Customers want new features in their chips and Moore’s Law enables ever more powerful chips, and Arm has to keep pace.

Third, they are to a victim of their own success. They grew incredibly rapidly with the explosion of smartphones post-2007. That had a number of cultural side effects. Quasi-monopolies are not known for their customer service. Many customers dreaded working with Arm, complaining of a too-slow roadmap (see point #2 above, that roadmap is expensive to build), and we have heard negotiating a new license with them compared to a visit to the dentist, or the IRS. In addition, their success sparked ambitions for the company to expand into new areas – data center, IoT, AI, automotive, etc. For a variety of reason these have not really panned out with the company launching new product domains but often proving incapable of fulfilling initial ambitions.

Put simply, Arm has some issues with its pricing model and its company culture.

Finally, add to this to the rise of RISC-V, an open source alternative to Arm. At its heart, Arm’s IP is basically software, so a few U.C. Berkeley professors realized that they could build chips on ‘free’ software and created this new open source project. Because working with Arm pained so many customers, RISC-V has been growing steadily for years. Today, every major chip company has at least a few RISC-V projects kicking around inside. These range from science projects to full blown investments in the RISC-V ecosystem. Comparing the two architectures is beyond our scope here, suffice it to say that RISC-V is at least ‘good enough’ and ranks as a credible threat to Arm, albeit not an existential threat at this stage.

Fast forward to 2016, Softbank acquired Arm for $32 billion in 2016, a valuation that caused many to scratch their heads. Since then, both Arm and Softbank have had their share of woes. Softbank’s are well publicized (e.g. WeWork, etc.). Arm’s mostly fall into that category above of underperforming initiatives, but also the occasional headline worthy tale, like their China JV.

We do not want to overly disparage Arm. They have immense capabilities and some incredibly talented people. They have also hard their share of successes in recent years, not least Apple’s shift from Intel to Apple Silicon, which runs on Arm. That being said, scan Apple’s comments on Apple Silicon and you will not find any mention of Arm.

We are not close to the situation or Softbank in general. So our best guess is that Softbank needs to rationalize its holdings a bit, maybe improve its capital structure, and are seeking to sell some assets which are not meeting their goals or whatever criteria Sotftbank is using. Softbank has been telegraphing this change for at least a year. We recall last summer when we got a half dozen out-of-the-blue calls from hedge funds asking our thoughts on Arm’s valuation. Something was clearly afoot.

Softbank has two options: sell the company or spin it off via an IPO. We have not seen any recent Arm financials, but our guess is that it would be very hard for them to achieve an IPO valuation anywhere close to the $32 billion Softbank paid for it. Our best guess is closer to $20 billion.

So maybe a sale, but who would buy it? This presents one of those thorny M&A problems. If a chip company, say Qualcomm or Intel or Apple, bought Arm, that would prompt all those companies’ competitors to find alternatives. Five years ago there was no place to flee to, but today there is RISC-V, a very credible alternative. We cannot rule out a chip company buying Arm, but we imagine few companies would want to spend tens of billions of dollars only to see 80% of the value exit the day the deal closes.

More likely is some form of Financial buyer. This could take the form of a traditional private equity fund. There is a case to be made that Arm could benefit from a ‘disciplined and focused’ owner to prompt some improvement. However, we question whether those entities would be able to justify a valuation sufficient for Softbank. Alternatively, Softbank sits at the nexus of a number of sovereign wealth funds, who have some incentive to see that company shore up its position. If we were to be overly imaginative, we could see a combination of Global Foundries with Arm. GloFo is a chip manufacturing foundry company, so we could paint a PowerPoint slide with some creative synergizing. GloFo is owned by the Emirati sovereign wealth entity Mubadala which has close ties to Softbank. To be clear, this is unlikely, but it can be fun to dream up pitch decks.

Another theoretical possibility is a consortium of several chip companies contributing to a buyout under a private equity fund. This would present an arms length relationship sufficient to retain Arm’s customers, but also allow some upside to the valuation provided by a private equity firm flying solo. This is similar to what happened with Dutch SemiCap equipment maker ASML, a key strategic player in the chip manufacturing supply chain who needed a few billion dollars to keep Moore’s Law moving.

One thing we can say with certainty is that no Chinese company will be the buyer, or probably even allowed into a syndicate. As we have detailed, China has big ambitions for chips and unlimited funds to meet those, but geopolitics will keep them out of the mix. That being said, it will be interesting to see how China’s government will treat this deal. Given that Arm China is technically a separate company, one could argue that China’s anti-trust regulator would have no jurisdiction over a deal. One more convolution of that regulatory wrinkle.

In the end, Arm is incredibly important to the industry but sits at a precarious strategic point.

Photo Credit: D2D Advisory, another auction process.

2 responses to “Arm for sale?

  1. Pingback: Death, Taxes and IP Licenses | DIGITS to DOLLARS·

  2. Pingback: Nvida to Acquire Arm – first impressions | DIGITS to DOLLARS·

Leave a Reply