A great deal for Softbank. Masayoshi Son's group, which still controls 87.5% of ARM's capital, acquired it in 2016 for $32 billion. However, the British company's enterprise value is now almost four times that amount.
The growth performance is real—ARM's revenue has doubled in five years and quadrupled in ten years—but it remains difficult to justify a valuation that represents such a huge multiple of sales and annual profit.
Analysts are therefore betting on continued—or even accelerated—revenue growth, as evidenced by their three-year forecasts. This comes at a time when ARM is embarking on a new strategy of producing its own processors.
This transition—from a lucrative business model of monetizing patents through royalties to an integrated manufacturing model—is a double-edged sword, as it risks putting ARM in direct competition with some of its customers. So, it looks like a strange game of three-cushion billiards is in store.
Analysts nevertheless welcome the vision, which is becoming increasingly clear following Softbank's acquisitions of Ampere and Graphcore. The bets seem risky here too: Ampere has no commercial activity, while the technology developed by Graphcore, which is taking on Nvidia head-on, is still unproven.
That hasn't stopped Softbank from paying $6.5 billion for the former and between $500 million and $600 million for the latter, in the hope that their expertise will complement ARM's capabilities and position it strategically in the AI race.
While impressive on paper, all of the ambitions associated with the Stargate project remain subject to considerable uncertainty. However, Masayoshi Son is certainly no novice. For more on this topic , see SoftBank Corp.: All about chutzpah.
