A PROFIT of $20 billion in seven years is not bad going. It’s enough cash to build and fit out one of Taiwan’s most advanced microchip manufacturing plants — or fund the UK government’s long-delayed semiconductor strategy 16 times over.
Yet SoftBank’s two-thirds mark-up on Arm, the chip designer it acquired for $32 billion seven years ago this week and has been trying to sell for the past four years, is far from exceptional when set against the growth seen elsewhere in the industry.
Over the same period, the PHLX Semiconductor Sector index has risen more than fourfold, and Nvidia — Arm’s one-time suitor thrilling investors with its potential to power the artificial intelligence era — is 30 times more valuable than in 2016.
This initial public offering, the year’s largest and the most closely watched in London precisely because it is not happening here, was never required to shoot the lights out. If it merely switched them back on it could be judged a success for Wall Street investment bankers who have been starved of juicy transactions and fees.
Still, Masayoshi Son, the magpie-like SoftBank founder, will be hoping that the price pops when trading starts on Nasdaq next Thursday, so that the 91% stake he retains appreciates further to fund another acquisition spree.
SoftBank’s IPO record isn’t great, a reminder that struggling new issues are by no means peculiar to London. What is notable about Arm — founded 33 years ago in Cambridge in a joint venture between Acorn Computer, Apple and its production partner VLSI — is that it does not promise the explosive growth that tech’s Next Big Thing very often does. But it turns a profit. And the fact that the industry’s leading lights all want a piece of the company speaks to its unique position amid warring device makers and geopolitical rancour.
The $735 million which the 10 including Samsung, Intel, Apple, Alphabet and Nvidia will collectively invest is a mere bagatelle given their cash positions, but it protects their interests. In fact, it is less than the profit Apple had turned when it sold its last share in Arm 20 years ago.
Why are they bothering? Arm’s secret sauce is a rulebook, known as an instruction set architecture (ISA), whose few thousand instructions can be configured into four billion possible encodings that determine how a device’s central processing unit (CPU) is controlled by software. It has evolved into a foundational technology that is licensed 1000 times a second, on which more than 15 million developers design software. Best for the giants to keep their hands in.
Aside from them, Arm must please investors looking for growth, which chief executive Rene Haas says can still be squeezed from the sluggish smartphone space, as well as by gaining share in designs for computer server chips and the car industry.
It was never meant to be this way. In trying to win approval for the doomed Nvidia takeover, Arm petitioned the Competition and Markets Authority, warning that it should not be left to the mercy of “unsentimental” public markets that would demand “profitability and performance” and no doubt strategic changes and cost-cutting.
I am not sure I would buy anything 28 banks are employed to sell, but I suspect Arm will be fine. And I’m sure it could have prospered in London. We must learn the lessons from its flight so that our future world beaters need never look abroad for capital.
James Ashton is Chief Executive of the Quoted Companies Alliance. His story of Arm, The Everything Blueprint, is published by Hodder & Stoughton.
