It’s awkward when things have gone awry, the investigations that could deliver a killer blow are just dragging on, and everyone seems to recognise it’s over but you.
I’m talking, of course, about Nvidia’s deal to buy chip design company Arm from Japan’s SoftBank, as close to a dead deal walking as you’ll ever find.
Its demise is starting to feel inevitable. The only questions are how long the end will take, and what happens next?
The answer to the first, amid stories that the parties are preparing to abandon the deal, is probably: a while yet. A deal that was always going to be controversial, because it handed Arm’s neutral, open to all licensing model to one of its customers, is bogged down in regulatory investigations around the globe. The US Federal Trade Commission last month sued to block it; authorities in Europe and China are scrutinising the deal.
Ironically it could be the UK, which welcomed the 2016 acquisition of then London-listed Arm by SoftBank as a vote of confidence in post-Brexit Britain, that delivers a crucial blow. The Competition and Markets Authority in November referred the deal for an in-depth investigation, both on antitrust and national security grounds, in a first for a regime now superseded by new government vetting powers.
Quite why Arm’s sale to Nvidia, a California-based chipmaker, threatens national security when its acquisition by SoftBank, a Tokyo-based investment behemoth, was “in the national interest” remains a mystery. But the CMA’s first report was sufficiently damning on the chances that remedies or behavioural commitments could address competition concerns to think the deal is in trouble regardless. It is set to give provisional findings in March, with a final decision in May.
The deal will limp on in the meantime. The agreement struck in 2020 doesn’t lapse until September 13, two years after it was signed, or until it is blocked by regulators. All parties have committed to bring best efforts and a sunny disposition to getting it done until then — and would like to emphasise how enthused and engaged they are about that task. For a start, SoftBank is sitting on $1.25bn of Nvidia’s cash, paid over at signing, which could be refunded if it fails to live up to its many and various undertakings under the original deal.
Where would the deal’s collapse leave Arm? Hardly thriving. The self-interested commentary sent to regulators by the deal parties makes a fair point that Arm now isn’t Arm then. In 2015, it had top-line growth of 22 per cent; it has eked out mid-single figures in recent years. Its operating costs have exploded, in part because of hiring to hit the doubling of its UK workforce that was a commitment of the 2016 deal. The SoftBank pitch to pile into the Internet of Things hasn’t yielded results.
The prescription is lashings of investment to make Arm competitive in new areas such as chip designs for data centres, if you ask Nvidia. Or a refocusing and rebuilding of the business based on its traditional strengths, if you ask others. Either way, it looks like a turnround situation, a far cry from the jewel in the crown status that Cambridge-based Arm has traditionally enjoyed in Britain.
Would Arm return, triumphant, through an IPO to the London market? Perhaps. It’s neither here nor there that 95 per cent of its revenues come from outside Europe, let alone the UK. The FTSE is home to plenty of international companies. But Arm was always more global than many admit, founded with British tech but US money from Apple and VLSI Technology. It has become more so since 2016: most of its executive team are now based in California.
“All roads lead to the US,” commented one banker when asked what would happen. “The big unknown is the UK government.” Which suggests that a masterful combination is possible: that having blessed Arm’s 2016 takeover in a fit of jingoistic flag waving, the government could in effect block its sale six years later, before having to launch a charm offensive to try to win the listing back again.