This is such a bad time for a US$115 billion tech deal

TIMING is everything when you’re trying to pull off a mega deal. And it’s hard to think of a worse moment for semiconductor designer Marvell Technology to unveil a takeover of laser specialist Lumentum at a US$115 billion enterprise value.

After a month in which the market has become increasingly jittery about the economics of artificial intelligence, this attempt at one of the largest tech industry takeovers has gone down terribly.

Evidently, now is not a great time to be doubling down on data centre kit.

Marvell was worth US$221 billion before Bloomberg News revealed last Wednesday (Oct 28) that the pair were in advanced talks on a tie-up. Investors were immediately shaken, and the deal’s formal announcement on Monday has only made things worse.

The buyer’s stock price is down 30 per cent since its interest emerged, against a Nasdaq 100 that has fallen roughly 5 per cent.

Under the all-share terms, the two companies are to be mashed together so that Lumentum shareholders would own a third of the enlarged firm.

Based on the undisturbed stock prices, the share-swap ratio gave the target company a 24 per cent premium.

But Marvell’s stock-market plunge means the deal is now worth less than US$800 per Lumentum share – 10 per cent or so below where the laser company was trading before news of the talks broke last week.

This is an inauspicious beginning for a transaction that needs approval from both sets of shareholders, as well as regulators. And yet there is some solid technological rationale to a combination, as analysts at Morgan Stanley have pointed out.

To the uninitiated, it’s hard at first to get your head around why exactly. Marvell-Lumentum’s stated aim is to create “a broad full-stack connectivity portfolio that is intended to enable end-to-end extreme co-design, high-volume production of fully optimised, pre-validated optical solutions”.

Put in simpler terms, this means Marvell will be able to incorporate optical technology within its semiconductors from the earliest design stage.

That could bring cost and energy-efficiency benefits. Optics and chips are increasingly converging, indicates Raymond James & Associates research.

Moreover, there are some chunky cost savings from joining forces, put at more than US$1 billion annually. It’s not clear when these will be achieved.

But they could in theory justify the US$20 billion premium if the market valued them on the kind of profit multiples recently commanded by the two stocks.

Cyclical conditions and competing customers

Why the bad reaction? The chief reason may be that Lumentum shares had already risen by more than four times in the 12 months before deal talks broke, comfortably outperforming their suitor’s.

Marvell is going all-in on an acquisition when the price of its target is high and investors are feeling distinctly risk-averse in general about the AI spending boom.

This may be a deal at “peak cyclical conditions”, as the Morgan Stanley analysts put it.

Compounding these worries are the uncertainties about the impact on both sides’ commercial relationships, given that Marvell has customers that compete with Lumentum, and vice versa. Marvell boss Matt Murphy stressed that clients’ reaction has been positive.

Lumentum may well be sensible to seek to sell itself at a premium when it is riding high. But by taking 100 per cent stock rather than a cash-and-shares mix, it is fully exposed to the negative reaction from Marvell’s shareholders.

Both sides now have a tough job to sell this deal. As for other bosses contemplating tie-ups in this industry, investors are sending a simple message: Not now, please. BLOOMBERG