Semiconductor Barna Jam
I’ve been watching aghast from the sidelines for many weeks as an M&A storm gathers in the semiconductor industry. Each announcement has been more astounding than the last, but after the eruption of Intel as a potential player on Friday last week, I have finally been moved to words. And the words are primarily: this is entirely crazy. The fish eats fish metaphor extends to four levels. On the face of it, it appears NXP might be swallowed by Qualcomm, which might be swallowed by Broadcom, which might be swallowed by Intel.
And the numbers keep escalating. Qualcomm is now bidding $44 billion for NXP, up from an initial $38 billion. Broadcom is now offering $117 billion for Qualcomm, down from about $121.5 billion before the latter raised its bid for NXP. And no one knows what Intel might have to drum up to gulp down the entire chain of fish, but whatever that number is, it would certainly be counted in hundreds of billions. Broadcom’s current market capitalization is just north of $100 billion, before it swallows any smaller fish. For that kind of doh, you could buy a small country in Eastern Europe or Latin America.
Never mind the Byzantine regulatory labyrinth any of these transaction would have to navigate, only the first, Qualcomm’s potential acquisition of NXP, is even friendly. Hostile takeover seems to be the order of the day, the hostility perhaps siphoned from elsewhere in our society, which seems to have an abundance of it at the moment.
But why would anyone think a hostile takeover of a technology company would work? People even remotely acquainted with the industry understand that the real assets go home at night and might or might not turn up in the morning, depending on how they feel about who’s running the show. And, you can’t rely on the current bag of technology in a tech company’s portfolio. The industry moves so fast that what matters most is what’s in the pipeline next year, the year after, and five years out. All those plans go up in vapor if the people who figured them out are disgusted with management’s behavior and take jobs elsewhere. Tech workers are not slaves to be bought and sold at will. They need to be wooed — individually — to remain loyal. Anyone who thinks otherwise is fooling themselves.
So, what is all this merger nonsense? Well, there are probably too many moving parts for me to conceptualize the full panorama, but there are some fixed points on the map.
Qualcomm’s reasons for wanting to buy NXP are fairly straightforward. Between them, the companies would have an important — perhaps dominant — position in the burgeoning smart automobile market. Among the various Internet of Things (IoT) verticals, automotive is the largest, most coherent, most concentrated, and nearest at hand. The merger makes sense from a purely industrial and financial standpoint.
Broadcom’s reasons for lusting after Qualcomm are more opaque. Primarily a financially oriented company under CEO Hock Tan, Broadcom’s pattern seems to be modeled on the mid-1980s leveraged-buyout craze (à la Barbarians at the Gate: The Fall of RJR Nabisco). Basically, you enlist a pile of hot money from all your private equity pals and whichever banks are willing to play and go after a big company. Once having breached the walls, you throw out management, shut off the R&D spigot, and squeeze enough cash out of SG&A expenses to impress the market and pay your raiding party back for its help. Everyone thinks you’re a genius until the raiders eject the former company’s husk onto the side of the road at the edge of town with no blood left in its veins. And even then, traders on the Wall Street side of the transaction don’t really care.
So, is it that Hock Tan wants to run the company likely to be dominant in the up-and-coming smart auto market? Not half likely! Other theories: perhaps he’s in league with Huawei and the Chinese government to cripple Qualcomm in the 5G market, opening a way for Huawei to take over. The 5G market, utilizing the next generation of mobile technology, is in fact a bigger near-term prize than automotive. Maybe it’s more like Apple, which is locked in a bitter lawsuit with Qualcomm, put Broadcom up to this and is perhaps even supplying funding somehow indirectly. Apple certainly has enough money to pay Broadcom to make Qualcomm go away.
Finally, there’s Intel. The sad story of Intel is that it mostly missed the mobile market and stands a good chance of forfeiting the next big thing, the IoT market, as well. Resting on its laurels from its lock on the PC business, Intel embraced the wrong philosophy for the mobile market (choosing performance over low power consumption) and got left behind. The consolation prize, and a good one at that, was the server market, which the company now dominates mightily.
But where to from here? Intel is trying to tiptoe into mobile by way of the assets it acquired from Infineon in 2011. The company has been making inroads into Qualcomm’s position in mobile radio modems with keystone-customer Apple. But mobile modems are an uphill slog for a company that lives by selling expensive, high-performance chips. The mobile and IoT markets are severely constrained by both power and cost.
So, what is Intel disrupting if it wolfs down Broadcom, either alone or with one or both of the others in its belly? Unlike Broadcom, Intel is a real operating company whose main business is to create products, deploy assets, generate revenue, and satisfy customers. Is it possible that Intel would actually run all these combined businesses? Perhaps. It could claim Qualcomm-NXP’s position in automotive, gain a much better position in mobile, particularly 5G, and potentially pull a lot of its acquisitions’ parts into its own factories, thus keeping them running closer to full capacity. It could “Americanize” Broadcom and thus help that company with its U.S. regulatory problems. Maybe Intel would shut down its own nascent 5G business and run everything through its acquisitions. But it could go the other way, killing Qualcomm’s products and swapping in its own. Only its barber knows for sure.
But that brings me back to where I started. None of this is really going to work as an operating entity. If the skilled employees aren’t on board, nothing good can happen. The main thing accomplished by taking over a rival in tech is to kill it off and open up a patch of sky.