Financial and market analysts that I talked to agree that the $121 billion deal is in the best interests of stockholders for the short-term. But long-term, they say that it’s bad for the industry and that Broadcom’s growth-by-acquisition model will not be sustainable.

“I think the deal will be good for Broadcom’s stock price, but as a former engineer, I will weep about it,” said one stock analyst who asked not to be named.

“The Broadcom model is good for stockholders and bad for the industry,” said an industry watcher who also asked not to be named.

They refer to the former Avago’s history of carving up its acquisitions, keeping the best bits, managing their costs tightly, and selling off the rest. It happened with LSI, Broadcom, and Brocade. Qualcomm is the next and largest target of what some former Broadcom executives call Hock Tan’s “financial engineering.”

Tan would sell off Qualcomm’s RF360 joint venture with TDK, its Centriq ARM server group and perhaps a big chunk of its patent portfolio — perhaps even throwing in some patents from the former LSI and classic Broadcom. There’s some debate about whether he would sell off the former CSR Bluetooth and former Atheros Wi-Fi groups.

The crown jewel, of course, is Qualcomm’s cellular baseband franchise and, to a lesser extent, its mobile application processor business, which Tan would tightly manage. Hock Tan once told one analyst, “I sit on the mountaintop, watching farmers plant their fields, and when it’s time for the harvest, I come down and pick it.”

This is a harsh model, but business is harsh. The problem with Tan’s model is that he requires others to plant and water seeds for fruits that he eventually harvests.

Two or three years after what would be by far the biggest merger in the semiconductor industry, his investors would be hungry for growth again. It’s a kind of addiction. Eventually, the king of the mountain looks like the fool on the hill.

Qualcomm is the victim of its own hubris. For decades, it has reportedly charged some of the steepest patent royalties in the cellular industry to recoup its R&D investment. Since the CDMA era, its demands have drawn lawsuits and regulatory actions, culminating in the falling out with Apple and another top customer, said by some to be Huawei.

Qualcomm deserves a premium for leadership engineering in cellular. But how much? I’d love to see a detailed analysis of its research spending relative to its royalty charges, but numbers on the latter remain confidential in today’s opaque patent market.

Clearly, the company makes big investments, but Apple claims that Qualcomm charges five times the royalties of all other cellular patent holders combined. I note that Samsung — another frenemy of Apple — plunked down a staggering $26 billion on capital equipment for its fabs — a whopping 40% of its semiconductor sales in 2017.

The value of a patent is whatever people are willing to pay for it, and Qualcomm has consistently pushed the envelope to the breaking point. It could at any point — even now — recalibrate its IP business model, but I doubt that it will.

My gut tells me that Paul Jacobs would rather let Hock Tan pull out his fingernails than let him buy the company that his father co-founded — even if he got the 13th seat on a merged company’s board. Similarly, I can’t imagine him reforming the business model behind the storied patent wall in its headquarters.

The patent wall in Qualcomm’s San Diego headquarters.
The patent wall in Qualcomm’s San Diego headquarters.

So I suspect that shareholders will decide the company’s fate — more specifically, the few dozen executives who control proxy votes for big institutional investors such as Blackrock, Vanguard, and Fidelity. Experts tell me that they take input from their in-house stock analysts, but in the end, these proxy execs make their own calls.

I called Qualcomm’s top 10 institutional investors. Not a single one would speak with me. So I suspect that there may be no resolution for this matter until the proxy vote on competing slates of directors is announced on or about March 6.

Even then, analysts say that Qualcomm’s deal to buy NXP could throw a monkey wrench into the works. Qualcomm could increase its bid for NXP, and China could approve the deal any day, scaring Broadcom off.

Hock Tan wants a few big leading product franchises to manage and a few juicy pieces to sell off. NXP is not seen as a prize in his eyes, say analysts.

There are many shoes yet to drop in this semiconductor soap opera. None of them look very pretty to me.

— Rick Merritt, Silicon Valley Bureau Chief, EE Times