The news that Broadcom is getting out of the cellular baseband chip business brought a widespread sigh of relief to its shareholders, and its stock price immediately turned upwards, rising some 16% since the market first sniffed the news this Monday.
That all seems to be based on it doing more of what it does profitably – WiFi, broadband, set-top SoCs, cellular where it’s not than baseband and networking.
It is odd when both logic and unreliable and rather emotional investors align, but this is one of these rare cases. Broadcom put it best when it said that it would save $700 million a year in R&D and other costs, and this means that it can shore up its broadband and WiFi businesses, where it clearly leads. It is doing so well in its other businesses, that Broadcom was able to say it will drop out of baseband chips and still keep its revenue forecasts and says it will actually beat its margin forecast.
The news came as a surprise to the market despite Broadcom regularly braking out the cost of trying to mix it with the increasingly giant-like Qualcomm, and the emerging Chinese giant MediaTek, who between them hold 76% of the market, other contenders include Intel, Ericsson, Marvell and Spreadtrum.
Broadcom has now entered yet another game of pass the parcel. The larger part of its modem business came from Nokia, being sold to Japan’s Renesas Mobile, which it bought last September for $164 million. Renesas had previously paid $200 million for the Nokia wireless modem business. It is likely that even if a buyer can be found for its business, Broadcom will get back far less than it paid. Giving it away would mean no write-downs (or smaller ones) or closure costs.
Nvidia is another company that bought itself into the business in 2011 snapping up Icera for $367 million, but now laboring, like Broadcom, with having come to the market too late and landing too few design wins. Recently Broadcom made its first big LTE modem win at Samsung, for its M320 dual core LTE SoC. As the handset business converges in to a smaller and smaller number of larger handset vendors, design wins are far harder to come by and the smartphone leads of Apple and Samsung now make it essential to get into at least one of their phones.
Broadcom said it may sell or close the business and has hired JP Morgan to help it consider its options.
In its latest quarterly report Broadcom reported $846 million in mobile and wireless, but this will include many chips onboard smart phones, including Bluetooth and WiFi, so it’s hard to see how much of this related to baseband, potentially very little. However costs last quarter outweighed revenue, leaving it with a negative operating income (before SG&A) of $32 million for the first time last quarter, so it was obvious that something had to be done. At the time Broadcom said that it was taking time for LTE sales to come through, implying that it was prepared to wait.
On the surface it now looks like Qualcomm can takes shots at Broadcom with impunity, hiding behind its captive SoC customers, safe in the knowledge that most companies cannot build a phone without Qualcomm. But there may be quite a lot of wisdom in the move by Broadcom, and it may not all be seen as an investment failure. WiFi connectivity is rising in importance in smartphones, which is what Broadcom is great at, especially as cellular architectures take in HetNets, and fixed line players offer it in Homespots for offload. Also the LTE chips could be seen as commoditizing, lowering in unit price and becoming slightly less strategic, than they have been, now there are multi-mode connections and WiFi First phone services.