On Dec. 16, Oppenheimer issued a downgrade on shares of Broadcom (NASDAQ: BRCM). The reasons cited? The argument appeared to be principally focused on rising structural risks to the company's core connectivity (that's Wi-Fi, Bluetooth, GPS, etc.) business. In particular, Oppenheimer is worried about share loss in connectivity to cellular chip leader, Qualcomm (QCOM +0.24%), as well as threats from potential in-house solutions from Samsung (NASDAQOTH: SSNLF) and Apple (AAPL 0.01%). Are these concerns off-base, or should investors be concerned?
Vertical integration is the new "buzzword"
While the new "fad" is to assume that every device/systems company is going to bring all silicon content development in-house, the truth of the matter is that chips are becoming more expensive to develop as semiconductor manufacturing technology becomes more complex. While some companies (admittedly like Samsung and Apple) do have the scale to perhaps justify a meaningful investment in doing things in-house, it very often doesn't make sense to since the merchant chip vendors can amortize the development costs over many more units.
Further, it's not correct to assume that as long as Apple or Samsung decide to throw money at the problem that the optimal solution is bound to emerge. For instance, Samsung develops its own modems and applications processors and yet it frequently sources both (especially for very important devices such as the Galaxy S and Galaxy Note) apps processors and modems (or integrated single chip solutions) from others. If doing in-house implementations of very complex chips were so easy, then the entire idea of pure-play chip companies would be broken and just about every chip vendor would be a short-to-zero
Share loss to Qualcomm? Perhaps
The first and, admittedly, very legitimate concern is share loss to Qualcomm. Broadcom has been adamant in claiming that it sees its market share position, particularly in high end LTE phones (where connectivity content is quite high and performance matters), is defensible. This seems plausible if one works under the assumption that Broadcom's connectivity products have a real performance and/or power consumption advantage over competing products. At the firm's analyst day, the GM of the company's mobile and wireless division showed a slide that that the company does indeed have a performance advantage over a major competitor.
Now, this is not to say that the performance numbers are Qualcomm against Broadcom (there are many players in connectivity but one would suspect that it's Qualcomm), but the broader point here is that just because Broadcom faces competition from Qualcomm doesn't necessarily imply that the share loss is a done deal; if Broadcom engineers the better/cheaper/faster solution, it can still defend its share.
Share loss to vertical integration? Probably not
Apple has probably designed what is arguably the best mobile system-on-chip today with its A7. Samsung's chip efforts have been less impressive, but its acquisition of CSR's mobile business seems to signal that it's interested in this space. Further, Apple has – by way of numerous job openings for RF IC engineers as well as engineers experienced in Wi-Fi and Bluetooth – signaled that it, too, could be looking to do a wireless chip of its own.
While Samsung and Apple could succeed in developing a Broadcom-class connectivity solution, it'll be tough – really tough. Broadcom's R&D in mobile and wireless is nearly a billion dollars per year, and the company has also amassed plenty of design expertise and IP from all of those years of working on these products. Could Apple or Samsung conceivably outgun Broadcom at its own game? It just doesn't pass the smell test – at least for any high end handsets where performance is king.
Foolish bottom line
Are there risks that Broadcom loses connectivity share? Of course – this is a competitive business and the folks at Qualcomm and Marvell aren't dummies, and from conversations with folks within Intel, it looks like they're getting more serious about low-power connectivity (in order to pull off SoFIA, they'll need to).
On the flip side, Broadcom can gain share in LTE modems and apps processors (where it has no presence today) – the other side of that "share loss" argument. While Qualcomm's cellular position is likely also very defensible, it's important to note that cellular basebands are much more expensive than connectivity combos, meaning that each win in cellular helps Broadcom much more than each win in connectivity that Qualcomm wins. Also, it'd take something spectacular to convince this Fool that Apple or Samsung will – at any point in the near future – be able to encroach on Broadcom's connectivity business with in-house solutions.
