Broadcom - Time for Stellar Execution

Wednesday, February 27, 2008

My next article as published on Sramana Mitra's site..

In the last part of this series, I talked about Broadcom’s aggressive campaign to become a leading vendor in the cellular IC market. In this piece, I will talk about the plan, the challenges and the need to execute these plans in style.

So what is the plan? CEO Scott McGregor said in his Q4 2007 conference call, “Our goal in cellular baseband is to reach a 10% unit market share by the time we exit 2009. The plan remains the same -- 2007 was a year to win designs; 2008 is the year to get products and phones done; and 2009 is the year to grow share and generate significant revenue.”

I like the plan. I also think that the execution has gone quite well so far. Broadcom announced a design win with Nokia last year for its single-chip EDGE chip. Nokia’s dominating market share in this space, coupled with the fact that there will be over 400 million EDGE phones for 2009 augers well for Broadcom. These factors imply that even the design wins for ‘selected future EDGE phones’ translates to large unit shipments instantly getting it closer to the target envisioned. Add to this the 3G business with Samsung, the 10% target looks very reasonable.

The catch, however, is this. 2008-2009 are the peak years for EDGE. Later, these phones will give way to faster phones using WCDMA/HSDPA. I also estimate that the per-unit cost for the EDGE chips will be much lower than the $23 price for the 3G chips. Broadcom has also not won any substantial designs for its 3G solution. While we may see some design win announcements later this year, my worry is two-folds. For data-centric phones, cost advantages apart, performance becomes a key criterion for design wins. Broadcom purchased Zyray in 2004 but is yet to deliver in the 3G space. This makes me think that Broadcom is battling basic performance issues. Secondly, the engineering challenges associated with the complicated design and the level of integration in the new product may push the delivery and in some cases, also lose customers.

Broadcom has a good track record of delivering on its promises. So, despite the odds, I am willing to give it the benefit of doubt. If the execution is good, then we will see 3G design wins this year resulting in significant revenue contributions in 2010. I anticipate that in the long-term, pending delivery, Broadcom is positioned to capture around 20% of the 3G market. This will help it to about 15% market share in 2012. The drivers will be its continued relationship with Nokia, its complete solution and its aggressive pricing. Most of these gains will come from TI and the other smaller vendors as I expect Qualcomm to retain its engineering leadership, and hence also grow its market share, in this space.

In summary, the cell-phone market appears promising for Broadcom. In the next piece, I will complete the picture with a synthesis of the various revenue streams in its Mobile and Wireless business.

Disclaimer: These are my perspectives on Broadcom and does not necessarily reflect the views of Atheros Communications or Tensorcomm.

Posted by Vijay Nagarajan at 9:00 AM  


Vijay - this comment is not directly related to this post, but a more general question.

I have read a few of your very interesting posts on QCOM, and am quite confused by the ASP numbers.

For instance, QCOM reports average ASPs of 215, 214, 214, 203 for 2005-2008.

In one post, you say that average IPR is $11, 5% of handset price. This matches those numbers.

However, in another post you discuss that handset asp is ~$200 for WCDMA, and $125 for CDMA/EV-DO. Given volume mix of those, should not average ASP be far lower than the $200 range?

housemaidsknee said...
February 27, 2008 at 11:35 AM  

Thanks for your comment. The average IPR of $11 corresponds approximately to the currently reported ASPs as you rightly pointed out. On the other hand, the handset market ASPs will continue to ramp down. The $200 and the $125 ASPs are my projections for early next decade. With the product mix, you are right in that the net will come down. But this will still be closer to $200 than to $125 since WCDMA/HSDPA will be in a far higher number of handsets than CDMA/EV-DO.

So, the data that I have provided is consistent. I hope to have addressed the cause of confusion here. Please let me know if you have any further questions.


Vijay Nagarajan said...
February 27, 2008 at 7:35 PM  

I think it might be time to revisit your Broadcom analysis. Scott claims he is no longer projecting market share for 2009 - still targetting 10%, but to reach that will take time. I take that as they have had major setbacks and lost market share at Nokia and Samsung to Infineon and TI. Suspect Broadcom is today overvalued at $25-$30 unless the other businesses like cable modem, ethernet, wlan continue to keep them going.

Anonymous said...
May 2, 2008 at 1:36 AM  

Thanks for your comments.

Broadcom will be largely resilient to one year delays in achieving market share. Even in the event that the company's market share is delayed by over three years, my valuation will drop only around $3-$4. The broad portfolio spanning the products you have mentioned will keep the bottomline intact. I agree that there have been trends recently (e.g. NXP-STM merger) that are worrisome for Broadcom. But, unless I see more tangible signs such as shift in the competitive landscape or its customers losing market share, I will still maintain that Broadcom is undervalued.


Vijay Nagarajan said...
May 3, 2008 at 9:46 PM  

Post a Comment