RF Micro Devices, Inc. (RFMD)

[eastunder]

RFMD, SWKS Drop on QCOM News; Raymond James Downgrades, Longbow Defends.

By Tiernan Ray

http://blogs.barrons.com/techtraderdaily/2013/02/21/rfmd-swks-drop-on-qcom-news-raymond-james-downgrades-longbow-defends/?mod=yahoobarrons

Shares of wireless chip vendors are crumbling today, after Qualcomm (QCOM) this morning said it will start selling a so-called “front-end module,” the parts that handle the RF portion of a wireless device, in addition to its existing franchise in the baseband portion of wireless circuitry.

Qualcomm shares are down 66 cents, or 1%, at $64.61, but shares of RF Micro Devices (RFMD) are down 74 cents, over 14%, at $4.44, TriQuint Semiconductor (TQNT) stock is off 35 cents, or 7.4%, at $4.40, Avago Technologies (AVGO) is down $1.10, or 3%, at $33.66, and Skyworks Solutions (SWKS) shares are down $3.06, or 13%, at $20.56.

Raymond James analyst Tavis McCourt cut his ratings on all three RF vendors today, reducing Skyworks to Outperform from Strong Buy, cutting RF Micro to Market Perform from Outperform, and cut TriQuint to Underperform, calling Qualcomm’s announcement a “bombshell.”

McCourt notes that the $4 billion to $5 billion front-end RF market is growing by 10% to 15% annually, propelled by the advent of “long-term evolution,” or LTE, cellular, which is generally increasing RF content in chipsets.

McCourt notes that not all of the technical aspects of Qualcomm’s offering are proven, or even known, for that matter:

Qualcomm expects to have commercial products with its RF front end in market in 2H13. Qualcomm has had internal projects around RF front end components for a number of years, but this is its first commercial deployment that we are aware of. Of note, Qualcomm is using a silicon CMOS-based power amplifier, and it has not published any power efficiency stats, so its actual performance versus existing GaAs-based PA solutions is uncertain. Additionally, Qualcomm is not integrating its RF front end with its snapdragon processor, so this is a discreet system that Qualcomm will attempt to cross sell (i.e., vendors can choose the Qualcomm RF front end or competitors’ offerings freely.) We expect the Qualcomm solution will target the Chinese market initially, but over the longer term, Qualcomm’s ambitions are obviously much larger.

McCourt expects P/E multiples will be “capped” for the RF vendors given that he thinks it is a matter of time before Qualcomm finds success in the front-end market:

Historically, Qualcomm’s entry into new chip markets have had only modest success in the first 1-2 years, but almost universal success over a three- to five-year time frame given its substantial R&D scale and systems expertise. We expect Qualcomm’s entry will have no impact on Skyworks’ analog, WLAN, or other non-cellular end market businesses.

But there were also defenders today. Longbow Research’s Joanne Feeney reiterates Buy ratings on Avago, RF Micro, and Skyworks, and a Nuetral rating on TriQuint.

Feeney thinks the Qualcomm part is too generic to pose a real threat:

QCOM’s new front-end module (FEM) provides an off-the-shelf solution for OEMs, but the generic nature of this device—and a number of technology shortcomings—will be very likely to limit adoption. QCOM’s solution is based on all-silicon (SOI or SiGe) technology whereas leading-edge PAs are GaAs-based because that material provides superior performance, so we suspect this solution would be adopted only in lower-end smartphones. The FEM can only perform as well as its worst component – so handset OEMs which emphasize performance will likely steer clear of this one. Superior performance claims likely reflect QCOM’s baseband and the integration of envelope tracking (ET) technology—but others will have ET-PAs at the same time. Sub-par filter technology: QCOM is using SAW filter technology and this underperforms FBAR/BAW on certain bands – particularly LTE bands. Once again, this points to QCOM’s RF360 being targeted at lower-end devices. Uniformity for OEMs: QCOM’s solution removes the ability of OEMs to differentiate handsets. Handsets will end up having the same performance specifications. This would undermine the ability of the OEMs to segment its own customer base into high-spending customers and low, and would remove the ability to differentiate from competitors. No customization and no second source.

And Lazard Capital Markets’s Ian Ing stakes out a middle ground. For him, Qualcomm’s product may not be about competing at all, but rather about protecting its baseband franchise:

One could argue that QCOM intends to gain handset content by entering the RF market (despite unattractive gross margins), which would be a negative for RF vendors. More likely, we would argue that Qualcomm has viewed “band fragmentation” as a threat to building single- platform world phones (QCOM’s default combo chip) as it opens the door for other modem suppliers with more limited solutions (such as BRCM, Mediatek, etc). Given that RF vendors are without an easy turnkey solution to support the 40 radio-bands in the world, QCOM may be attempting to push the RF IC vendors to better match the bar set by the QCOM RF360 solution. QCOM has ‘kick started’ markets before: We note QCOM will often enter, then exit, markets just to “kick start” them to better help its QCT and QTL model. Recall, it produced CDMA handsets in the mid-to-late 90s, and recently sold its stake at an Indian BWA carrier (JV with Bharti). These moves were not meant to become a Motorola or a Verizon, but to create better opportunities for QCT and QTL in the long run.

Ing maintains Buy ratings on Avago, Skyworks, and Qualcomm, and Neutral ratings on RF Micro and TriQuint.