Trying to peer into the future, a variety of sell-side analysts attempted to pry specific information out of Marvell Technologies Group (MRVL) CEO Sehat Sutardja on his vague predictions of increased sales of application processor chips into the smartphone and tablet market in the second half of 2013. They did not succeed (see "See Marvell Q1 2014 Earnings Call Transcript"). During the presentation portion of the fiscal Q1 2014 call on May 23, 3013, Q2 revenue guidance was given as $770 to $810 million. But beyond that, there were just a few details other than the assertion that growth would be driven "by increased traction in areas such as mobile handsets [smartphones], tablets, connectivity and SSDs." The guidance would be up sequentially between 5% and 10%, but down 1% to 6% from the $816.1 million Q2 last year.
The storage controller chip business, which accounted for 53% of Q1 revenue, is somewhat seasonal so it would normally improve in the second half from back-to-school and holiday sales. Marvell is projecting market share gains as well, after a 5% share gain in fiscal 2013 (which ended Feb. 2, 1013). With a typical attach rate of 1 hard drive, hybrid drive, or SSD per PC, final numbers will probably come down to how many PC units sell, plus whatever demand there is for drives for data center and cloud storage. Other analysts have pointed to SSD unit growth as a big positive, but at this point it is such a small part of Marvell revenue that even rapid growth is no cause to get excited.
Mobile and wireless results were below expectations in the second quarter. The segment represented 18% of revenue or about $132 million, a 24% sequential decline. This segment is where much of the Q2 and second-half growth is supposed to take place. It includes ARM-based application processors and wireless technologies such as cellular, Wi-Fi and Bluetooth, which may be integrated on a single chip. The story here is complex. At one time most Marvell sales in the smartphone market went to RIM, but Marvell never served the CDMA standard. Marvell does not seem to have any slot in the new BlackBerry 10 models, at least not in the U.S., but then again those have not seen high sell through yet. For two years Marvell smartphone chip sales have been concentrated in China, where it saw and seized an major opportunity with the TD-CDMA standard.
For several quarters now Marvell has looked forward to selling a "global" SoC that can run most modern cell phone protocols. Sehat believes they are ready with a chip that is powerful enough to run tablet computers, yet conserves power enough to run smartphones, and can be deployed globally. But Sehat would not name customers. This is nothing new for Marvell, and it results from the nature of the markets Marvell served. It is not like the PC industry, where people care or at least know whether the CPU is made by Intel (INTC) or AMD (AMD). Marvell seldom takes center stage. The company is bound by agreements with OEMs, and in most cases cannot announce a product win until the product is announced by the OEM. Sometimes the rest of us don't learn about Marvell wins until a new device is torn down and analyzed by a technology review site.
We continue to make strong progress in mobile at multiple tier-1 OEMs with our integrated 3G as well as 4G platforms. Recently a leading Asian tier-1 OEM started production of a new smartphone and a tablet based on our dual-core platform … we are confident that additional devices will be introduced by this OEM as well as other tier-1 OEMs in the coming quarters.
That sounds great, but how do you put a dollar amount on it? Even if we knew the major OEM involved, and its current market share, we don't know if all its new smartphones will be Marvell-based or if it will offer a variety, including those based on SoCs from competitors. "We are also seeing faster adoption for our recently introduced quad-core 3G platform solution." Initial shipments have already started. Sehat also pointed to the announcement of the PXA1088 quad-core 5-mode LTE SoC earlier in the week. "As a result, we expect at least one leading handset OEM to bring to market a 4G LTE smartphone using our solution this year, with more following early next year."
In answer to one analyst question, Sehat did indicate that the growth driver in Q2 and the second half would be the 3G smartphone solution as well as tablet computers. However, the tablets would likely use only Wi-Fi, adding that "by the end of the year we will have a much better position in this market." But after that, further questions on this subject were answered with no additional concrete information. Sehat reiterated a previously announced goal of gaining 10% of the cell phone handset market, at the same time saying it would be a difficult target to achieve. Also, it seemed like the LTE goal was to have at least one tier-1 OEM customer in North America, in addition to any in Asia, but it did not appear that was certain yet.
All I have been able to conclude from this dialog is that revenue is likely to improve in the second half and into 2014, but how much improvement there will be depends on sell-through of the handsets. I would hazard that style and brand-appeal sells smartphones in developed nations, while developing nations are price-sensitive. Predicting sales for tablets other than iPads has been particularly problematic so far.
I have acquired MRVL stock over time as part of a long-term investment strategy, so for me the decision is between selling and holding. I believe solid market share in networking and storage minimizes downside risks. The normal second half seasonal ramp is expected. If there is good acceptance of the new smartphone chips, the main benefit will be in calendar 2014, but we should see some revenue in calendar 2013. I am willing to wait for that. Sehat did say the smartphone revenue "will be more meaningful next year," in answer to a question.
Finally, if you are thinking of investing in Marvell you should understand the patent litigation situation with Carnegie Mellon University (CMU). A jury in Pittsburgh, where CMU is located, found in favor of CMU. Unless CMU and Marvell settle, the appeal process could take years. I would certainly trust an appeal judge's opinion on the matter more than a hometown jury of people with no experience in either patent law or storage controller algorithms. I believe the danger of a large award against Marvell may hold back the stock price for some time. I'd like to see a quick settlement, even if it is higher than is reasonable given the likelihood of the results of an appeal. A settlement would do more than anything, short of a full victory on appeal, to improve the stock price. For Marvell's take, see "Marvell Technology Group Seeks To Overturn Jury's Patent Infringement Findings."
Marvell closed on May 29, 2013, at $11.43, giving it a market capitalization of $5.7 billion and a trailing non-GAAP P/E ratio of 13.9. The 52-week high was $13.27 back on May 29, 2012, and the 52-week low was $6.98 on Dec. 27, 2012, which followed the CMU jury trial outcome. Marvell ended fiscal Q1 2014 with a cash and equivalents balance of $1.73 billion. The dividend is currently $0.06 per share per quarter.