Marvell Technology Group (MRVL) reported revenue of $513 million for their quarter that ended January 31, 2009 (4th quarter fiscal 2009). That was not a good number for Marvell; it was down 35% from the October 2008 quarter, and down 39% from the year-earlier quarter.
I own Marvell stock. I had hoped for better, but not expected it. Since a lot of Marvell's revenue comes from the data storage sector, I hoped people's need to store ever-increasing amounts of data might partially offset the recession. As we have seen, for instance, at Dot Hill. On the other hand, I have a backup drive that once seemed capacious and now is getting pretty full, but I am delaying buying a bigger capacity drive right now; I have more pressing needs for cash. Multiply by 100 million.
Marvell is a relatively new company that grew big quickly and seems to aim at dominating the entire mixed digital analog sector. In addition to making chips for Western Digital drives, it makes processors that RIM uses in some of its Blackberry products, chips for printers, chips that combine Bluetooth and WiFi, and video processing chips. It seems to be able to move its engineering expertise from one technology arena to another, taking on the likes of TI, Motorola, Intel and Broadcom. Sometimes it takes Marvell a while to get traction in a new area, but when Sehat Sutardja says "In a year or two, " it is a good idea to take him seriously.
In a year or two, Sehat believes there will be architectural convergence for many devices, including netbook computers, smartphones, and next generation data and communication appliances. He believes Marvell has all the expertise necessary to get there with the best products at prices that enable the magic of mass consumption with good profit margins for Marvell. That looks right to me. Marvell has shown it can do ARM processors as good as anyone, and it is unrivaled at putting analog circuitry on chips.
But I would not discount, quite yet, the ability of rivals to out market, and perhaps even out engineer, Marvell. Motorolla, TI and Intel all have deep pockets and large research establishments.
Of its $513 million in revenue in the quarter, Marvell spent $208 million for research and development. Obviously they could have sent a lot of engineers home in 2008 and pleased short term investors. That is not the Marvell way.
Whether Marvell was profitable in the quarter depends on how you want to do your accounting. On a GAAP basis they lost $65 million. They claim non-GAAP net income of $32 million. On a cash basis, they had free cash flow of $95 million. They ended with $952 million in cash and equivalents, so they are well positioned to keep up their drive for industry dominance.
They have cut costs, and are reducing the work force by 15%. If I know Marvell's corporate culture, that means those lucky enough to be employed are going to work even harder and smarter, if that is possible.
For more details, see my Marvell analyst conference summary for March 5, 2009.
The current economic and stock market troubles provide for a unique opportunity to buy into Marvell, but there is risk in investing in the highly competitive semiconductor industry, so ...