SunPower IPO Shines Light on Weak Wireless Revenue
Cypress Semiconductor Corp (CY, Buy). Reiterate Buy Rating as SunPower IPO unlocks shareholder value
Cypress' SunPower subsidiary makes solar cell, photovoltaics, and other opto-electronic components. Its IPO is sure to take the headlines. While Cypress's weak revenue results ripple through our model, lowering revenue and EPS estimates, the real story is in the underlying value of SunPower. We're sure to find out what that is soon as Manny Hernandez steps down as Cypress CFO to take the CFO role at SunPower in preparation for a third- or fourth-quarter IPO.
Cypress plans to hold a majority stake in the IPO, and as such should benefit from the huge potential that SunPower holds. While our revenue estimates are reduced as a result of new management guidance, we still feel there is upside to shares.
What the Bulls will point to:
SunPower will be launched in an IPO that could take place as early as 3Q05. We believe the IPO is likely to be conservatively valued at 5 to 6 times next fourth quarter sales of $125M, or $500M to $600M. This is based on the only other pure-play solar cell company Evergreen Solar (ESLR) that trades at 4 times its project 2006 sales. Cypress is expected to retain 80 percent ownership in the company after the IPO (with potentially two classes of SunPower shares). The impact is certain to shore up the company's very leveraged balance sheet.
Strong Bookings: Book-to-bill ratio was very strong at 1.09 with bookings of $217M (down 23 percent from the same quarter last year, but up 16 percent from last quarter). The mid-point of guidance for the 2nd quarter is $217M and this could be met with a stable bookings pattern. Three-month backlog was $152M.
Restructuring Impacts 2Q: Restructuring results are expected to be fully achieved in the 2nd quarter of this year. The restructuring was more than work force reduction; it realigned the company around systems, with three new business groups created - Memory and Image Sensor, Computation and Consumer, and Data Communication Business. We were impressed with the gross margin guidance snap-back from 1Q to 2Q as utilization should increase from 63 to 71 percent (break-even results are forecast at 71 percent utilization.)
Exiting pSRAM: The company is investigating options for its pseudo-SRAM business as it does not fit in with the company's long-term goals. This business is presently $11M per quarter and is sourced from foundry partners. While divestiture will mean lower revenue levels, it is likely to result in more results consistency.
What the Bears will point to:
WAN/SAN Business Remains Weak: Based on the prior categorization, revenue for the WAN/SAN segment would be down nearly 30 percent from the same quarter last year, and down quarter-to-quarter for the fourth consecutive quarter. The calendar year is clearly being impacted by the TCAM3 product cycle in the Network Search-Engine (NSE) market, and revenue is likely to be flattish through 2005.
Wireless SRAM Dynamics: Even including the FillFactory acquisition, the old wireless business would be down nearly 30 percent from last year, and down for the third consecutive quarter. SRAM units and ASPs were down sequentially in 1Q. Growth in this category is essentially going to rely on the image sensor acquisitions.
Moderation in the 2006 Revenue Ramp for SunPower: In response to a question on the conference call, SunPower's CEO pegged the 2006 revenue target at $200M. This is below the prior estimate of $250M issued by Cypress' management, and below what we believed was a cautious target of $225M for us. Clearly, detailed planning of the timing of various manufacturing ramps has played into this more realistic scenario.
June Quarter Outlook (Key Numbers):
Revenue between $210M and $225M.
ASPs forecast down 5% quarter-to-quarter - same as 4Q to 1Q decline.
June quarter is 70 percent booked.
SunPower forecast revenue at $15M
End Market comments:
Overall end markets are expected to be flat.
Revenue growth is coming from market share and new products.
Wireless base station improving
AmTech's own estimates for the June Quarter
Revenue of $216M, slightly below mid-point of guidance.
Operating expense of $90M, $2M below guidance as we believe restructuring impact on guidance was likely conservative.
Gross margin is forecast to increase to 43 percent on higher utilization levels.
SunPower revenue forecast is $16M, $1M above guidance as yields and ramp accelerate modestly.
Optical Segment growth: MID up 6 percent quarter-to-quarter, DCD up 1 percent, CCD up 8 percent
Summarizing 2005 estimates
We have taken down our CY05 revenue to $890.6M from $915M, as a result of slow growth in the data com business unit.
Our SunPower revenue estimate remains at $75M for the calendar year.
The 2006 Numbers
We have taken down our 2006 revenue estimate to $1108M from $1275M, as a result of slower growth in the network market, slower growth in wireless SRAMs, moderation in SunPower's revenue ramp, and conservative growth in the image sensor business.
Our SunPower revenue estimate declines to $206M from $225M as 2006 guidance was lowered to $200M on the conference call. We believe there was some amount of conservatism in the updated revenue guidance.
We are lowering our model for 2005 and 2006 with lower revenue estimates. But we will soon realize the true market value of the SunPower business and have reason to believe our internal valuations may prove conservative.
Basically, we are looking for an incremental $205M in 2006 revenue at a 35 percent gross margin (versus Cypress' corporate gross margin of about 45 percent) and a 21 percent operating margin. Blending the two, we get $1,108M in revenue.
Equally, if we try to look at the SunPower IPO value in the third quarter of this year, we believe a safe estimate of valuation is $500M to $600M — and in 12 months the IPO value could be 4 to 5 times sales of $200M or $800M to $1B. Adding back the Cypress semi business at two times sales of $900M we get $1.8B for a total value of $2.8B.
Join our online Radio Show on Friday 11th July starting at 2:00pm Eastern, when EETimes editor of all things fun and interesting, Max Maxfield, and embedded systems expert, Jack Ganssle, will debate as to just what is, and is not, and embedded system.