Solar Power Hype

May 13, 2006

Solar Power Hype

We read EETimes on a regular basis. We publish links to their articles which we think point out industry trends which are being missed by the mainstream press. We also periodically rant about aspects of their reporting which read more like cheerleading and/or low-wattage parroting of "research" whose results leave more questions than they claim to answer.

We also enjoy pointing out when they put out a piece which makes us think. When they motivate us to investigate something where even we were accepting the hype. This week it's solar power.

Cypress Semiconductor (CY) gained a ton of press over the last couple quarters affirming the genius of T.J. Rogers for his 'vision" when he invested in SunPower and their solar power technology. Rogers describes himself as a "wealth creator" although, as a former long-term CY shareholder (10+ years from shortly after the IPO), I would like to see a list of who, besides himself and fellow insiders, has gotten rich investing in his companies.

As of May, 1996 CY was $13.875. At Friday's $16.27, that's a 17% gain over ten years which less than a 1.5% CAGR. T-bills exceeded a 10-year investment in Cypress Semi many times over.

In the case of SunPower, it does appear another batch of people have benefited from being early financiers of the company. The stock zoomed from $27.10 to $36.40 (+34%) since its November 17, 2005 debut. However, sales are only $78.7M and they lost -$15.8M over the last 12 months. With 61.1M shares outstanding (that we know of, since we haven't read their disclosures and cannot say if this accounts for insider options), SunPower's market cap is $2,224M which is 28(!) times sales!!! MSNMoney projects $0.24 CY06, with $0.04 in 2Q06. Accepting what sound like extremely optimistic assumptions for the moment, that's 226x P/E on 2Q06 projected earnings or a 152x P/E for CY06!

Part of the rampant optimism on the stock stems from quotes like "we're sold out for the next six months" and "we've already sold everything we can make." Usually within a sentence or two of those comments, you also hear about a "shortage" of polysilicon material (i.e., wafers). So which is it? Are they sold out? Or they can't get enough materials? We also read a comment spot prices for these materials have gone from $30 to $40 to even $100. Which price is SunPower paying? How does this affect their selling price? Isn't there a lot of elasticity in the demand for solar power? (i.e., doesn't demand increase as price decreases? And vice-versa?)

Somehow these comments always lead to the most optimistic assumption possible for SunPower's sales.

And you thought the technology-stock bubble was dead?

Before you accept the growth story and the inevitability of a positive return on a SunPower/ Cypress investment, the figures above, compounded by the EETimes article below, bear consideration. If you don't see serious holes in their "story," feel free to drop us a line and explain. If Cypress or SunPower has responded to these allegations, we haven't seen it.

* * * * * *

Data presented in solar-power article doesn't add up

EE Times
(04/24/2006 9:00 AM EDT)

I am an electronics engineer and have been interested in solar power for quite some time. Thus I read with great interest Loring Wirbel's article titled "Cypress' solar-power gambit bears fruit in SunPower IPO" (Dec. 12, page 1). Cypress' attempt to make solar cells is worth applauding, but the data presented in your article is confusing and full of inconsistencies. It states that a solar panel that uses seventy-two 5 x 5-inch solar cells is rated to deliver 150 watts. It also states that Cypress' solar cells have a conversion efficiency of 21 percent, while the industry is at 14 percent. A simple calculation shows the total area of a 72-cell panel to be around 1.16 square meters. The number I am familiar with in the industry is that on a clear day, we get around 1,000 W/square meter from the sun at noon on the equator. Most of the industry uses this as a base number.

Thus, for a 1.16-square-meter solar-cell surface area, the incident power from the sun is 1.16 kW. If the output from the solar panel is 150 W, the conversion efficiency from these numbers comes out to be only 12.9 percent [(150/1,160) x 100]. This is much lower than the quoted 21 percent. To achieve the 21 percent conversion efficiency in this case, either the incident energy has to be reduced to 715 watts or the output of the solar panel has to be increased to 243 watts. The article does not clarify any of these statements and makes exaggerated claims.

I found other problems in this article with the cost numbers. I understand that these numbers are not real, but the problem is that they are not even consistent. At one point the article quotes Cypress CEO T.J. Rodgers as saying that solar power today costs $2.99/W. I assume that this is the cost to the user. But is it the installed cost or only the material cost? If it is the material cost, then these numbers are also inconsistent.

The article further quotes Rodgers as stating that in rural distribution systems, the cost can be as low as $1.50 per watt. There is no explanation as to how the material cost can come down in the rural distribution system. If I assume that it is the installed cost and the labor cost is low in rural distribution systems, then this cost is ridiculously lower than what is further quoted in the same article. When I do a simple calculation from the cost of a 150-W panel of $700 (as mentioned in the article), I get a cost/W of $4.66.

I also have some issues with the cost of one 5 x 5-inch solar cell (priced at $6 in the article) vs. the cost of a solar panel that consists of 72 cells. If I multiply the cost of 72 cells on one panel, it comes to $432. The actual panel cost of $700 is much higher. What costs an additional $268? Wiring, a frame and some protective glass?

It is articles like this, with overblown marketing numbers and exaggerated projections, that create confusion in the market and give the industry a bad image. It is the responsibility of the writers to do some simple verification on their own of the claims made by the industry before publishing such misleading and erroneous numbers.

Jagdish Pathak, President
Sub Micron Circuits Inc., San Jose, Calif.

* * * * *

The above is not intended as advice to buy, sell or hold any stock, bond nor any other financial product. Invest at your own risk.