SunPower Corporation (SPWR) 2006 Q1 法說會逐字稿

Good morning and welcome to SunPower Corporation's first-quarter 2006 earnings release conference call. Your lines have been placed on a listen-only mode until the question-and-answer segment of today's call. This call is being recorded. If you have any objections, you may disconnect at this time.

I would now like to turn the call over to Mr. Tom Werner, CEO of SunPower Corporation. Thank you. Sir, you may begin.

Welcome to SunPower's Q1 2006 earnings call. We are pleased to report another quarter of operational results that exceed our guidance, as well as a number of achievements that we feel extend our leadership in the technology arena. During this call, we will report Q1 financial results, review highlights of our business, and provide guidance for Q2 and update our fiscal 2006 guidance.

Let me start with the operations. We continue to expand our manufacturing capacity at our Philippines facility ahead of our plan. We began production of our third 25 MW solar cell line during the quarter. That third line is already exceeding yield expectations, and that's even while we are running 190 Micron thin wafers. So, we exceeded yield expectations and exceeded improvements on silicon utilization on this line. We attribute this to the significant investment that we have made over the last several years, advanced semiconductor manufacturing systems.

Let me say a few words about technology. You are aware that SunPower makes the world's highest efficiency solar cells commercially available. We're the only company in the world that is shipping solar cells with a minimum of 20% conversion efficiency. This compares to a market average of 14 to 15% according to industry analysts. Yet, we continue to significantly invest in research and development to extend our technology advantage. By the end of this year, we plan to be shipping our second generation technology solar cells, which will have a minimum conversion efficiency of 22%. In fact, we have already successfully prototyped the generation to 22% efficiency solar cells utilizing our current manufacturing lines in the Philippines.

Now, let me talk about silicon. You are all aware that silicon is in short supply. SunPower takes a portfolio approach to its silicon supply, much the same way that you would manage your personal financial portfolio. So we have short, intermediate and long-term contracts. We also work with established and emerging suppliers, and we work with established and emerging technologies.

For 2006, we have 100% of our silicon needs either on purchase order or contract, based on mean Street estimates. During the quarter, we secured a five-year, $55 million silicon supply agreement. This agreement begins in 2006 and increases our coverage under contract for 2007 silicon supply to over 80% of mean Street estimated requirements. We continue to work on numerous other deals, consistent with our portfolio strategy.

Additionally, our silicon position is enhanced by our high-efficiency technology, giving us superior silicon utilization relative to conventional solar technology. Let me quantify this. We are improving our silicon consumption from 2005 levels of 9 grams of polysilicon per watt, less than 7 grams of polysilicon per watt on our Line 4 that again we will implement the end of the year. This is over a 20% gain in silicon utilization efficiency by the end of this year. Based on industry analysts, we believe that conventional technologies use between 11 and 13 grams of polysilicon per watt. We continue to work to improve this metric by increasing our cell efficiency, lowering our wafer thickness and improving our manufacturing processes. As we look forward, more silicon will become available and prices will abate. We expect to have a significant cost leverage because of superior silicon utilization.

Now, let's look at our expansion plans. At the end of the quarter, our nameplate capacity -- or the capacity of our first facility -- reached 75 MW. Our fourth line is on plan to begin production by the end of this year, and our capacity, our nameplate capacity at our first plant will be over 100 MW at that time.

Last quarter, we announced that the SunPower Board had approved engineering of our second facility. The Board has subsequently approved construction of that facility. Construction will start during this quarter. You recall that that building is designed to house six Generation 2 compression lines. That means that this facility will have an aggregate incremental capacity of 200 MW, and we expect to add the first two lines in 2007.

Importantly, we accelerated our investment and construction of our new, previously unannounced, advanced panel manufacturing plant. This panel manufacturing plant will have an initial capacity of 30 MW. It's in the Philippines, by the way, near our cell plant. And it will have an eventual capacity of 90 MW. We expect production to begin in the third quarter of this year, and the benefits of this facility will be lower manufacturing costs through advanced automated production. We call this automated production our auto line, borrowed from Cypress Semiconductor's back-end chip-packaging auto line. This also gives us another advantage, because production will be closer to our research and development team, which will allow us to have faster cycles of learning by running our ideas in our manufacturing facility.

Now, let me move onto product advancements. We began shipping to customers, in volume, our upgraded solar panel product line. What that means specifically is customers are now receiving a 220 watt standard 72 cell rated panel. Doing the standard configuration, this is the highest-rated power panel in the industry. This, of course, still continues to leverage our aesthetic advantage. Lastly, we were the first company to have a panel completed through the new, more rigorous IEC product qualification standards, which is further evidence of our technology leadership.

Now, let me finish with guidance. Recall that we provide topline and EPS guidance. We expect Q2 revenues to be in the range of 50 to $52 million. We expect Q2 non-GAAP net income to be in the range of $0.05 to $0.07 per share. Finally, we expect 2006 full-year revenues to be in excess of $220 million.

Now, I'd like to turn the call over to Manny Hernandez, who will provide details of our Q1 financial results.

Thanks, Tom. Good morning and thank you for joining our conference for the first quarter of 2006.

Before I go over the financial results, I'd like to remind everyone that, during this conference, management made and will continue to make statements that are not historical in nature. Please consider these statements as forward-looking, pursuant to the Private Securities Litigation Reform Act of 1995. Those statements are based on our current expectations and are subject to certain risks. Please refer to our press release and our SEC filings for a more detailed discussion of those risks.

Let me now give you a summary of our 2006 first-quarter financial results. Revenue for the first quarter was 42 million, a 43% increase from our prior quarter revenue of 29.3 million, and a 378% increase from the year-ago first-quarter revenue of 11.1 million.

Demand for our solar products remained strong in the first quarter, and our revenue was aided by incremental output from the continued ramp our second 25 MW line, and our third 25 MW line that went operational during the quarter.

On a GAAP basis, we posted a net income of 0.3 million. This is an improvement from our net loss of 0.6 million in the prior quarter and a net loss of 7.3 million from the year-ago first quarter. Incidentally, our GAAP results in the first quarter included the initial impact of FAS 123R, which cost us $1.2 million in the quarter.

On a non-GAAP basis, which is adjusted to exclude non-cash charges for amortization of intangible assets and stock-based compensation, our first-quarter net income was 2.7 million for a diluted earnings per share of $0.04. This is an improvement from the prior-quarter net income of 1.5 million or $0.02 per share, and compared to the net loss of 6.1 million from the year-ago first quarter.

Our non-GAAP gross margin in the first quarter was approximately 17% versus last quarter's gross margin of approximately 20%. As we discussed at the last quarter's earnings conference, we expected our gross margin to decline to allow for increase in silicon prices and incremental costs from our ramp. The silicon price for the quarter was actually as we projected, and it was partially mitigated by the acceleration of our thin-wafer initiative. This gave us a gross margin -- a base gross margin of 18%. However, during the quarter, as Tom mentioned, we accelerated our investment in our advanced panel assembly operations, and we also converted a significant portion of our Philippine employee-based stock options to stock units, which resulted in an initial compensation charge. Both of these factors impacted our gross margin by 1 percentage point to 17%.

Going forward, we expect our silicon price to be stable and our thin-wafer programs to continue. Also, with additional scale from the ramp of our manufacturing lines, we expect our gross margin to be at 20% in the second quarter and consistent with our prior guidance, we still expect to exit 2006 with gross margins around 25 to 26%.

Briefly on the balance sheet, we ended the first quarter with cash of approximately 117 million and no debt. Our DSO improved with Accounts Receivable increasing less than our sales growth, and our net inventory closed at 45 days. Our capital expenditure for the quarter was approximately 20 million, and depreciation for the quarter was 3.3 million.

Let me now turn it over to Tom, who would lead us through the Q&A session.

Thanks, Manny. I will open the call to questions now. I do have with me Peter Aschenbrenner, our VP of Sales and Marketing, and Julie Blunden, our VP of External Affairs. So, we will take questions now.

Thank you. (OPERATOR INSTRUCTIONS). Tim Luke, you may ask your question.

Thanks very much. Just a quick question on the gross margins, Manny, if you can give us a sense of how you would expect the -- you (indiscernible) 20% approximately in the second quarter -- how you would expect that to trend in the third and fourth quarter in terms of the degree of trajectory to exit. I think you said 25 to 26%.

Sure thing, Tim. Gross margin in the second quarter at 20% is practically coming from our contribution margin on incremental sales of almost 30 to 35%, so that's something we're looking forward to.

Going into Q3, we see that trending up to 22 to 23%, and Q4, like I mentioned, ending the year at about 25 to 26%.

I was also wondering (indiscernible) if you guys could provide some updates on just where you've seen the demand in terms of regional trends and on regional trends in terms of demand and how you see that go forward.

Yes, I will answer that question, Tim. This is Peter Aschenbrenner.

The high-level trend in regional demand is relatively faster demand growth in some of the new markets like southern Europe, Spain and Italy and Korea, and relatively slower demand growth in Germany compared to last year, which was very rapid.

Okay, thank you. Just from an overarching view, you've provided for the quarter, the second-quarter revenues and EPS and for the full year, you provided a revenue range. Is there an EPS range that we should think about in conjunction with that new revenue range?

Yes, we normally don't guide to that, Tim, but one thing I could say is that's better than the Street mean that are currently out there today.

Okay, thank you so much, guys.

Paul Leming.

Soleil Securities. I've got a couple of questions if I could? I may have missed it, but did you indicate what your shipment or ASP level was for the first quarter?

No, we didn't. This is Peter again. Our shipments in Q1 were about 11.8 MW with a blended ASP, as we had forecast, around 330 -- $3.30 a watt.

Okay, thanks. The second question, what were startup expenses and costs of goods sold in the first quarter for -- if you had any remaining on the ramp of the second line -- and maybe there were none there, but particularly on the ramp of the third line that you bought on in the quarter?

This is Manny. Two things that impacted our startup costs, if you will, in terms of continuing the ramp of Line 2 and turning on the ramp of Line 3, our depreciation expense essentially increased by single-digit million to cover for those incremental fixed costs. They were practically absorbed by getting more units out of the factory, obviously.

And no real -- other than the added on depreciation, no other real startup costs out of the ordinary related to -- (technical difficulty) -- issues? Is there anything? I know you said that the line ramped up quite quickly but nothing we should be thinking about, an unusual cost that would go away?

So, this is Tom Werner. I would say two things -- one, that recall that we started using 190 Micron wafers on our third line, and that meant that we were also wafering at the center wafer thickness, so there is a small cost associated with that. Then there was a relatively significant cost that we accelerated the development of our new panel manufacturing facility, and we chose to -- our concept happen faster, and when we did that it there was some impact in the quarter because of that. But that will allow us to ramp that facility faster. Obviously, we thought the upside was worth it.

Thanks. One last question -- you talked about purchase orders and contracts on polysilicon, and I presume assuming that some wafer, as you look for '06 and '07. Could you talk a little bit about the difference between having a purchase order and having a contract? Are the purchase orders in your mind absolutely as firm as guaranteed as some of these long-term contracts, or even though you've got a purchase order, is there some uncertainty in your mind as to whether or not, maybe in the case of the ingot, whether that supplier has got the polysilicon lined up to produce the ingot?

Okay, I will take both parts of that. The first question is our contracts versus POs and then the second one, the way I would split it is, you know, the likelihood of supply and concerns over of an ingot supplier having poly. So the first, you know, sort of broadly, I would say that we would consider a contract to have more teeth, and that's because contracts tend to be longer in term and that's why they are contracts. They also tend to have penalties -- or because you spent a lot of time creating them, they inherently have sort of penalties built into them for getting out of them. Then sometimes explicitly they do.

For purchase orders, they tend not to be multi-year. In some cases, for us, they are a couple of years. But they tend not to go beyond that timeframe, so there's a timeframe difference.

But on both of those, our experience has been and our expectation is that both will be honored. So from a running our business standpoint, we plan on both of those being honored and that has been our experience.

You asked a very good question about ingot supply. Just to remind everybody, we do buy ingots, wafers and poly. Most of what we buy is either ingots or wafers, however. Now, we of course don't hope or we don't assume that our ingot suppliers have poly; we spend quite a bit of time with them ensuring that they do and in some cases helping them get the poly or collaborating with them to get the poly. So our confidence level on poly for our ingot suppliers is certainly well-bedded and just like anything else in the near-term, we have a great deal of certainty. As you go further out in time, the forecasting becomes a bit more difficult. But we spend quite a bit of time with our suppliers and have a really good sense for that and we are confident in their poly supply.

Thanks very much.

Sanjay Shrestha.

First Albany. Good afternoon, guys. Just a couple quick of questions here -- first, on the faster than expected ramp of Line 3, if it was going to be just Q2, being based on some of the numbers, Manny, you talked about, your gross margin number would have come at around 19% for the quarter? I just want to confirm that.

No, Sanjay, our base margin, inclusive of all that already, would have been 18%. The impact of the two initiatives that we accelerated, as we've mentioned earlier, cost us 1 point of margin. We thought those were all worthy of accelerating because it benefits us sooner than later.

Got it, got it; absolutely. Talking about the panel assembly that you guys are looking at right now or I should say starting to have presence for, I mean is that more from a cost standpoint, the integration benefit that you're going to get, or is that also the mix of like a higher quality and continuing enhancing sort of the reputation in the industry as like one of the top selling the panel provider? What kind of really dictated the ramp of the panel assembly for you guys? What was your thought process behind that?

Yes, so this is Tom. You are right. There were multiple reasons -- cost quality and R&D effectiveness. What we're doing is we are automating steps in the process, working towards ultimately what we call an auto line where it will literally be automated, period. In fact, our Q3 version of this will be very close to being automated period. What it does for us is it -- even though we are in the lowest-cost regions of the world -- that being China and the Philippines -- you still can, on certain steps, make the product more cost effectively via automation, because of the incredible throughput.

Then on quality, having automation means that the steps are done identically each time, so once you've honed in on the quality level that you want, the automation ensures that it is identical every time thereafter. So we get both at the same time; we get higher quality and lower cost.

Then lastly, we are investing significantly in panel research and development. As we learn about some of our new ideas and we work through those new ideas, we can do that more rapidly by running it on our own panel manufacturing line. I will give you a specific there. Because we have an all-deck context cell design, we believe that enables us to automate more effectively. It also allows us to do some things in terms of automating the whole line and by having the manufacturing and the research and development sort of co-located, we believe we can get there faster.

Got it, got it. Okay, that's great. Also, Tom, you did mention in your prepared comments that, from a poly situation standpoint, (indiscernible) doing a portfolio management approach and you're talking some of the established and the emerging players in the industry. Now, given that, given your technological leadership, given what you're doing with the consumption reducing from 9 to 7 grams, are you seeing more of the emerging and even the established players wanting to align themselves with somebody like yourself in the industry, and giving you a higher level of confidence even in this kind of a poly environment here in the near-term that you're going to get what you need to kind of grow here in the near-term?

Yes, as we work with our -- there's a number of aspects there that I will comment on. As we work with our customer base, I think the validation of our technology and the merits of our technology are coming in quite positively. So, I think it's clear that certainly a large share of the people if not all the folks in the supply chain recognize that SunPower has a strong technology if not a superior technology in terms of conversion efficiency. People that make polysilicon are well aware of the fact that it goes in cycles and in fact has just come out -- (multiple speakers) -- that being the '90s where it wasn't such a great cycle. Those people are making big investments in their capacity, so it's logical that they would look to sign up with people that will have a strong position in the market for the long-term. Even if you can sign up or contract with somebody where there's money upfront and the cycle changes in the future, that may not ensure that you signed up with the right partner. So, just by putting money upfront isn't sufficient.

So yes, the short answer to your question is yes, there are poly producers that are definitively saying I want to sign up with somebody who I believe will win the long-term, because I don't want to go through this cycle that the industry has typically going through. So, I think strong validation of our technology in the marketplace has caused a strategic view for some of the poly producers.

On the technology front, clearly emerging technologies in poly supply would want to align with a company like ours, because we are the ones who have taken technology bets and continue to do so, so it's logical that we would do that together.

Okay, that's great. Once again, congratulations on a great quarter and a great outlook, guys.

Michael Horwitz.

Pacific growth. I just want to understand. What number are you working with for 2007 mean estimates to come up with your 80% contracted silicon?

Fair enough, Mike. This is Manny.

We're going to dig through our databook here to make sure that we have the precise number, so give us a few seconds.

The 2007 mean revenue to which we attribute greater than 80% coverage for is 346 million.

Okay. Then so what happened here in January, you were at 75% of a $335 million number, I believe.

Correct.

Okay, I just needed to understand the way those numbers worked.

Then, I show the Street at 352, but that's splitting hairs at 346, so I'm not going to worry about that.

With MEMC's announcement, can you explain kind of the dynamic there? Now you have two very big players, MEMC and then of course REC aligning themselves with competitors. How do you think that plays out with silicon? Maybe it's a different way of getting at all the same questions.

Sure. So, the incumbents are tending to, on diversified -- (technical difficulty) -- sort of the inverse of what we said and sort of diversify their customer base and also their style of arrangement with their customer base. So, you've cited a couple of deals, two that I know of, that are kind of longer range -- well, not kind of; they are longer range, in terms of duration. Neither of those were surprising to us at all. I think we've been, in our communications, pretty clear that MEMC was not giving us any signals that we were going to be on their list of partners, yet we knew there was activity in that space, so no surprise there.

The incumbents are doing the same thing we're doing, and we do business with most of the other big incumbents. So, not a big surprise and similar to our strategy is what I would say.

With regard to the ramp in the Philippines and building out, what are the capital needs there? I think you've discussed that before.

Sure. Manny will go through his book and give you a number here momentarily.

The capital expenditure in fiscal '06, which is essentially where the line for capital will be incurred and also the new building, has been estimated at 90 to 100 million last quarter, and that's still pretty much a good number.

Okay. Then lastly, I think you do it in your Q sometimes but maybe you can help me now. What did the sensor business contribute this quarter, on a revenue basis or a gross margin basis?

On the top line, we got about $3 million from the sensor business in Q1. Margins were in line with historical margins. We don't break that out typically.

So the split would be, obviously, substantially the solar business with the $3 million contribution from the imaging business. Then splitting the solar business, we're not doing that because we can't, because it basically exposes pricing with one our customers and of course, we are under confidentiality agreements to not do that. So, the split that we do give is between imaging and solar, and you can do the math. It's obviously substantially the solar business.

Perfect. Then lastly, I think Julie is on the call. There's up bill on Capitol Hill right now to extend obviously the federal solar incentives. Then I believe they are also talking about renewable energy credits going national and renewable portfolio standards going national. You usually have a pretty good handle on things. Do you have any more color there?

We definitely to have a lot of activity at the federal level, and we're pleased to be able to say that our National Trade Association is doing a wonderful job representing the whole industry and looking to grow the markets across the U.S., helping to support all of the great work that's going on on a state-by-state basis.

All right, perfect. Thanks, everyone.

Jesse Pichel.

Piper Jaffray. Your gross margin guidance for the balance of the year was a big relief. I think investors were concerned there that your gross margins in Q1 were due to poly prices. So let me ask you. What confidence do you have that you're going to hit these strong gross margin increases there for the balance of the year? Maybe you could talk about your targets there in terms of where you see poly prices going. In other words, what are the poly price increases that you've built into your guidance? Maybe you can also shed light on how much of your contracts and POs are under fixed-price.

You bet. Okay, so let me parse that up a little bit. First of all, when we do guide, we guide because we believe we're going to deliver that, so we're confident in our guidance. Then let me of course your question is sort of heal that back.

As you look at Q1, we had previously discussed Q1 as a transition quarter. Why did we say transition? That's because we brought on depreciation of Line 3, but it didn't that materially contribute to Q1 where it does in Q2.

Secondly, we accelerated investment in this panel manufacturing facility, which is a purposeful decision, and is now built into our ongoing run-rates, whereas we decided to make a bigger investment sooner.

Then thirdly, we had one-time transition from stock options to stock units, which essentially moves the expense from where options normally show to COGS. That's largely a one-time, although there's an ongoing number component to that as well. So, you have those three things. In one case, in one case the expenditure is built into the future. In the other case, other cases they were one-timers, largely one-timers and then you have revenues increasing substantially. So that's how we come up with our gross margins and our confidence level.

In terms of polysilicon prices, I know, Jesse, you know this, that we buy mostly ingot and wafers, so the concern with poly prices would be deeper into the supply chain, although we do buy some poly. I don't plan on giving specific costs of what we're paying. What I would tell you this that, when we started the year, we gave guidance on what we thought we would be paying per kilogram on ingot, and that is consistent with what we expect to be paying the rest of the year. That's pretty consistent with poly prices in the $75, plus or minus range. That has a lot to do with Sanjay's question of poly producers aligning strategically. If a poly producer were to sell poly on the spot market, they could get a higher price than that. When they sell in a long-term strategic arrangement, they sell typically in the 75 plus or minus range. But on ingot, which is what we typically buy, that's sort of the starting point.

So, assuming that you might have to go to the spot market to -- (technical difficulty) -- the expectations, gross margins really should not be impacted?

Yes, that's really the point about managing to a portfolio -- is the short, intermediate and long-term, the short-term stuff is spot. We would manage that such that the spot percentage is a low percentage of what we buy. The reason why we wouldn't have all long-term is because that's sort of a buy high/sell low strategy. Poly obviously is at its peak pricing in many years if not decades. So, we try to have a balance and we're not buying a whole lot on spot, so the impact of spot wouldn't be big because it's a small part of the mix.

Then of course, as we've talked about previously, we've ramped up our yields, our conversion efficiency and our ability to run thin wafers slightly better or ahead of plan for the year, and so we are able to absorb some of the spot pricing by virtue of better operating performance.

Is there any difference for 2007 versus what you've commented here for 2006, in terms of that 80% number? Is that also built around a pretty stable poly price? Are you building in expectations there for spot prices into 2007?

Yes, we are at a high level. It's like any other business. We are really focused on the next few quarters and then setting things up of course for '07 but the specifics for the next few quarters. What we're building into our plans is paying, on the spot basis, paying more for poly, because we do believe, similar to you, that the shortage goes through 2007. So we're building into our financial forecast paying more for poly in that time frame.

I will also answer a question that was asked previously that, in terms of pricing, I would say a majority of our ingot and/or wafers is priced for this year. Then well over half of it is priced for next year. when I say that, the percentage that is not priced yet, we do have discussions in general agreement with our suppliers. That isn't necessarily why we haven't concluded things. There's other variables that have to be concluded, the ultimate quantity that we are going to receive, for example; the timing of shipments, whether it's in wafer or ingot form. So there's a number of other variables that we are concluding with our supply base, and that will firm up simultaneously pricing.

Thanks for that clarity. Great quarter.

Rob Stone.

It's Cowen & Company. Two questions, if I may? Tom, could you elaborate a little bit on what is in the cost of investments in the modular assembly area that would flow through COGS right away? If you are not yet assembling modules, is it additional people, some kind of startup costs for the facility? Is that in the depreciation figure that Manny mentioned?

Yes, let me give you kind of a broad answer and then Manny may add one a bit to that.

Yes, you are correct; it's both. We accelerated hiring people, so that obviously is ongoing. We also moved into a new facility, and we started to facilitize, of which some of that we're not capitalizing, so we've got a one-time hit for some of that.

We are also running some prototypes with vendors that we are funding that would ultimately be run in this facility. Those prototypes get expensed as well, so it's a mix of ongoing costs as well as some one-time hits. I would say it's about a quarter one-time and 3/4 ongoing.

This is Manny. I don't want to minimize the issue but I just want to put it in perspective. The two factors that we've talked about here, the acceleration of the panel plant and the conversion of the unit plant, cost us $400,000 in the quarter, and they were about half/half for both factors. Unfortunately, at this level of sales of 42 million, that's the percentage point of margin, but I just want to put some color on that in terms of size.

No, that's extremely helpful. Thank you, Manny.

My other question, Tom, was just going back over your progress on grams per watt. It looks to me like you're going about twice as fast as people might have originally anticipated through a combination of wafers and efficiency. Can you run through the three lines and give us a sense for what wafer thickness Lines 1, 2 and 3 are running? I assume Line 4 is going to start at 190.

I sure can. So, very quickly and very specifically, Line 4 will start at 194 (indiscernible); Line 3 is running 190; Line 2 is running 190 and 220; and Line 1 is running exhausting the 255 inventory as well as running 220. So as we speak, we have three thicknesses running on the line. Then in the future, it will converge down to all 190 or thinner as we get to the end of this year and early next year.

I can't help but comment a little bit more about the importance of not only grams of polysilicon per watt but the conversion efficiency advantage that helped us get to that. The benefit of that is also accrued as you go deeper into the value chain, so when you make a module, you have more watts for the same effort, so that effort is divided by more watts. Then clearly, when you install that system, you need to install less panels to get the same size system up. So as we reach industry-leading or certainly top-tier polysilicon per watt numbers, we have the added advantage of the supply chain advantage, both in modules and in installation.

Yes, on the modules, the recently announced higher power modules, were you also able to wring a little more efficiency out of just the module assembly itself? In other words, more watts out for the same wafer content?

Certainly, I will comment on that just really quickly, and Peter may add to it. So yes, that's exactly right. For the same effort, you are now dividing by 220 instead of 200 or 210. Simultaneously, we're getting better at module manufacturing with our partner in China, who has been an excellent partner with us. So, we are seeing some economies of scale while simultaneously being able to divide by a bigger number. That's why partially why we accelerated our vertical integration, which by the way will be in parallel to our partner in China. We plan on doing both. So, I think that answers your question.

Can you roughly split the -- as you go from 9 to 7 grams per watt, can you parse proportionally how much of that is from wafer thickness versus conversion efficiency?

I can. I will be doing it off the top of my head and live, so we trust you will appreciate that it's a rough estimate. I should clarify, by the way, that the less than 7 grams would be on our fourth line, and then our plan is to retrofit the first three lines in time, so I think you've got the idea that we will have a mix of the lower gram lines with still low but not as low lines.

The short answer to your question, as I work the math through my head, is roughly 50-50 conversion efficiency with thinner wafers. By the way, when I say thinner wafers, it's thinner wafer and less curve or what we call pitch.

Great; thanks very much.

John McGinty, you may ask your question.

Credit Suisse. Good afternoon. I guess the first question is can I just get a clarification on the issue of price? Even though some of the contracts have not been fixed because you are still negotiating, is essentially, for all practical purposes, the price set that you're going to pay over the balance of the year as far as you are concerned?

The short answer is yes.

On the 80% next year, how much more variable or potentially variable is the price on that 80%? Not on the other 20, but on the 80%?

So, I will give you a little bit of data so you have a sense of the dataset that I have when I'm answering that question. Our ingot suppliers have a number of sources of polysilicon, and so as they look at their sources of polysilicon and their portfolio short line and intermediate-term deals, they build that into their pricing forecast. Most of our ingot suppliers have reason to believe they are balancing the short spot prices with better long-term agreements and/or cost reductions by virtue of being able to use lower-cost poly in higher proportions to get our ingot specs. So, given all of that data, that suggests to us that pricing would be materially the same as '07 or marginally up.

Okay. The second question, if I could? Manny, you said in essence that your gross margin was 18% before the two factors that brought it down to 17%.

That's correct.

If I go back to the transcript, which is kind of unfair, but if I go back to the transcript in January and read your answer to a question, you said we actually have a shot at staying at the 19% or even staying flat with the fourth quarter just ended in the quarter. So to some extent, a quarter ago, you thought you might do 19 or even 19-plus%. I guess what I was trying to understand is you did 18. Is that because you accelerated the ramp of Line 3? Was that because silicon prices were higher, or were you just a little too optimistic or could you kind of square the 18% with what you had said in the Q&A a quarter ago?

Sure. Several elements -- and I want to at least enumerate which one spanned out exactly and which one came a little short. Silicon cost was as projected. (indiscernible) cost was as projected.

I'm sorry, the what cost?

Depreciation. So, if there's any element there that we were counting on the upside that could have potentially helped the margin back to 19%, it was more benefit from the thinner wafer initiative that could have offset some of those costs. Albeit we got to where we wanted to get to during the quarter; there was a lot more pain getting there. So, we had to pay a little bit along the way. So, we did benefit from the thinner wafer initiatives, but that was comprehended on the 19% was get a little bit more juice out of that, and we didn't get it all.

Going up to the 20 and 21 or 2 and ending at 25 to 26, 20 to 22 3 to 25 6, that's all just basically incremental volume and with a 35% incremental margin? Is that by there?

The biggest contributor obviously is that. You do get a lot of contribution margin from incremental sales, just from scale, and we also expect thinner wafer initiatives to continue to help that as well.

Then the final question, when we look at -- you're giving us earnings before -- the non-GAAP earnings that you're giving us have the stock-based compensation and the amortization of intangibles. The question I have is do those numbers that we saw in the first quarter -- are those essentially what the numbers will be in the second, third and fourth quarter?

Yes, they will be. The amortization of intangibles is pretty fixed at 1.2 million for the rest of the year and even next year. The stock-based comp, which is predominantly the FAS 123R, was 1.5 million in total in Q1 and expect that to stay on for the rest of the year.

That's after-tax or pretax?

Those are all pretax.

So 1.2 per quarter for the amortization and 1.5 for the stock-based comp?

That's correct.

Okay, I just wanted to get that because some of us have to put our numbers out with a policy after those costs. I understand you give it the other way, but I just want to make sure in doing that. Thank you.

Jeff Bencik.

It's Jefferies & Company. I just want to get a clarification in terms of the percentage of your silicon contracts that are on your purchase order, and the percentage that are under contract for '06 and '07.

Yes. For '06, the sum of the 2 is 100%, actually slightly north of 100%. Actually, most of that is on contract; it's about two-thirds/one-third. Then for 2007, as we said, it's north of 80% coverage, and that is still about the same of two-thirds/one-third -- contract to P&L.

Okay. In terms of the general market conditions, how much are you seeing that is actually even available on the spot market to purchase, if you want to?

That's a good question, and that's why we shouldn't spend a lot of time on the spot market, because there isn't much of a spot market.

You can occasionally -- our experience has been if you have a strategic relationship, you occasionally get a call, maybe once or twice a quarter with some small amounts of -- and it varies also, Jeff by quality of the silicon. So, to get pure, new polysilicon, it's maybe once a quarter event that you have an opportunity to do that -- for us, anyway -- maybe twice.

There is [tot] scrap in tops and (indiscernible) and wings that show up a little more frequently, and if you're willing to use stuff that was even less pure than that, the frequency is even higher. The less pure stuff is not something we pursue. So you're right; there is a spot market but it's a couple of times a quarter sort of thing.

Okay. Just in terms of can you give us an update on how your negotiations are going for additional silicon supply for '07 and '08 and beyond? Should we expect something in the next three months on that, or just an overall update?

Sure. We believe that discussions are going well. We have quite a number of discussions going on simultaneously, or I should say negotiations going on simultaneously. You know, my experience with doing deals is it's not over until it's over, so I would be loath to put a timeframe on any of the above. It's certainly rational to expect activity in the near-term, whether that's a month, two months or five months; we will see how things play out.

But we do bring to the party the technology as well as we do have expertise in that side of the value chain that we can bring to partners. The combination of those things means that we have quite a few things that we are in negotiation on.

Sure. Are the numbers you're talking about, in terms of tonnage of silicon, sort of enough that if you are successful you would be able to meet like consensus estimates for '08 and '09?

Yes.

Okay. Same thing in terms of pricing.

Yes.

Okay. If we can switch gears a little bit, in terms of the panel manufacturing plant, when exactly do you expect that to be fully operational?

The plan is during the first half of Q3. We would start that -- and that is like a new facility, so there will be a ramp associated with that. That would substantially happen in Q4.

By the way, we are ramping our partner in China substantially. Again, I want to put a plug in for them; we are happy with that relationship and intend on keeping that relationship.

I guess what I'm trying to get to with the next question is what is the incremental benefit of you manufacturing it yourself instead of using them? (multiple speakers) -- you talked about?

We measure, just like you do cells, you measure cost per watt. This year, in 2006, during the first two quarters of the year, it will be dilutive, neutral roughly in the last two quarters and then accretive starting in 2007. You're talking about changes of pennies per watt.

Now, the longer-term, though, when the get out, say, three or four quarters, we're looking at being able to pull 10 or 20% of the cost of panel-making out, minimally, so I'm not sure it longer-term. Four quarters is sort of intermediate term.

Okay. Then, just in terms of your R&D and G&A costs, what should we expect for that throughout the rest of the year?

It's Manny. Just for near-term guidance, our OpEx, which include R&D and SG&A, was 5.1 million or 12.3% of sales in the quarter just ended. For the ensuing quarter, we see that going up to 6.1 million.

As we fund incremental R&D effort for our second generation technology, the one Tom already talked about, we're going to run more experiments and materials outside for calibration, etc., and for sales and marketing as well, so we continue to invest in the efficiency of our outbound channel. That's where most spending is going to be. That would put us just a touch below 12% of sales, in terms of operating, and we should exit the year closer to about 10% of sales for those two items.

Okay, that's all I have. Thank you very much.

Robert [Shrem], you may ask your question.

Independent Research-Germany. Congratulations on a good quarter. I'd just like to -- a couple of questions. The first one, regarding the megawatts of production underlying your 2006 and 2007 revenue guidance, if you could give us an indication on that, please.

This is Peter. I will take the first part. Our projections for 2006 are based on megawatt shipments of between 60 and 63 megawatts.

Okay, thank you. Regarding the wafer (indiscernible) and the improvements you are able to make, in comparison to competitors like [Ucell], how do you think how long is your competitive advantage in terms of months until competitors are able to get to the same level as you are currently?

Your question is specifically on grams of polysilicon per watt?

Yes, and also on wafer thickness, because when you compare your 190 microns, it looks extremely good compared to what is currently regarded as another leading edge of the industry (indiscernible) 200 microns, which they just said they plan to achieve in the first half of 2006.

Robert, this is Peter Aschenbrenner. I will take that question. There's really two issues related to thin wafers. The first has to do with action the fabricating the wafers in the factory, and there, it's really a question of wafer handling and processing. I think we've shown that one of the things that SunPower does very well is understanding process conditions, specing equipment properly to do the job. I think that's one of the reasons that our toolset is actually quite different than a conventional solar cell factory. I think all of those things together allow us to actually fabricate thinner wafers relatively easily, more easily than our competitors.

Secondly, unlike a conventional solar cell, our products actually become more efficient as they get thinner, whereas a conventional solar cell tends to lose efficiency as you get thinner. So, in one case, you're kind of swimming upstream on one hand, gaining silicon efficiency and on the other hand, losing conversion efficiency, whereas in the case of the SunPower solar cell, we've got the wind at our backs and we're gaining in both metrics. So we have a more -- a clearer driver in terms of pushing to thin wafers that's rewarded both in lower silicon utilization and higher efficiency.

Finally, something that's a little underestimated I think is what you do with the solar cell once you get out of the cell factory. That is, when you put it into a module. I think that as the cell manufacturers are pushing on wafer thickness reduction very hard, what we're finding in the field is that the module assemblers are having an increasingly difficult time taking those thin wafers and turning them into modules with high yield. This is again one of the core advantages we think we have going forward, in that our all-back contact cell enables fundamentally different interconnection techniques and is, we believe, much more amenable to module production with low-cost cells. Keep in mind that SunPower has manufactured very thin cells in its history, 150, 160 microns, so we are comfortable with how to handle those.

Okay, great. Thank you. The last question from my side would be you mentioned earlier in the conference call that the German market kind of showed a slower growth compared to other markets, especially in Asia. How do you see the importance of the German market development going forward in 2006 and 2007? Do you see any danger of losing support from the political side because of high module costs and less drivers for gaining cost advantages in the market?

We see the German market as a critical market for the industry. It's clearly the largest market today. It grew very rapidly between 2004 and 2005. My information says that it grew something in the neighborhood of 45%.

We expect slow growth this year because of higher module prices, as you indicated. But we think that that's being moderated by improvements in the channel, in the delivery channel, so that the cost to the end-user is actually tracking somewhat to the decrease in the feed and tariff of about 5% this year. So we think there's a high level of awareness of the need to continue to reduce end-user price over time to keep the political support we have in Germany. I think that most of the major suppliers are aware that and taking the appropriate actions.

Okay. For background, could you give us a little indication of what your pricing power is in terms of percentage difference in regional markets? What typical costs of a module for the end-user can differ in different regions?

Yes, I will take that again. This is Peter. We don't see pronounced differences in our regional pricing any more. I think this has been one of the effects of a short supply, is that significant differences that may have existed in the past due to currency fluctuations or local market (indiscernible) tariffs for instance have largely been washed out and we now see relatively similar pricing all around the world.

Okay, thank you very much.

Kevin Monroe.

Thomas Weisel Partners. Good afternoon. I was just wondering if you guys could give me your thoughts on -- given the tight silicon supply in 2007, would you ramp capacity, even if you don't have silicon supply, in anticipation that supply comes on in 2008, so maybe you sacrifice margins in '07 to ramp up really quickly in '08 when supply is a lot looser?

We are constantly balancing capacity addition with silicon supply and technology -- what technology we're bringing online as well. I guess I would broadly state to you that yes, we have a forward bias, an aggressive bias, so we will add capacity aggressively. If we do end up erring, it will be on the side of having capacity before we have silicon but we don't plan on erring, of course.

Okay, thank you.

(technical difficulty).

Yes, it's National Bank Financial. Tom, you mentioned that Line 4 will have both the high-efficiency panels and the thinner waferness, and then Lines 1, 2 and 3 will be retrofitted. How difficult is that? Is there a lot of cash involved in doing that?

It is not all that capital-intensive to retrofit lines. I would say, on a percentage basis, it's less, far less than 20% of the original cost of the line, probably on the order of 10. Of course, that's evolving because we haven't implemented the fourth line yet. Until we do that, we can't nail that cost down completely.

We think we're going to be able to do that pretty readily in terms of timing, so we are again finalizing those plans as we get closer and closer to Line 4 coming online. But on the order of 10 to 20% of the original cost is a rough estimate, and the timing would be quarters, single-digit quarters certainly, maybe just a few quarters to get it done.

That's likely in early '07.

During '07, first half is what I would say.

This new automated panel assembly facility, what's roughly the cash cost of that? It's obviously in your 90 to 100 million CapEx for the year, but is it a big amount?

Correct. It's just a touch higher than 5 million, 5 to $6 million, so it's not terribly capital-extensive. But the benefit you get from that investment is something we're looking forward to.

Okay. Manny, I think, last call, you talked about the total capital expenditures for your whole expansion plan, right up to the second facility and six lines. I think you said $230 million. Is that still a roughly good estimate?

Yes, they are still looking as a good estimate, trying to beat that obviously. I think Tom recapped it last time as between 200 to 250, so the 230 is still right there. That is predicated on a $0.90 per watt capital on 200 MW lines and 50 million worth of buildings, which is already budgeted for '06.

Okay. Then, just talk a little bit about your top customers in the quarter -- any changes there from previous quarters?

No. Our top customers are still Conergy and SOLON, both active in Europe. Those are the two that exceeded 10% and are growing sales to (indiscernible) under our new long-term supply agreement with them.

Okay, the same sort of ratio as was in your 10-K?

Yes.

Okay. Just coming back to wafer thickness again, how low do you think you can get wafer thickness? You mentioned maybe below 190. I note that Sharp is commercially a 180 and I think [Qcell's] is down something at 130.

Yes, we tend to talk about what we are actually shipping, so I want to be a little careful about speculating on the future. The heritage of the Company, though, is thin wafers. I believe, in Cypress' annual report last year, they had the Helios aircraft that had sun-powered solar cells on it that were built in the late '90s. Those are actually 165 Micron thick wafers. So the Company has a heritage of much thinner than 190. Let's just say that we are working towards those kinds of numbers, but we are constantly reevaluating the entire supply chain, the wafering costs, the costs of running it on our cell line and the cost of module making. So we are sort of committing to 190 and a little below that as we look at Line 4 and then evaluating below that, subsequently.

Okay, great. Thanks.

Paul Leming, you may ask your question.

Soleil Securities. I just wanted to come back if I could for a minute to the issue of cost of goods sold and margin. Revenues, Q4 to Q1, were up $12.5 million in round numbers and costs of goods sold are up a little bit less than $12 million. We've kind of gone through lots of pieces of that over the course of the call. But I'm just trying to get my hands around kind of what your view of the real, underlying core incremental profitability of the business is as you grow it. As I try and model Q2 numbers off of your guidance, it looks to me again like you're going to have a relatively low fall-through of incremental profit in the second quarter. Am I missing something? Is that right? What have we not talked about that did cause such a larger quarter-to-quarter increase in costs that there really was not a large flow-through to gross profit in the first quarter?

Manny will take that question. Just so everybody on the call knows, we have two more questions after this and then we will end the call. So, go ahead, Manny.

Yes, let's start with Q4 to Q1, Paul. I think we've talked about all the elements now that contributed to the cost of goods sold there. Obviously, silicon was expected to increase, and it did increase as we expected -- (multiple speakers).

Have you ever indicated what the actual percentage increase in silicon costs were, or in raw material cost were, quarter-to-quarter? You've said they came in in-line with your estimates, but do we really know if your estimates were up 3% quarter-to-quarter or 200% quarter-to-quarter?

Our silicon -- I think we covered this in the last call, but our silicon costs we were saying in the range of 25 to 30% quarter-on-quarter increase.

Thank you.

That's, for the most part, Paul, was the biggest element of the increase. But we did expect that, and it did come in as expected. Depreciation also came in as expected. So by the time you take the silicon costs and the depreciation costs, the Company could actually gotten worse, but all the improvements at the factory, the thin wafer initiative that we projected, got us back to 18% base.(multiple speakers)

Have you indicated what silicon costs will increase, Q1 to Q2, underlying increase?

Fairly stable at Q1 levels, Paul. We know that already because we've got contracts put out.

Given your contracts, do the poly costs stay pretty flat over the course of the year and it's just a January 1 reset?

Materially, yes. There's single digit percentages changes throughout the year, but materially, yes.

Okay, thank you. That really clarifies what I was after. Thank You.

Tim Luke.

Lehman. Manny, just a quick couple of factual things. Your other income this quarter was 1 million. How should we think about that going forward? I was wondering if you could just give us a tax number that we should use going forward over the next few quarters. I think you said that SPG was 3 in the quarter. Should we just sort of flat line that through the rest of the year?

Do the last question first, yes, I would flat line SPG at -- (multiple speakers) -- level.

To the interest question, no change from last quarter's guidance. We expect Q2 at 500,000 income, Q3 at 300, and Q4 at 100. Just representing utilization of cash.

So you exit the year at around 50 million in cash, is that how we should think about it?

About 60 million, Tim.

60, cool. Then the tax number?

Tax is still guided at 5% for '06 and 10% for '07. For those of you who just want to model the truly effective tax rate -- 25%.

(indiscernible) you said it was 20% for '07, no, a slight reduction?

It's looking good at 10 now, Tim.

Okay. Thank you very much, Manny.

Jesse Pichel.

Jesse Pichel from Piper. To quick questions -- what will the automated module assembly line do for your panel efficiencies? And two, I think the bigger issue long-term for investors in SunPower is does really higher efficiencies mean higher margin longer-term? Sure, poly is a major variable in that equation and you're using less of it, but how much does it cost you to get higher efficiencies, in terms of either more equipment or -- which is depreciation -- or higher variable costs, in terms of labor or something else?

So, the new panel manufacturing will have R&D changes built into it that will affect the amount of watts as you go from cells to modules. That doesn't materially drive our decision, but we are able to influence that. We've made some progress there during this quarter as well. Of course, every extra watt really matters because people buy panel watts, not cell watts. So, and the other thing is that, in time, we've gotten a number of other R&D projects that we will be able to iterate or do cycles of learning in our new panel facility, we believe, faster because they are essentially co-located.

I will comment briefly on the importance of conversion efficiency and certainly Peter will comment briefly and then I will close the call. Conversion efficiency is leveraged throughout the value chain, all the way from a wafer through cell-making, module-making and ultimately installation in the field. I'd don't think that there's any solar company or any polysilicon company or silicon company that is suggesting that silicon prices will remain at today's prices. I think everybody is forecasting that they will abate, come down in time, that it's just a matter of when. When that happens, the power of efficiency -- it will clearly flow through our income statement and you will see the power of efficiency. We've obviously done the math. We've been very open that we do have higher capital costs. That higher capital cost is more than offset by the higher efficiency of just cell-manufacturing. Then we accrue a benefit in the module and installation side of the value chain. So, we are still very bullish on the power of conversion efficiency and as the silicon price abates, you're going to see that flow-through and of course as we scale our manufacturing operation.

Do you want to add something, Peter?

Yes. I wanted to say that, you know, we are still early in the game here, in terms of bringing this very much higher-efficiency product to the market. I think you have to keep in mind that, historically, the difference between products from most of the major manufacturers was very small in terms of efficiency. So, the channel is really just starting to discover the types of efficiency value that are recognized when you transport the product to the site, put it up on the roof, and the labor savings, etc.

In terms of overall cost reduction, I will point to one of our competitors who recently, in their conference call, stated that a 1% increase in cell efficiency leads to a 7% cost reduction at all levels of the value chain.

Now, obviously, this is slightly different for each process, but I think this is a growing -- an issue of growing awareness in the industry, and we're out there pushing very front lines.

Great, thank you.

Okay, thank you all for joining us on the call today.

In conclusion, I would like to cap our conversation. We are executing on our plan and hitting our numbers. We grew revenue 43% sequentially from Q4 and 278% from Q1, 2005. We were GAAP net income positive in Q1. Including intangibles and stock expense, we delivered Q1 non-GAAP net income of $0.04 per share. We expect to grow revenue again by 24% or more in Q2. And we increased our revenue guidance for the full 2006 fiscal year to in excess of $220 million.

Once again, thank you for joining us. We appreciate the support of our stockholders, and we're working hard to grow SunPower into a market leader.

Thank you. That concludes today's conference call. Have a nice day.