MEMC Sees Solar Boosting Prices

Nabeel Gareeb, chief executive of MEMC Electronic Materials (WFR), which makes polysilicon wafers for computer chip manufacturers, has taken the stage at JP Morgan’s technology conference in Boston. He’s making a sharp, rapid-fire presentation of how the solar panel industry will boost prices and volume of his company’s wafer production over the next several years. He also seemed to tip his hand that the company may get into making wafers for exotic types of semiconductors based on, perhaps, silicon germanium — though that’s just my conjecture.

At any rate, MEMC shares are trading down 1.79% at $68.61, after trading up slightly in the pre-market, during the talk. Nothing I heard in the talk sounded too negative.

Gareeb starts with a few stats contrasting his traditional semiconductor market to the solar market:

The semi wafer market is a $10 billion market growing at 10% per year, compounded, over the next several years. Solar, on the other hand, is a $4 billion market growing at 40% compounded. MEMC, he says, has signed long-term agreements that will be worth $15 billion to $18 billion over next several years. Nabeel foresees earnings growth of 30%, compounded, “for years to come,” while getting to $4 to $5 billion in sales in 2009. MEMC’s total wafer production will rise from 4,000 metric tons of polysilicon — the basic ingredient of its wafers — in 2005 to 15,000 metric tons in 2010. “That will be worth $5 billion, and the vast majority of that is in solar wafer contracts,” he says.

Next comes the Q&A: How’s demand trending to polysilicon in the industry? “Very robust in the solar market,” says Gareeb. “And it continues to increase in density.” Asked about prices for wafers, he says “Spot prices have continued to go up. What’s happening – it’s a very interesting phenomenon – as new entrants come online, their poly[silicon] isn’t necessarily of the quality to make a wafer in its entirety, with the result that they end up mixing their poly with our poly. Our customers have told us they prefer the quality of our poly, which allows for us to price it at a premium.”

The volume ramp in solar wafers: “When we signed Suntech (STP) the year before last – that was $5 to $6 bil worth of product over the next ten years. The way you can think about it is, start with products in the hundreds of millions [of dollars] and just triple that over ten years. We want to grow much faster than the 40% rate of the industry because we want to get from zero to midteens market share."

Solar, says Gareeb, is going to bring up the company’s average selling price on wafers. Each deal the company signs, prices for solar wafers go lower, but solar has higher margins than semi wafers on average, and the deals the company is signing are staggered to achieve a favorable blend. “The whole intent was to layer [the contracts] over time, mitigating the effect of price reduction,” in each contract, says Gareeb. Because margins of solar wafers are higher than the company average, as the volume increases, "the mix leads to an improvement in our margin.”

On the of bringing adequate polysilicon capacity on line at the company’s Unit 3 manufacturing facility: “I have to tap dance around this a bit, or I would have to file an 8K every week,” jokes Gareeb. “When we did our earnings call in third week of April, we said that unit three was running at 80% to 90% capacity. After a proactive shut-down following a chemical leak, those units are back running in the same or better fashion prior to the shutdown.”

Last quarter, MEMC was down to 12 days of inventory. Is that a problem? “There’s some confusion about how much inventory contributes to sales in our case,” says Gareeb. “In Q4, our inventory went down but sales were up about $60 million from Q3. What happened was we had some inventory write-down. We wrote most of that stuff off and took a hit to our gross margin and dumped it as scrap. But where we are today, with 2 to 2.5 weeks of inventory, I think we’re okay with that level.”

When asked about the prospect of solar achieving grid parity with other forms of energy, Gareeb had this to say: “Our plans have assumed oil is at $40 a barrel and that solar is not subsidized. Obviously, with oil being where it is, grid parity can be achieved a lot sooner than the middle of the next decade.” Customers of MEMC that are focused on achieving grid parity will be the companies that will win in solar, he observes, and then feels inclined to remark, “We’re building a 30- to 50-year business model that should sustain itself very profitably.”

If solar is so fast-growing, asks an audience member, why not move the entire wafer business to solar, and just forget about semiconductors? “The semi market isn’t dead,” says Gareeb. “Look at the revenue multiplier that semiconductor polysilicon offers. Take a dollar’s worth of poly at $60 to $90 per kilogram. That’s dollar’s worth gets turned into $2 for solar [wafers] but $4 to $7 for semi [wafers]. So, the multiplier effect of silicon, combined with a large customer pool and our experience, makes the semi market still relatively attractive.”

How will wafer output ramp this year? “With Unit 4, with our poly reactors in Pasadena, Texas, and with our existing capacity, it’s roughly $800 to $900 million in dollarized capacity per quarter,” which is up, adds Gareeb, from the $700 million to $800 million a month that the company has currently with its Unit 3 facility.

People would expect you to acquire other firms in your industry. What should we expect? “If you can install an asset base for a couple hundred million dollars, why would you pay someone else $3 to $4 billion for it?”

What about going beyond solar and semi wafers? Any other interesting opportunities in the market given your technical skills? Gareeb:

Look at the wafer [total addressable market]. There are other types of wafers in the world –- I can’t talk about to much about what they are, because it would give away what we’re working on. People have for years tried to get us into the materials business. For years peole tried to convince me to get into making photomasks [the templates used for etching circuits in semiconductors]. I said no, if you just focus on wafers, there’s an opportunity for significant growth in the double digits. We will look at some other opportunities in the wafer space over time. There are similar opportunities [to solar] out there.

And then, when asked specifically about making wafers for exotic semiconductors, it sounded to me that Gareeb was tipping his hand, without directly admitting he was going into making so-called III-V style semiconductors:

The alternate substrate market is huge in terms of potential applications, but it’s the cost effectiveness that is the issue. When something becomes consumerized, you get tremendous price pressure. You have to be able to manage that with an excellent asset management model.

And guess who has an excellent asset management model, in Gareeb’s view?

Lastly, the moderator asked Gareeb, What will be the ratio of solar to semi in your wafer production at the end of the year?

You may see solar become a third of our revenue stream over three to five years. I can’t say what this year will be, but perhaps in our 10K next year.

Comments

[sausage]

The comments miss the essential point. Any industry with astronomical margins is going to attract new competition and cause exiting players to expand. This is especially true of commodity-like products like silicon wafers. Customer loyal counts for very little here. We should be asking what they are doing to keep their costs down below that of the coming rush of competition.