From Barrons TechBlog:
What’s the outlook for chip stocks?
That’s the question at issue tonight at the 12th annual Churchill Club Semiconductor Forecast dinner at the Hyatt in Santa Clara. The lineup is a good one:
Bill Tai, general partner of Charles River Ventures, is the moderator.
Christopher Danely, analyst at J.P. Morgan.
Dan Niles, CEO at Neuberger Berman Technology Management.
Mark Lipacis, analyst, Morgan Stanley.
Sangeeth Peruri, managing director and portfolio manager, J.&W. Seligman.
Stay tuned for live coverage of their observations and picks; keep reloading this post if you’re reading Wednesday night. The highlights follow:
Last year there was a slightly different set of players. Jim Covello of Goldman Sachs picked Sandisk (SNDK) as a long and Novellus (NVLS) as a short. Joe Osha of Merrill Lynch picked Texas Instruments (TXN). Niles picked Skyworks (SWKS) and Trident (TRID). Tai picked Microtune (TUNE) and Integrated Device Technology (IDTI). All of them expected the SOXX - the major semicondcutor stock index - to be a lot higher than it is today.
Nasdaq down 5%, Dow down 9%, SOXX down 17% since the last dinner a year ago. The only money makers of the picks last year: Skyworks as a long, and Novellus as a short. the other 5 picks were down. Net-net, the group was down 2%. Trident was the worst pick, down 78%, but Skyworks was up 34%, and the best long.
Danely’s overall outlook: His slides are entitled, “Is It Time to Back Up the Mini Van?” (His apparent answer: no, it isn’t.) He notes semis in ‘90-’99 had 13% revenue growth compounded, but slowed to 10% 2003-2008. That is still double the U.S. GDP, and better than almost any other industry out there. Business is tougher than it used to be, but far better than just about any other industry out there. Unit growth ‘90-’99 about 9%; 2003-2008 unit rowth is actually higher. What has hurt is pricing trends. They rose 3% compounded in the 90s, but since have been down an average 1% a year. One key reason: consumer is greater and greater percentage of the chip industry end market. Barriers to entry lower, product life cycle shorter. Also have seen increased investment. Large cap semi revenue growth from 16% in the 90s to 9% in 2003-2008 period. Total semis down to 9% topline growth, from 14%. He notes that multiples on the S&P 500 in the ’90s ranged from 13-16; still in that range now. But semi multiples have narrowed. Intel range was 10-40 in the ‘91-’99 period; Intel now trades in a range of 15-20x. Seen multiple compression. But still above the S&P.
Danely sees unit growth of 7% this year and 8% next year; he sees ASPs down 3% this year and another 3% next year, for growth in revenue of 3% this year and 4% next year. (That’s compared to down 4% in 2006 and 8% in 2007.) The first three-year period of negative ASPs since the mid-90s. (Or in this case, four years.) He says inventories are coming down in the channel, and are close to normal.
Danely says he actually hopes for a bad second half: easier to upgrade the stocks with inventory and margins low, and demand recovering.
Danely notes that the curent downturn is 8 months; he says the worst is over. He does not think it goes below 350, or another 10%-12% lower.
Danely’s overall trends: lower growth, but still good growth. Semis look less like growth companies and more like ‘normal’ companies. They need to pay more dividends, lower operating expenses, and increased focus on cash management…they need to be “more Philip Morris and less Google.” The highest yield in chips: Microchip, at 3.6% yield. Compare that with MO, which pays 5.1%; Pfizer pays 6.5%.
Conclusions: Worst is over. Hard to get excited. Demand lackluster. Margins/leverage not attractive. I’m concerned about demand, he says. Valuation is OK, on 2009 estimates. Inventory is not excessive, but not too low either. He sees another year of so-so returns.
Next: the outlook from Morgan Stanley’s Mark Lipacis. Where are we in the semiconductor cycle, he says, is a question he gets asked all the time. He says the question is not that relevant anymore. He thinks communications infrastructure is interesting for an 18 month time horizon.
In the past, valuation did not matter - as long as there was momentum, as long as the next set of data points was better than the last set of data points, you paid more, Lipacis says. In the past it was really a momentum group. Before 2001, if you simply paid the second derivative on revenue growth, you made money, without having to do a lot of analysis on individual companies. Bu he says over the last five years, this relationship stopped working. What is working? Fress cash yield was the best predictor of mid/large cap semi stock performance in 2006 for the first time in 9 years, “if not ever.” Why? Lipacis says it was a signal that semis are maturing. Growth 2005-2010 is in the 3%-9% range, down from 20-year average of 13%. Also seen margins stabilize. He contends one result is that the cycle call is now much less important.
What’s the next growth driver? Has to be something in the consumer/mobile area.
Today reminds us of the late ’90s, Lipacis says. Telco balance sheets are healthy. Telco pricing flattening out. In late ’90s, was Internet driver. Today, YouTube is the new solitaire. Debt levels in the telcos are at the lowest levels in 20 years. Interest coverage is improving, similar to where it was in the late ’90s. Conclusion: these guys can spend money if they are motivated.
So what is their motivation? Cable companies capex increasingly threatens the telcos. Aggregate telco and cable capital spending looks like the late ’90s. Telco price per minute from 1930 to 1995, dropped 7% a year. In late 90s, there was big buildout, and started falling 15%-20% a year. But now is flattening. Perhaps the excess capacity has been soaked up.
Now we get some reaction from the buysiders.
Niles says he looks at GDP cycles, not semi cycles. “I am incredibly concerned about the next 6-12 months,” he says. “I don’t think people really know what end demand is.” Oil price spikes, home values declining. Food inflation and inflation in general is something we have not had to deal with for a while; 10-year bond rate is going up, making credit more expensive. $168 billion of rebate checks will be gone. Next president will have to raise taxes to pay that back. I’m worried about recession late this year, early next year, as the rebates dry up, he says. The big problem is still inventory - an inventory of unsold homes, at 11 months. That’s the issue, he says. In 2001, there was ton of inventory on balance sheets; now its home inventory. Until then, we have a problem; I’m playing very defensively, he says.
Seligman’s Peruri says cycles had big deltas 5-10 years ago, when it was more rapid growth industry. Now a slower growth industry, and not as cyclical. If you look 10 years ago, if you got cycle wrong, it was hard to get investments right. As industry has slowed, cycles are more muted; you don’t need to call the cycles. More important to understand specific company fundamentals now. What you really have now is variation in investor psychology more than macro environment shifts. He agrees we will have a recession, but once stimulus money is spent, you will get slowing growth. The worst may or may not be over, I don’t know. Global growth likely muted, and semi growth also likely muted. So you can make money with companies in good product cycles. There are always companies ramping and innovating, and always some struggling and getting dislocated.
Danely notes SOXX has gone from 340-350 to 410; he gets calls from people looking to invest, and asking what to buy. Where they are now, people are conditioned to expect some upside. You need consensus estimates move up; but he does not think that will happen. “I don’t feel very good,” he says. Danely also says he has a view of caution on the economy. Not sure why the PC industry has not felt any slowdown.
Niles says we are seeing some slowing in the PC industry. Cisco orders in emerging orders are slowing. Most recent quarter at Cisco, they showed slowing in India. Some handset vendors saying no pick up in handsets in China post-earthquake. China stock market is down almost 50% peak to trough; isn’t this the bullet-proof market? China is building up to the Olympics; what happens on the other side? There are some interesting in this decoupling trade that people are playing; it may be different than people think if emerging markets slow.
Peruri says PC strength could be strong for longer than people think. PCs are the only end market prices in dollars; cell phones priced in local currencies; telecom priced in local currencies; industrial priced in local currencies; but weak dollar makes Intel-based PCs effectively cheaper around the world. The combination of CPU competition between Intel and AMD and the dollar depreciation could get to point where demand for PCs is accelerating around the world. PC demand could last longer because of the dollar.
Niles noted that Microsoft (MSFT) missed PC numbers in the latest quarter; they saw downshift in demand. They tried to blame other things; it will be interesting to see how that pans out; it was masked by massive Xbox demand.
Tai wants the panel to say what the growth driver is post-consumer. Danely says you never know what the next growth driver will be. He was an engineer who worked on DEC Vaxes. He though the IBM PCs were so slow, that they would never work; 3 years later, the Vaxes are gone, and everyone is using PCs. You don’t know what the next innovation will be.
Niles notes that there are interesting things in the alternative energy sector. He says the good news is, we have to get off dependency on oil. Gasoline is $4 here, but Europe and Japan pays $6-$8. That is probably the next big wave of growth in terms of chips. The problem is; semi industry is so big, you can’t have one sector drive it. You need something huge to drive it.
Niles cautions that betting your career on this is dangerous; everyone wants to be a solar company, and Applied Materials (AMAT) is making that possible.
Peruri thinks Nvidia (NVDA) has an interesting play in the use of graphics processor for computing, with 100x performance improvement over conventional microprocessors. They created the GPU market from scratch; makes you wonder if they can create another $10 billion to $20 billion market from scratch.
Sox predictions one year out:
Danely sees SOX 450, about 10% up from here.
Lipacis says 451. Take the 12 month price targets for the Morgan Stanley stocks in the SOX.
Niles: 452. I have no idea, he says. S&P and SOX will trade in 30% band. You could have another bank go under. (”Hope it isn’t ours,” he says, given that Lehman owns Neuberger.)
Peruri: I really feel likes things are going to drift up, then go lower. I’ll pick 425.
Tai: 488, but things get worse before they get better.
PICKS!:
Danely like Microchip (MCHP), which competes against “stiffs” like STMicro (STM) and Freescale. Tough to find semi company expanding addressable market noticeably. The other pick is Advanced Micro Devices (AMD), which is trading like it is going out of business. (Which they are, someone in the peanut gallery just grumbled.) Gone from grabbing market share, to retrenching and restructuring. Could be profitable by the end of the year. And could double.
Lipacis likes PMC-Sierra (PMCS), the class pure-play on communications infrastructure. Once a $250 stock, now at $8. The best performing stock in 2003. I believe if infrastructure theme is right, will see people remembering good old days. The second name is Texas Instruments (TXN), in part on sentiment call. People hate that stock. When everyone runs in one direction, you might have an idea. Nokia shifting away from them; for 8 months people have been obsessing about it.
Niles: sticking with Skyworks (SWKS). Gaining share at Nokia, RIMM, LG, Samsung, etc. Plus growth in linear business. The other long, a solar play, is Applied Materials (AMAT). Orders are down 60% year over year; semi business for them will be at some point will turn. (He owns SWKS, but not AMAT, yet.) Taiwan Semi is raising prices, sign of tight capacity. And real upside is the solar business. Start recognizing orders from solar business late this year. Applied will supply a lot of the survivors; make it easy for anyone who wants to be in the solar business to get in. And a short: the TLTs - expect rates to go much higher. Ticker is TLT, which is an ETF.
Peruri: A theme since the bubble is every company wants to invest in analog space, especially high end analog. I wonder if digital space, which everyone is shying away from, is now the place to go, and wondering if commodity analog is also place to go, since people are de-emphasizing them. Marvell (MRVL) is the digital pick; lot of great product cycles. Street has them at $1 for next year, but could do $1.20-$1.50. With 20x multiple, then $24-$30 stock, at $17 today. About 70%-75% off the bottom; a year from now it could be meaningfully higher. The second is ON Semiconductor (ONNN), which is focused on lower-end analog. And can buy assets on the cheap. They could earn $1.20-$1.50 a share as well; the stock is around $9; if you use a 10x-15x multiple, you could get 50%-100% upside. The stock is however up 100% off the bottom.
Tai: Also picked Marvell, for different reasons. They are in good segments; the drive market, WiFi space, high volume areas that are growing. The company historically executes very well.
Will we see wholesale consolidation? Danely says we should in theory, but that we aren’t likely to actually see it. Maybe some smaller analog companies getting bought. Power management is highest growth end market in analog; Analog Devices or Texas Instruments would like to get in there in a big way, he ways. Lipacis thinks consolidation is possible in the memory sector.
Niles says he can make a case that this is 2001 again. There were 6 40% rallies in the semi index on the way to losing 80% of your money in the index. This is the analog to being in the end of 2001, when the stocks dropped another 40% from there. Niles notes that investment banks are running at 25-30x leverage. Between that and the housing market, that has me kind of concerned, Niles says. The president who comes in to deal with this may be good on the economy, and protectionism would make it worse.
Someone is asking about the cost of raw silicon. Peruri notes that the market is being driven by the solar sector. The bear case, he notes, is massive over supply is coming in polysilicon. The bull case is that the cost of energy is rising, and silicon solar panel costs coming down. If reach grid parity in 2-5 years, demand could go up astronomically. About 0.1% of world power from solar. Demand could go up 10,000x. So prices could go higher. Premature to assume polysilicon prices will go higher or lower.
And that’s it.
What’s the outlook for chip stocks?
That’s the question at issue tonight at the 12th annual Churchill Club Semiconductor Forecast dinner at the Hyatt in Santa Clara. The lineup is a good one:
Bill Tai, general partner of Charles River Ventures, is the moderator.
Christopher Danely, analyst at J.P. Morgan.
Dan Niles, CEO at Neuberger Berman Technology Management.
Mark Lipacis, analyst, Morgan Stanley.
Sangeeth Peruri, managing director and portfolio manager, J.&W. Seligman.
Stay tuned for live coverage of their observations and picks; keep reloading this post if you’re reading Wednesday night. The highlights follow:
Last year there was a slightly different set of players. Jim Covello of Goldman Sachs picked Sandisk (SNDK) as a long and Novellus (NVLS) as a short. Joe Osha of Merrill Lynch picked Texas Instruments (TXN). Niles picked Skyworks (SWKS) and Trident (TRID). Tai picked Microtune (TUNE) and Integrated Device Technology (IDTI). All of them expected the SOXX - the major semicondcutor stock index - to be a lot higher than it is today.
Nasdaq down 5%, Dow down 9%, SOXX down 17% since the last dinner a year ago. The only money makers of the picks last year: Skyworks as a long, and Novellus as a short. the other 5 picks were down. Net-net, the group was down 2%. Trident was the worst pick, down 78%, but Skyworks was up 34%, and the best long.
Danely’s overall outlook: His slides are entitled, “Is It Time to Back Up the Mini Van?” (His apparent answer: no, it isn’t.) He notes semis in ‘90-’99 had 13% revenue growth compounded, but slowed to 10% 2003-2008. That is still double the U.S. GDP, and better than almost any other industry out there. Business is tougher than it used to be, but far better than just about any other industry out there. Unit growth ‘90-’99 about 9%; 2003-2008 unit rowth is actually higher. What has hurt is pricing trends. They rose 3% compounded in the 90s, but since have been down an average 1% a year. One key reason: consumer is greater and greater percentage of the chip industry end market. Barriers to entry lower, product life cycle shorter. Also have seen increased investment. Large cap semi revenue growth from 16% in the 90s to 9% in 2003-2008 period. Total semis down to 9% topline growth, from 14%. He notes that multiples on the S&P 500 in the ’90s ranged from 13-16; still in that range now. But semi multiples have narrowed. Intel range was 10-40 in the ‘91-’99 period; Intel now trades in a range of 15-20x. Seen multiple compression. But still above the S&P.
Danely sees unit growth of 7% this year and 8% next year; he sees ASPs down 3% this year and another 3% next year, for growth in revenue of 3% this year and 4% next year. (That’s compared to down 4% in 2006 and 8% in 2007.) The first three-year period of negative ASPs since the mid-90s. (Or in this case, four years.) He says inventories are coming down in the channel, and are close to normal.
Danely says he actually hopes for a bad second half: easier to upgrade the stocks with inventory and margins low, and demand recovering.
Danely notes that the curent downturn is 8 months; he says the worst is over. He does not think it goes below 350, or another 10%-12% lower.
Danely’s overall trends: lower growth, but still good growth. Semis look less like growth companies and more like ‘normal’ companies. They need to pay more dividends, lower operating expenses, and increased focus on cash management…they need to be “more Philip Morris and less Google.” The highest yield in chips: Microchip, at 3.6% yield. Compare that with MO, which pays 5.1%; Pfizer pays 6.5%.
Conclusions: Worst is over. Hard to get excited. Demand lackluster. Margins/leverage not attractive. I’m concerned about demand, he says. Valuation is OK, on 2009 estimates. Inventory is not excessive, but not too low either. He sees another year of so-so returns.
Next: the outlook from Morgan Stanley’s Mark Lipacis. Where are we in the semiconductor cycle, he says, is a question he gets asked all the time. He says the question is not that relevant anymore. He thinks communications infrastructure is interesting for an 18 month time horizon.
In the past, valuation did not matter - as long as there was momentum, as long as the next set of data points was better than the last set of data points, you paid more, Lipacis says. In the past it was really a momentum group. Before 2001, if you simply paid the second derivative on revenue growth, you made money, without having to do a lot of analysis on individual companies. Bu he says over the last five years, this relationship stopped working. What is working? Fress cash yield was the best predictor of mid/large cap semi stock performance in 2006 for the first time in 9 years, “if not ever.” Why? Lipacis says it was a signal that semis are maturing. Growth 2005-2010 is in the 3%-9% range, down from 20-year average of 13%. Also seen margins stabilize. He contends one result is that the cycle call is now much less important.
What’s the next growth driver? Has to be something in the consumer/mobile area.
Today reminds us of the late ’90s, Lipacis says. Telco balance sheets are healthy. Telco pricing flattening out. In late ’90s, was Internet driver. Today, YouTube is the new solitaire. Debt levels in the telcos are at the lowest levels in 20 years. Interest coverage is improving, similar to where it was in the late ’90s. Conclusion: these guys can spend money if they are motivated.
So what is their motivation? Cable companies capex increasingly threatens the telcos. Aggregate telco and cable capital spending looks like the late ’90s. Telco price per minute from 1930 to 1995, dropped 7% a year. In late 90s, there was big buildout, and started falling 15%-20% a year. But now is flattening. Perhaps the excess capacity has been soaked up.
Now we get some reaction from the buysiders.
Niles says he looks at GDP cycles, not semi cycles. “I am incredibly concerned about the next 6-12 months,” he says. “I don’t think people really know what end demand is.” Oil price spikes, home values declining. Food inflation and inflation in general is something we have not had to deal with for a while; 10-year bond rate is going up, making credit more expensive. $168 billion of rebate checks will be gone. Next president will have to raise taxes to pay that back. I’m worried about recession late this year, early next year, as the rebates dry up, he says. The big problem is still inventory - an inventory of unsold homes, at 11 months. That’s the issue, he says. In 2001, there was ton of inventory on balance sheets; now its home inventory. Until then, we have a problem; I’m playing very defensively, he says.
Seligman’s Peruri says cycles had big deltas 5-10 years ago, when it was more rapid growth industry. Now a slower growth industry, and not as cyclical. If you look 10 years ago, if you got cycle wrong, it was hard to get investments right. As industry has slowed, cycles are more muted; you don’t need to call the cycles. More important to understand specific company fundamentals now. What you really have now is variation in investor psychology more than macro environment shifts. He agrees we will have a recession, but once stimulus money is spent, you will get slowing growth. The worst may or may not be over, I don’t know. Global growth likely muted, and semi growth also likely muted. So you can make money with companies in good product cycles. There are always companies ramping and innovating, and always some struggling and getting dislocated.
Danely notes SOXX has gone from 340-350 to 410; he gets calls from people looking to invest, and asking what to buy. Where they are now, people are conditioned to expect some upside. You need consensus estimates move up; but he does not think that will happen. “I don’t feel very good,” he says. Danely also says he has a view of caution on the economy. Not sure why the PC industry has not felt any slowdown.
Niles says we are seeing some slowing in the PC industry. Cisco orders in emerging orders are slowing. Most recent quarter at Cisco, they showed slowing in India. Some handset vendors saying no pick up in handsets in China post-earthquake. China stock market is down almost 50% peak to trough; isn’t this the bullet-proof market? China is building up to the Olympics; what happens on the other side? There are some interesting in this decoupling trade that people are playing; it may be different than people think if emerging markets slow.
Peruri says PC strength could be strong for longer than people think. PCs are the only end market prices in dollars; cell phones priced in local currencies; telecom priced in local currencies; industrial priced in local currencies; but weak dollar makes Intel-based PCs effectively cheaper around the world. The combination of CPU competition between Intel and AMD and the dollar depreciation could get to point where demand for PCs is accelerating around the world. PC demand could last longer because of the dollar.
Niles noted that Microsoft (MSFT) missed PC numbers in the latest quarter; they saw downshift in demand. They tried to blame other things; it will be interesting to see how that pans out; it was masked by massive Xbox demand.
Tai wants the panel to say what the growth driver is post-consumer. Danely says you never know what the next growth driver will be. He was an engineer who worked on DEC Vaxes. He though the IBM PCs were so slow, that they would never work; 3 years later, the Vaxes are gone, and everyone is using PCs. You don’t know what the next innovation will be.
Niles notes that there are interesting things in the alternative energy sector. He says the good news is, we have to get off dependency on oil. Gasoline is $4 here, but Europe and Japan pays $6-$8. That is probably the next big wave of growth in terms of chips. The problem is; semi industry is so big, you can’t have one sector drive it. You need something huge to drive it.
Niles cautions that betting your career on this is dangerous; everyone wants to be a solar company, and Applied Materials (AMAT) is making that possible.
Peruri thinks Nvidia (NVDA) has an interesting play in the use of graphics processor for computing, with 100x performance improvement over conventional microprocessors. They created the GPU market from scratch; makes you wonder if they can create another $10 billion to $20 billion market from scratch.
Sox predictions one year out:
Danely sees SOX 450, about 10% up from here.
Lipacis says 451. Take the 12 month price targets for the Morgan Stanley stocks in the SOX.
Niles: 452. I have no idea, he says. S&P and SOX will trade in 30% band. You could have another bank go under. (”Hope it isn’t ours,” he says, given that Lehman owns Neuberger.)
Peruri: I really feel likes things are going to drift up, then go lower. I’ll pick 425.
Tai: 488, but things get worse before they get better.
PICKS!:
Danely like Microchip (MCHP), which competes against “stiffs” like STMicro (STM) and Freescale. Tough to find semi company expanding addressable market noticeably. The other pick is Advanced Micro Devices (AMD), which is trading like it is going out of business. (Which they are, someone in the peanut gallery just grumbled.) Gone from grabbing market share, to retrenching and restructuring. Could be profitable by the end of the year. And could double.
Lipacis likes PMC-Sierra (PMCS), the class pure-play on communications infrastructure. Once a $250 stock, now at $8. The best performing stock in 2003. I believe if infrastructure theme is right, will see people remembering good old days. The second name is Texas Instruments (TXN), in part on sentiment call. People hate that stock. When everyone runs in one direction, you might have an idea. Nokia shifting away from them; for 8 months people have been obsessing about it.
Niles: sticking with Skyworks (SWKS). Gaining share at Nokia, RIMM, LG, Samsung, etc. Plus growth in linear business. The other long, a solar play, is Applied Materials (AMAT). Orders are down 60% year over year; semi business for them will be at some point will turn. (He owns SWKS, but not AMAT, yet.) Taiwan Semi is raising prices, sign of tight capacity. And real upside is the solar business. Start recognizing orders from solar business late this year. Applied will supply a lot of the survivors; make it easy for anyone who wants to be in the solar business to get in. And a short: the TLTs - expect rates to go much higher. Ticker is TLT, which is an ETF.
Peruri: A theme since the bubble is every company wants to invest in analog space, especially high end analog. I wonder if digital space, which everyone is shying away from, is now the place to go, and wondering if commodity analog is also place to go, since people are de-emphasizing them. Marvell (MRVL) is the digital pick; lot of great product cycles. Street has them at $1 for next year, but could do $1.20-$1.50. With 20x multiple, then $24-$30 stock, at $17 today. About 70%-75% off the bottom; a year from now it could be meaningfully higher. The second is ON Semiconductor (ONNN), which is focused on lower-end analog. And can buy assets on the cheap. They could earn $1.20-$1.50 a share as well; the stock is around $9; if you use a 10x-15x multiple, you could get 50%-100% upside. The stock is however up 100% off the bottom.
Tai: Also picked Marvell, for different reasons. They are in good segments; the drive market, WiFi space, high volume areas that are growing. The company historically executes very well.
Will we see wholesale consolidation? Danely says we should in theory, but that we aren’t likely to actually see it. Maybe some smaller analog companies getting bought. Power management is highest growth end market in analog; Analog Devices or Texas Instruments would like to get in there in a big way, he ways. Lipacis thinks consolidation is possible in the memory sector.
Niles says he can make a case that this is 2001 again. There were 6 40% rallies in the semi index on the way to losing 80% of your money in the index. This is the analog to being in the end of 2001, when the stocks dropped another 40% from there. Niles notes that investment banks are running at 25-30x leverage. Between that and the housing market, that has me kind of concerned, Niles says. The president who comes in to deal with this may be good on the economy, and protectionism would make it worse.
Someone is asking about the cost of raw silicon. Peruri notes that the market is being driven by the solar sector. The bear case, he notes, is massive over supply is coming in polysilicon. The bull case is that the cost of energy is rising, and silicon solar panel costs coming down. If reach grid parity in 2-5 years, demand could go up astronomically. About 0.1% of world power from solar. Demand could go up 10,000x. So prices could go higher. Premature to assume polysilicon prices will go higher or lower.
And that’s it.
Comments