Jump to content

All Activity

This stream auto-updates     

  1. Yesterday
  2. sunnypease

    First Solar (FSLR)

    CS upgrading FSLR to neutral. ■ Section 301 Fears Overblown: Solar manufacturers and developers, including FSLR, reacted negatively on Friday 6/15 (down 3-7%) owing to concerns that the fresh round of 25% tariffs on Chinese imports (under Section 301) will raise module prices for developers and squeeze margins for manufacturers. Solar cells and modules are listed in USTR's new list under subheading 8541.10.00 ("Diodes for semiconductor devices, other than light- emitting diodes"). However, we believe the negative reaction is overblown for the space, as c-si manufacturers are supplying solar products to the US market from their Southeast Asia factories, which were set up in prior years to circumvent the ~30% Antidumping (AD) and Countervailing duties (CVD) imposed by the US on Chinese imports after 2014. ■ ..for FSLR as well: The negative reaction for FSLR is even more overblown, as the company doesn't have any Chinese factories (they are in Ohio, US and Southeast Asia) and instead stands to benefit from tariffs on competing products. That said, we still embed $2.60/sh for pricing benefits that resulted from separate Section 201 solar tariffs that were announced in January; this value would be at risk should the US soften or eliminate these tariffs as a negotiating strategy. See our 4/9 Trade War Impacts–Renewables and Utilities are Both Insulated and 6/6 FSLR: Near Term Protected, Long Term Still Needs Proving. ■ Valuation: Our $53 TP remains unchanged and is based on a multi-stage valuation, including DCF through 2022 and P/E and EV/EBITDA for core business thereafter. We see potential downside of $2.6/sh should Section 201 tariffs be eliminated and another $3/sh if firm-price bookings face cancelations. Our Grey sky value of $35/sh gives no value to growth from Series-6 deployment. Key risks include Series 6 production ramp, cost reduction, oversupply, and changes in US trade policy.
  3. Mark

    First Solar (FSLR)

    Ha, yeah... this is something that's ticked me off since I started with solar. We aren't considered energy (unless energy is selling off, then we are) and we aren't considered tech or this would be Tesla-like. Just a commodity I suppose. May as well be trading soybeans. Here's hoping we get some follow through tomorrow. I really need to unload and was hoping to do so much closer to 60. Biggest trading snafu of my life is what this decline has been. Set your stops, I know, set your stops. Burned. Bad.
  4. Last week
  5. sunnypease

    First Solar (FSLR)

    it may have been the quad witching. i don't fully understand how options work, but i understand it requires stocks to be bought and sold. However, options do not expire until the next day, so this must be brokers & MMs transacting ahead of expiration to settle up accounts somehow? I messed up buying calls Friday and wound up with 35 calls expiring in the money. Oops. So this morning I had 3500 extra shares. Yikes! Somehow my margin allowed this but I was trembling... PLEASE GO UP. The solar gods listened and sent FSLR up 50 cents or so and I sold those shares like a hot potato. Wound up making a little something on the deal too. Then later we broke over 52 and something.. wish I'd held on, but could not afford to be long that many shares if it dropped. Anyhow... its good we are turning positive again. Hope we can stay that way for a little while. rest of market seems to be acting giddy and crazy, so hopefully some of that bleeds over. FSLR is a tech company too!
  6. sunnypease

    Canadian Solar (CSIQ)

    Thank you SCSolar for the further detail & clarifications. It seems that these days the BOS costs are greater than the module cost & so I would hope that could be reduced. Perhaps by automating installation, especially in the US and Europe (higher cost of labor).
  7. Mark

    First Solar (FSLR)

    Ha! It has nothing to do with all the downgrades, including the Credit Suisse one which was already bearish to begin with. Or the reason for the panic what with the whole sky is falling along with ASPs thing. Nah, quad witching maybe.
  8. sunnypease

    First Solar (FSLR)

    Wow. Don't knock yourself out. Here is the take from Credit Suisse on Friday's action: ■ Our take – Renewables: Solar manufacturers were down 5-10% on Friday morning possibly due to quad witching, market sell off, and concerns that the US would impose additional 25% duty on an expanded list of Chinese imports (which might include solar products). We believe the reaction is overblown as the oversupplied solar sector has enough non-China capacity to supply to the US market if required. We have not yet seen signs of consolidation upstream, but downstream micro-inverters have taken the first step. Declining cost of solar and storage continues to exceed expectations, while cities and states ramp renewable demand.
  9. SCSolar

    Canadian Solar (CSIQ)

    I would view a worst case scenario of companies being profitable with an ASP above of $0.175+$0.035=$0.21 I would view a better case scenario of companies being profitable with an ASP above $0.15+$0.03 = $0.18 I would view a best case scenarion of companies being profitable with an ASP above $0.13+$0.025 = $0.0155 I am using USD as a reference. Currency fluctuations would change the results. In my example I use a cost savings in areas that detail a drop in manufacturing of modules to atleast $0.175/watt. If you use a slower Mores Law adjusted 25% cost reduction per doubling of capacity for the next decade, Then a $0.24-$0.28 cost to manufacture today drops to $0.15 to $0.175 cost to manufacture. I am saying the ASP to be profitable is Cost to manufacture + Opex + Interest + Profit I am saying today Opex is at $0.04/watt including shipping look at the JKS and JASO ERS. I expect it to drop 25% or more per doubling of capacity. A 10% shipment growth per year places shipments aproaching 300GW. A 25% lower Sales cost(including shipping) on efficiency gains of 25%(aka 23-25% efficient modules) is $0.007 savings from todays costs of $0.027(JKS Q4). A 25% lower Admin cost per watt per doubling of shipments places General Admin around $0.0045 from todays $0.007/watt. A 25% lower RnD cost per watt per doubling of shipments places RnD at $0.00329 from todays $0.0516/watt Based on the market growth forcast at 10%+ per year for the next 10 years and savings at 25% per doubling of shipments Opex=$0.02+$0.0045+$0.00329 = ~$0.0278 or $0.03 or less. Yes. Depending on module manufacturer and volumes Interest Net Income runs $0.0035-$0.01/watt. This drops to $0.0027 to $0.0068/watt with a 2.5 times increase in shipments at a 25% reduction per doubling. I expect then in 10 years that the Opex and+ Interest will be under $0.04/watt and in the range of $0.03-$0.035. down from todays $0.045-$0.05. It could be lower.
  10. explo

    Canadian Solar (CSIQ)

    Thanks SC for all details, especially clarifying the Moore's law parallel. Thanks also for clarifying that OPEX (SG&A+R&D) reductions are not lagging COGS reductions as much as I indicated. Anyway one issue might be that both volume expansion and technology progress might have driven recent cost reductions. With a demand saddle point coming we might get a cost progress saddle point too before they both take off again. As disdaniel says there could be demand explosion at some point. I think the panel makers have done their job to make PV competitive. I'm certain that the next decade will be a glorious one for PV. The decade that we will hit peak install.
  11. explo

    Trading Strategy

    Yeah, the data is getting a bit big. Borderline that I need to manage it like that.
  12. disdaniel

    First Solar (FSLR)

    My OpenInvest portfolio (~$5,500 in S&P 500 type firms--but minus all fossil fuel, large mining, tobacco, chemical, utility and big banks) just purchased 1 share of FSLR on Friday. Normally solar stocks plummet 25-75% after I first purchase . Fair warning!
  13. disdaniel

    Canadian Solar (CSIQ)

    At some point along this cost curve demand explodes. By "explode" I mean demand curve goes non-linear compared to history. Already solar is eating nat-gas peaker and coal's lunch. Knock another 20-25% off the cost of solar (lowering total installed cost by 10%)--as china is now doing--and you eat nuclear, and regular nat gas's lunch too. I see storage as the final key to balancing the supply/demand equation in solar's favor. I think the cost of storage will drop by 50% (maybe 65%) over the coming decade. We may be within a decade of the last fossil fuel plant being built. All that said, the next 6 months could be very painful for module makers.
  14. sunnypease

    Trading Strategy

    Get your sheet together man! You could use hadoop or the cloud?
  15. sunnypease

    Canadian Solar (CSIQ)

    Thank you very much SCSolar & explo. SCSolar.. that was a really great future expense analysis breakdown. Just to verify, you think that 17.5c / Watt is the minimum price you see prices going to in about 10 years? Also.. because the supply chain is in China, these expenses are not really 17.5c / Watt, but are the equiv. RMB, correct? Then you say Opex, including shipping will be 0.04 / Watt more? And then you mention interest. Do you mean interest expense from LT debt? Thank you
  16. sunnypease

    First Solar (FSLR)

    I've realized yet another possible gremlin... First Solar has recently had some big projects in Turkey. As many are aware the Turkish Lira is down about 30-40% and had not yet started to recover. Meanwhile the fed tightening will cause further stress to other ROW markets. Thoughts anyone? Klothilde.. have you thought this point over?
  17. SCSolar

    Canadian Solar (CSIQ)

    My point was there are technologies that are available that can have double the efficiency of todays mass production technology. Only time is the impediment to the costs being driven down for these. Those efficiency improvements along with other technology changes and mass of scale will drive down what is perceived as a stagnant portion of Opex. There is also another 40-50% improvement to be gotten from todays crystalin productions. Moores Law often cited was a single decade forcast and has been revised several times. The comment at the time was basically the packing of transistor density would double every year per surface area over ten years when it was made. It held true for a Decade then was modified. The modification called for a doubling of density every 2 years. According to Wikepedia Dave House(whom I played golf with and worked for) indicated that he thought it was 18 months after the revision. Now Moores Law is sitting at 3 years for doubling density with the eventuality of no more gains. As miniaturization occurs, the ability of miniaturization slows due to technological and physical limitations. In essence what the modifications look like is the slew rate charging cycle of a capacitor(SRCCC). I will agree that Solar has finite limits for single wavelengths. That limit has always been around 30%. In the past decade the efficiency went from 12% to 18% for Multi and 15% to 21% for Mono. That is a roughly 50% improvement. It may take another decade to get to from todays max efficiency in production of 21-24. That would be a 75% improvement in raw efficiency gains. Or a 20% improvement in overall efficency. I might speculate that it takes another 10 yers to get to 27-29% efficiency after that which is around a 50% gain in raw efficiency vs the prior 10 years or a 7.5% overall efficiency improvement. That would somewhat track the revised Moores law and the what SRCCC If you look at the cost to manufacture the past decade, modules were around $2.80 when companies were producing 200MW a quarter. Todays costs are $0.30+/- at volumes of 2GW. That is a 10 fold drop in cost on a 10 fold increase. That is a 50% price drop per doubling of production trend. I would expect at least a 25% decrease in production costs per doubling of capacity going forward for the next decade. That would be inline with your Moores law pattern. The Operational expenses have tracked similar in that for every doubling of shipments. Opex was cut appx 50%. While 50% decrease in cost for doubling of shipments may not continue, I would expect at least a 25% decrease in production costs per doubling of capacity going forward. I would also expect the Opex to follow a similar SRCCC curve. For module manufacturers opex and interest is running at less than $0.05 depending on company for module manufacturer. $0.023-$0.025 for shipping(probably will rise with lower CN shipments) $0.08 GM, and $0.05 RnD. The Opex would fall from just under $0.04 to around $0.03 or less. Where do these numbers come from? Poly production costs used to be $20/kg + a decade ago. Now in new plants with favorable electricity prices it is well under $10/Kg and heading toward $5. Grams per watt with technologies keeps falling. Cells a decade ago were 210microns and fell to 180 160 and in some instances 140microns. With newer wire saw technology the thickness of the wafer falls to 100-120 microns. The amount of poly wasted has fallen from what was 200 microns to 140 microns to what is now 80 microns with diamond wires. You are looking at 180 Microns in the future vs 300 Microns today. Most of the Kerf loss is now recoverable and recyclable. This leads to a more than doubling of wafers from a similar amount of silicon consumed in the process. The manufacturing doping texturing and coating improvements has also lead to a 50% improvement from a decade ago. All this has lead to grams per watt falling from over 8 grams in the early 2000's to well under 4 grams/Watt today. 4 grams at an eventual $6/KG cost is only $0.024 cents vs todays $0.07. Wafer processing drops by 50% from todays $0.06 to $0.03. That is a $0.054 reduction in cost for wafering in the future alone. Cell processing may drop from the $0.06 range today to $0.04 as some of the new technology reduces but adds costs as well. Module processing falls from $0.10 today to $0.07 or less as efficiency gains and material costs decline from mass of scale. This all would lead to cost falling to $0.024+0.04+0.04+0.07 = $0.175 in the next 10 years at a minimum. As for Operational expenses, certain items are fixed such as shipping. The only variance that will come with that cost is the thickness of the panels and the efficiency and regionality of sales. Thinner wafers, thinner stronger and more rigid glass can help reduce that cost as can increased efficiency in panels which can lead to another 25% drop in shipping costs. Those costs used to be $0.05/watt 8 years ago and has fallen to 0.025 or less as of recent ER's. This should fall another 25% + with efficiency gains from 19% to 24% and thinner wafers and glass and diversification of the base where items are shipped. What is clear is that the rest of the operational expenses are running at $0.015 today. That is the pretty much fixed costs of General Admin and what has been a semi fixed RnD value depending on company. These cost in sum have basically over the past 10 years been a 50% reduction per doubling of capacity. The Opex is running for pure manufacturers in the $0.04 range. As the Capacity increases, the historical tend would be for these costs to continue a rate of decline of atleast 25% if not pushing 50% as history has shown. Similar trend is interest per Watt shipped. This should lead in the next decade an interest and cost per watt that runs at $0.025-$0.035 or less in the next decade. I see no problems with in the next decade or so for the cost to manufacture falling up to 50% from todays $0.28 for most to and with expenses being profitable at an ASP of $0.18 and above. In fact one of the driving factors in China's policy is to take the waste out of the supply chain near term as artificially inflated prices are supported by their past Tariffs. They want target projects at $0.60 and modules at $0.25 near term. That is down from the $0.70-$0.80/watt today in China. Taking a byte out of Poly is the first start, the second is the soft costs in building projects and then other upstream supply chain supply cost imrpvements.
  18. explo

    Trading Strategy

    I'm starting to get my sheet together. I think I need to buy a more powerful PC again.
  19. explo

    Canadian Solar (CSIQ)

    Yes, but not losing the photons hitting your single junction cells has been the cheapest way to grow conversion efficiency for now. And that physical limit is way below the mathematical 100% retention (the wave lengths convertible by a doped silicon crystal is only around one third of the wave lengths the sun shower the earth with, increasable, to a limit, with the much more complicated multi junction). Even the mathematical limit of the number of photons normally hitting a surface could be morphed a bit with lenses and mirrors, but I think the point is that before when a lot of photons were lost before conversion (and some electrons "lost" post conversion) there was much growth potential in the conversion efficiency improvement from these loss reductions. It has scaled like Moore's law for semi for now (but with at different factor) due to the similarities (improve tapping of small none moving structures, no mechanical parts). However unlike semi that are designed to convert electrons to calculations there are ceilings for the conversion of photons to electrons. In the former case there is no lack of electrons you just have to cool them down. In the latter case there are only so many photons to capture and therefore your loss reduction grow quickest in the beginning when you lose the great majority of them. Note that the conversion efficiency is only one part of the heavy cost reductions we've seen in panel production cost, but the part that is share with the whole chain, including the freight costs. Much of the PV version of Moore's law is the factory equipment improvement. CVD reactors in polysilicon plants that produce much more poly kg per electricity consumed for each generation. Ingot casting furnace that cast bigger blocks (more kg per electricity) with each generation. Wafering equipment that can slice thinner wafers (more surface per kg) without risking them break in cell processing. Thinner wires to reduce kerf loss (more kg retained). 6 inch cells instead of 5 inch (more surface output per cell line). 72 cell panels instead of 60 cell panels (less panels assembled, shipped, installed). Some of these have reach some limits too. Too big panels (cheap to produce) will have costlier human handling etc. I'm a strong believer that solar PV will dominate electricity generation as it is still going to get cheaper than any other electricity generation source if you calculate all costs (the only eagle couple in the region killed by wind turbine, ecosystem deteriorates, crops fail, etc.), but the panel production cost problem is already solved. No way that I could hope a panel would cost $0.30 already now 8 years ago. It's the cost of getting the panel from the Chinese factory to the rooftop (the rooftop space is basically free as is distribution of the electrical current to the close by consumption place) that will become the challenge now. The rate of these improvements will likely saturate a bit. I think this is the problem for panel producers. They can cover the cost to produce a panel than is sold at $0.30, but they have a harder time covering the expenses (sales, freight, admin, r&d) associated with the massive volumes and research it takes to achieve such low production cost. Example. If ASP goes to $0.25 and if 20% GM is achieved through a $0.20 cost then close to half of $0.05 GP per watt might go just to the freight of that panel (especially if we include trade barrier circumvention expenses), not to mention expenses to get it sold, expenses to handle a supply-chain for a massive scale of production that due to political reasons globally or in China cannot be too centralized. The point is that in 2010 there was 25% GM on $1.60 ASP or $0.40 GP per watt to cover all the expenses not related to production costs. We know the production cost was effectively cut from $1.20 to $0.20, no problem. But to cover OPEX with $0.05 instead of $0.40 when those costs are things you cannot affect (rising salaries of China sales personnel etc.) and reduce with scale in the same way as your production cost, then you are in a more tight spot when a major event disrupts the established cost sharing balance along the chain.
  20. SCSolar

    Canadian Solar (CSIQ)

    Solar Technology has a long way to go with efficiency gains from todays standard modules. There is still room to double cell efficiencies form todays single junction. NREL has shown over 46% efficiency years ago using Multi junction technology that utilizes more than a single wave length. The issue is reducing manufacturing costs and yields in a high production level enviroment to be cost effective. https://web.stanford.edu/group/mcgehee/presentations/McGehee2011.pdf https://www.nrel.gov/pv/high-concentration-iii-v-multijunction-solar-cells.html
  21. sunnypease

    First Solar (FSLR)

    Then there is the CdTe express lane. Direct from Ohio. The biggest swing state. Or from Vietnam (who doesn't love Pho.. and it's not China!) This is a company that is officially expanding at this point. It's an American "success story". If Trump gets too much heat on tariffs, First Solar is the American worker he can put his arm around. Plus, did we mention they have 2 years worth of production signed up for. So... FSLR's US future looks great. Meanwhile their Europe/India/ROW/China story looks worse. Lower Si module prices will mean very tough competition in those markets, correct? This, to me, is the biggest cause for concern in FSLR's story going forward... thoughts?
  22. sunnypease

    Canadian Solar (CSIQ)

    The problem is that lower panel prices will arrive in lower PPA prices, correct? And will CSIQ be able to drop their cost structure to meet the sudden drop in demand? One danger I can see is that panel price is a smaller part vs BOS. If panel price decreases by 20%, the overall system may only decrease by 10% or less. This might not be enough to boost demand. Panel price might be forced even lower to spur demand. This is only my fearful, worst case thinking.. Have there been studies of how demand might be effected with various changes to panel price? Explo has a good point in that the science of improving efficiency / driving down costs is hitting the walls of physics. The problem is that policy makers are not scientists. They expect things that cannot happen. For example, cars that can go 70 miles per gallon. Or cars that produce a certain level of diesel particulates.
  23. explo

    Canadian Solar (CSIQ)

    Spain was the first boom/bust market. The bust was so severe it died. Add to it a shortly following general construction boom/bust in the country. I think Germany and Italy are healthier examples of life after bust. UK as you mentioned picked up some slack within EU after German and Italy. They were hit quite bad by the tariffs. Maybe Brexit will make them a PV haven.
  24. sunnypease

    Canadian Solar (CSIQ)

    Shame. I just drove across Spain, from the south near Granada up to the north near Santandar. I saw only a smattering few solar panel installations & they looked old. There is so much empty land there and so much sun. UK has manage 26% of grid power supplied by solar. Why can't the rest of Europe do this?
  25. explo

    Canadian Solar (CSIQ)

    I agree. If I recall correctly a panel not assembled in China can use 29 Chinese cells and 31 Taiwanese or Korean cells to be tariff free in EU at any price. This is what Chinese companies decided to exploit and to "simplify" that they cancelled their MIP deals. EU is already open to them. EU is however past peak installation growth.
  26. SCSolar

    Canadian Solar (CSIQ)

    Most every country that has looked at solar and anti dumping has ruled that China was dumping and applied tariffs. Do you view India as any different? While China was selling in China in the $0.38 range, they were selling in India for sub $0.30. What is your perception and removal of the MIP in Europe? My understanding is most of the Chinese companies had withdrawn from the MIP and were using oversees manufacturing to supply Europe. The current MIP is $0.37 for Multi and $0.43 for mono. Those MIPS are dropping to $0.35 and $0.40. come July 1. I presume that the Chinese companies using oversees manufacturing are selling below those MIPS which would be inline with market rates in the low $0.30s for Multi and Upper $0.30s for mono.
  27. explo

    First Solar (FSLR)

    Think of it as a global pile up of panels in the distribution channels (including production sites, market central warehouses, close to consumer small stocks, in week long transit on ocean to maybe no longer optimal destination). The non-movement of products will cost money. Products will be moved at a loss if it prevents a higher loss of not moving them. Markets that can move products to US at low cost will move products there and the products those market need will be moved from piled up markets that have higher cost of moving their products to US. The sum of it all is that movement costs have gone up (US are blocking the existing highways) and demand has gone down (China is feeling satisfied and want to digest for a while). All will suffer from it, some more than others.
  1. Load more activity


×