tag:blogger.com,1999:blog-4815867514277794362.post3127232843792740521..comments2024-03-08T06:18:28.125+11:00Comments on Bronte Capital: Kick-back on First SolarJohn Hemptonhttp://www.blogger.com/profile/03766274392122783128noreply@blogger.comBlogger34125tag:blogger.com,1999:blog-4815867514277794362.post-31564776467451926612012-07-21T17:06:37.149+10:002012-07-21T17:06:37.149+10:00hey. thanks for sharing such a detailed review wit...hey. thanks for sharing such a detailed review with facts and figure. they were very useful . do keep on updating the post. will be here looking up for updates.<br /><br /><a href="http://www.thepvsource.com/index.php/inverters.html" rel="nofollow">SMA Inverters</a>Anonymoushttps://www.blogger.com/profile/14207908020496744478noreply@blogger.comtag:blogger.com,1999:blog-4815867514277794362.post-15554195525099319302010-05-16T18:19:22.078+10:002010-05-16T18:19:22.078+10:00The PolySirice would be very difficult to get belo...The PolySirice would be very difficult to get below USD$30 a kilo as most producers cannot produce it at USD$30 a kilo. You will get the big cashed up players forcing the little players out and the price of PolySi will increase again. <br /><br />Better advantages are to be made in improvements in the amount of Si required per Watt and increased efficiencies. Places where land is expensive will look to more efficient PolySi based modules to generate more electricity per acre. <br /><br />Thin film will be attractive where land cost is not an issue so efficiencies need not be as high - eg the desert.Craignoreply@blogger.comtag:blogger.com,1999:blog-4815867514277794362.post-89298237617205037072010-04-30T00:06:46.985+10:002010-04-30T00:06:46.985+10:00You were both right and wrong with your thesis at ...You were both right and wrong with your thesis at least based on Q1 earnings.As I said in my first comment ASP is just not a function of lower efficiency compared to c-Si modules,it is also bankability as FSLR improved its ASP in Q1,while other companies ASP's are decreasing<br /><br />However regarding costs they disappointed me a lot as it was flat with last quarter and efficiency has almost stopped increasing despite higher throughput<br /><br />I would short FSLR but would have a long position in a strong c-Si stock to negate the strong sector demandAbhishekhttp://greenworldinvestor.comnoreply@blogger.comtag:blogger.com,1999:blog-4815867514277794362.post-89052569250424741552010-04-29T22:42:41.633+10:002010-04-29T22:42:41.633+10:00Okay maybe I wasn't being too cute getting out...Okay maybe I wasn't being too cute getting out of the FSLR short. I think Q1 is interesting. It shows insane demand (priced up q/q! doh!) and yet their cost structure (core cost per Watt) and efficiency are flatlined. I'll still by my previous prediction. Place your shorts after the Q2 earnings call.C. Fuzzbanghttps://www.blogger.com/profile/01763956615872125922noreply@blogger.comtag:blogger.com,1999:blog-4815867514277794362.post-34211758923631270012010-04-26T17:45:35.463+10:002010-04-26T17:45:35.463+10:00Citi out with bullish poly commentary on 10.4GW de...Citi out with bullish poly commentary on 10.4GW demand 10e.<br />Talking about stabilised $50 poly prices with some rising spot prices (for what that's worth)...Nickhttps://www.blogger.com/profile/01164675502109841117noreply@blogger.comtag:blogger.com,1999:blog-4815867514277794362.post-25442273500553506872010-04-20T06:55:31.052+10:002010-04-20T06:55:31.052+10:00Because I enjoy this site, and maybe it helps, her...Because I enjoy this site, and maybe it helps, here is my quick primer (materials scientist perspective). If you know/remember a fair bit (and like equations), the wikipedia solar cell page is very good.<br /><br />http://en.wikipedia.org/wiki/Solar_cell<br /><br />--------<br /><br />A quick note on band gaps, simplified grossly <br /><br />The band gap is a measurement of the energy separating the conduction band electons (the metallic-type ones that can conduct electricity) from the insulative/valence bonding electrons.<br /><br />In Si, the photon enters into the material, and if the energy and momentum is right, it pushes the electron from the valence band into the conduction band. That free electron then can conduct away.<br /><br />If the system were completely symmetric, eventually, the atom would lose enegy, emit a photon (or phonons), and drop back into the valence band.<br /><br />The trick of a photoelectric cell, then, is to make lots of transitions happen, and to make an assymetric barrier--so electrons can go one way, but not back. Thus, you sweep up the electrons going one 'way'. <br /><br />Efficiency is the work in making sure that<br />1) you generate lots of electrons from the photons.<br />--this works well if the band is matched (so the band is less than the energy of a photon). Bands are moved using dopants. <br />--Si requires a bit more complicated transition, so is always less efficient than 'direct band' semiconductors. You need 2 things to happen to get silicon to give up the electron. having 2 things happen is always less likely than 1 thing.<br /><br />2) you collect all the generated electrons efficiently<br />--have a good oxide<br />--dope effectively.<br /><br />3) the conduction electrons aren't lost or converted accidentally.<br />--keep heat down<br />--have high conductivity and limited resistive losses<br />--keep defectivity down. Local recombination centers, where electrons are locally recombined due to material defects, are limiting.<br />--This is what manufacturing efficiency really gets you: defectivity, in a manner very similar to semiconductor.<br /><br />That is a short version. Manufacturing efficiency improvements are to<br />a) improve the conductivity (contact losses, contact resistance, doping)<br />b) Reduce defects that contribute to recombination losses<br />c) Improve optical absorption efficiency.<br /><br />Defects tend to follow exponential decats, optical absorption doesn't change quickly, and conductivity is mostly a cost question. So, that can provide you with a useful model. Max efficiency is then a property of the material system.Unknownhttps://www.blogger.com/profile/04773368635111455556noreply@blogger.comtag:blogger.com,1999:blog-4815867514277794362.post-35240725554423588592010-04-20T04:06:09.363+10:002010-04-20T04:06:09.363+10:00By no means is the approval for First Solar's ...By no means is the approval for First Solar's Topaz project in San Luis Obispo County a given. That 550MW plant will require CEQA-compliant environmental analysis and there are a host of issues there - from wildlife impacts to noise, traffic, loss of ag land, water and many other impacts and risks. The site is poorly chosen near a rural community and agricultural economy, and there will be a significant cost of bringing the power to market because the transmission lines will have to be upgraded.<br /><br />First Solar is working on technology to replace Cad-tel. That means it will have a backlog of Cad-tel panels to get rid of. As a commenter said above, you cant create fake demand through projects.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-4815867514277794362.post-31708812810669709962010-04-20T00:52:27.012+10:002010-04-20T00:52:27.012+10:00Hi Jon. I posted this on GTM and edited here for y...Hi Jon. I posted this on GTM and edited here for you. I liked your reply/post on your blog. <br /><br />The writer brings up some great points, and the readers bring up some great points. My guess is some of the comments have hardened his conviction and some have made him question his position. Here are some thoughts (from an investor who has been long and short solar companies.)<br /><br />1. What would the cost of CSi be if the solar grade silicon were free? I get about $1.20 a watt, but I would love to hear other investor’s thoughts. [I have never gotten as low as the .80 Yingli is stating. One other development here would be the use of metalurgical grade silicon in place of solar grade. Timminico and Globe Metals were trying.]<br /><br />2. What is the value of the glass in a FSLR panel? I view this as the lower bound of their pricing structure?<br /><br />3. The writer rightly points out this is a commodity. However, in commodity companies, I think the cycle is more important than the player. Cycle and price trump management every time. Even a crappy paper company goes up when the price of paper goes up. Where are we in the cycle? Could a Chinese or US FiT change the cycle? What if electricity prices went up (i.e. natural gas) or solar dropped below grid parity? Those would cause a surge in demand? Would that change your thesis?<br /><br />4. The writer might be attibuting too much emphasis on the space necessary of panels. The California Solar Initiative, probably a decent barometer for the world, shows that 90% of panels go to 10% of projects. Roof space is not an issue for large scale systems. Just a thought.<br /><br />http://www.energy.ca.gov/2009publications/CPUC-1000-2009-033/CPUC-1000-2009-033.PDF<br /><br />5. All the posters agree that FSLR is a very talented manufacturing company. This is a core stregth. What if FSLR put a magic CIGS process where their magic CadTel process sits? Surely they are trying. If we woke up and FSLR articulated a path to 16% efficiency and a 35% drop in cost per watt, would that get them to below grid parity and usher in a new era of solar demand? <br /><br />Just some thoughts. I just know we are not at the top of the cycle right now and I know that every politician loves solar. THe politics are almost impossible to predict. <br /><br />And finally, is there a way to play the MW cycle without the technology/panel risk? <br /><br />I have been long SATC as a way to play the MW growth, but you will easily see that if the cost for inverters drops.... so will SATC. So far it has been steady for about 3 years. If the cost of small scale inverters drops, SMA (S92.de) will implode. Just look at what an SMA inverter looks like, kind of like a DVD player, and look at the hundreds of inverter companies. Maybe we should be looking at this as a short? <br /><br />Good luck.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-4815867514277794362.post-28143276449067342072010-04-19T05:27:36.447+10:002010-04-19T05:27:36.447+10:00Hi John,
Sorry, that was a bit harsh. You are co...Hi John,<br /><br />Sorry, that was a bit harsh. You are completely correct, it is all about photons dislodging electrons. Photovoltaics is a fairly low energy process and modern physics is pretty much divided up by "energy involved". Particle physicist tends to mean the physics of sub-atomic particles, quarks and gluons etc. <br /><br />My point, somewhat nerdily made, was simply that solar power is a lower energy process and is the domain of molecular rather than sub-atomic experts. Apologies for the fine distinction, please email me directly.<br /><br />regards,<br />KeiranKeiranhttps://www.blogger.com/profile/17911141894010393966noreply@blogger.comtag:blogger.com,1999:blog-4815867514277794362.post-56716909028601977852010-04-19T00:49:19.335+10:002010-04-19T00:49:19.335+10:00As Keiran mentions, innovation can improve in fits...As Keiran mentions, innovation can improve in fits and starts. FSLR's huge cost advantage meant rapid production growth was the way to maximize profits. They put efficiency work on the back burner. With margins finally falling I imagine efficiency is now job #1.<br /><br />As for CIGS, I think production would be years but away a stock-lifting PR blitz could come at any time.doggydogworldnoreply@blogger.comtag:blogger.com,1999:blog-4815867514277794362.post-7150245640114263032010-04-18T21:08:59.682+10:002010-04-18T21:08:59.682+10:00Keiren - more to the point - this is a particle ph...Keiren - more to the point - this is a particle physics problem (although a low energy one).<br /><br />The photons come with an amount of energy which makes an electron jump a shell in the dopant atom.<br /><br />If the energy level is right it can jump through the gate of the semiconduct.<br /><br />CdTe semiconducts are good because the band-gaps match roughly the energy levels of sunlight. <br /><br />I am trying to educate myself about the band-gaps - but I am suffering. My business partner is an electrical engineer - so I guess he could do it for me - but it was fourth year electrical engineering two decades ago - so a pure physicist would help a great deal.<br /><br />Please contact me. <br /><br />JohnJohn Hemptonhttps://www.blogger.com/profile/03766274392122783128noreply@blogger.comtag:blogger.com,1999:blog-4815867514277794362.post-39652399766531644962010-04-18T09:20:21.643+10:002010-04-18T09:20:21.643+10:00Keiren -
If you can solve (a) a thermal efficien...Keiren - <br /><br />If you can solve (a) a thermal efficiency problem for me and (b) a fluid dynamics problem I would be appreicative.<br /><br />Wall Street physicists are rare.<br /><br />Give me an email.<br /><br />JJohn Hemptonhttps://www.blogger.com/profile/03766274392122783128noreply@blogger.comtag:blogger.com,1999:blog-4815867514277794362.post-4787446239258961532010-04-18T08:36:56.562+10:002010-04-18T08:36:56.562+10:00Hi John,
I really like your notion of how to be a...Hi John,<br /><br />I really like your notion of how to be a well pay blogger. Here is an interesting presentation at Stanford that you may find interesting.<br /><br />http://ecorner.stanford.edu/authorMaterialInfo.html?mid=2425狂猪https://www.blogger.com/profile/16599529315620633684noreply@blogger.comtag:blogger.com,1999:blog-4815867514277794362.post-69851171027571335362010-04-18T06:35:34.302+10:002010-04-18T06:35:34.302+10:00Kieren
I am not that stupid.
It is photons (mass...Kieren<br /><br />I am not that stupid.<br /><br />It is photons (massless, low energy) kicking out electrons from shells. <br /><br />At least that is my understanding of the electro magnetic force. <br /><br />And they need to kick the elctron out of the shell and it needs to find its way through a gate. <br /><br />JJohn Hemptonhttps://www.blogger.com/profile/03766274392122783128noreply@blogger.comtag:blogger.com,1999:blog-4815867514277794362.post-61024038002869089262010-04-17T20:41:50.565+10:002010-04-17T20:41:50.565+10:00Hi John,
As a molecular physicist turned banker I...Hi John,<br /><br />As a molecular physicist turned banker I cringed a little at your time series of conversion efficiency. First, though,<br />solar electricity is not a particle physicist's domain. I can put you in touch with some mates at Fermilab if you want but semiconductors in sunshine is a slightly lower energy scale from smashing protons into anti-protons at the speed of light. <br /><br />Second, the one thing you know for sure is that technology innovation proceeds in fits and starts. Engineering process improvements are often incremental which I think is what you were quoting, but a small change in thin film composition or structure could easily lead to a large jump in conversion efficiency. <br /><br />If your short on First Solar comes down to betting against sharp increases in conversion efficiency then you're playing the reverse of a VC strategy. These are classically long odds high payoff. Who knows, you might win but the expected returns are slim because of the skewed distribution of outcomes.<br /><br />great blog btw<br />KeiranKeiranhttps://www.blogger.com/profile/17911141894010393966noreply@blogger.comtag:blogger.com,1999:blog-4815867514277794362.post-27817067716726761222010-04-17T04:30:51.376+10:002010-04-17T04:30:51.376+10:00John,
I'm a thin film expert in the US and, a...John,<br /><br />I'm a thin film expert in the US and, as much as it pains me, I agree with your assertions. In my view, even if the polysilicon price bottoms out in the $30-40/kg range, the g/W figure will continue to decrease as new and innovative ways to efficiently cut thinner wafers are developed. This will certainly place significant pressure on FSLR over the next 5-10 years.<br /><br />As a fan of FSLR and, quite frankly, as an American citizen, what's frustrating to me is what seems to be a very non-level playing field when it comes to the competition between FSLR and the low-cost Chinese manufacturers you cited. Many of these Chinese companies have been given extremely generous subsidies from the central government (free land, low-interest or no-interest loans, etc.), have access to extremely cheap (and dirty) electricity, and most importantly, benefit from an imbalanced exchange rate between the two countries. All of these factors contribute to the low non-poly costs you referenced, thereby making FSLR's path that much more difficult. <br /><br />As a citizen in a country where consumption has long outpaced production and manufacturing, I for one would be happy to see a tariff put in place to level out the playing field and give companies like FSLR a fair fight. But perhaps this is a story for another day.Kevinnoreply@blogger.comtag:blogger.com,1999:blog-4815867514277794362.post-73580634308635611482010-04-17T02:35:49.518+10:002010-04-17T02:35:49.518+10:00Efficiency is key to First Solar's cost goals....Efficiency is key to First Solar's cost goals. I reached the same conclusion some months ago-- and the efficiency needed doesn't seem in the realm given their past performance. The follow on question should be: If I were FirstSolar, and have a bunch of money, what would I do to get those efficiency goals while best leveraging internal expertise? My conclusion was: They would switch to CIGS-- and have assumed that in my projections for FirstSolar going out to 2015. Recent reports (see links) indicate that this is likely and possibly by end of 2010 (which is sooner than I had expected, to be honest-- I thought by early 2012)<br /><br />http://guntherportfolio.com/2010/03/first-solar-skunk-works/<br /><br />http://www.reuters.com/article/idUSTRE6374W620100408<br /><br />Chitra Seshanchitra seshannoreply@blogger.comtag:blogger.com,1999:blog-4815867514277794362.post-3722703801833544212010-04-16T22:28:28.427+10:002010-04-16T22:28:28.427+10:00STP's non-silicon cost refers to the cost of p...STP's non-silicon cost refers to the cost of processing a purchased wafer into a module, that is why it is so low.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-4815867514277794362.post-15657592548046566762010-04-16T12:59:15.293+10:002010-04-16T12:59:15.293+10:00You are a hand-wringin', mouth-breathin', ...You are a hand-wringin', mouth-breathin', thumb-suckin', bed-wetter! Have the courage of your convictions, sir!Anonymoushttps://www.blogger.com/profile/02052452385743792288noreply@blogger.comtag:blogger.com,1999:blog-4815867514277794362.post-6052150868112338232010-04-16T07:27:26.158+10:002010-04-16T07:27:26.158+10:00Big improvement over your initial analysis. I stil...Big improvement over your initial analysis. I still think you put too much weight on poly cash cost. All-in cost is what matters over the full cycle. Sure, spot poly will follow the cash curve in times of oversupply, but FSLR's long-term contracts partially insulate them from such temporary effects. The timing is key here, if spot poly is nearing an uptrend by the time the bulk of FSLR's contracts roll off you're hosed.<br /><br />The 25-30 cent FSLR penalty sounds like a throwaway number -- i.e. 10% of module price back when the report was written. Why use a fixed percentage? It's not as if brackets get cheaper when module prices plummet, do they? Well, actually they do. It makes little sense to invest resource into reducing bracket cost from $0.20/W to 0.10/W if modules are 4.00/W. Different story when modules are 1.00/W. Add to this normal scale effects as cheaper prices grow demand. <br /><br />Look at 2010 module and system price estimates on pages 24-25 of your NREL report. In 2008 they estimated system price at 4.75. In 2009 they cut the system estimate 50 cents. But they only cut their module estimate 15 cents. Balance of system prices actually fell faster than module prices!<br /><br />For all their flaws, Nanosolar worked on balance of system cost from the very start. They knew a 0.30/W cell wouldn't revolutionize the world if BOS was still 3.00/W. They have some PR/white paper stuff dealing with BOS, including head-to-head comparisons with FSLR. Might be worth a read.doggydogworldnoreply@blogger.comtag:blogger.com,1999:blog-4815867514277794362.post-48904348204218103382010-04-16T05:43:59.006+10:002010-04-16T05:43:59.006+10:00John,
Suntech's non-silicon cost means someth...John,<br /><br />Suntech's non-silicon cost means something different than yingli's or trina's. Suntech purchases wafers, so the non-silicon cost refers to the conversion costs of wafers to cells and then to modules. The other players purchase polysilicon, and their non-silicon cost refers to the conversion of polysilicon all the way to module. Cheerios, SlurryAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-4815867514277794362.post-58289626916273962372010-04-16T04:16:38.880+10:002010-04-16T04:16:38.880+10:00As Abhishek points out, if poly is going to $20 yo...As Abhishek points out, if poly is going to $20 you are way better to short WFR, WCH GR or other poly guys.<br />No need to worry about efficiencies.<br />You just need to be sure about their semi poly businesses, which unfortunately are enfuego right now.Nickhttps://www.blogger.com/profile/01164675502109841117noreply@blogger.comtag:blogger.com,1999:blog-4815867514277794362.post-89286021234360086572010-04-16T04:01:58.265+10:002010-04-16T04:01:58.265+10:00John, I have no position in FSLR, and I have no wo...John, I have no position in FSLR, and I have no worthwhile independent view of the business itself.<br /><br />But I would be far more worried about your short if the push-back you got was about FSLR's manufacturing and research prowess than about their sales prowess.<br /><br />Anybody can sell anything. Motorola once was a "sales machine"--they just financed their own customers, including some who had no intention of paying.<br /><br />Selling is easy. Making stuff cheap and well is hard.<br /><br />JMJeff Matthewshttps://www.blogger.com/profile/04755227781599499378noreply@blogger.comtag:blogger.com,1999:blog-4815867514277794362.post-74977875493487972872010-04-16T03:14:52.977+10:002010-04-16T03:14:52.977+10:00John, I'm a trader, not a fundamental (this is...John, I'm a trader, not a fundamental (this is the right spelling by the way, not "fondamental" which is the french spelling) analyst, so I'll say this one more time at the risk of greatly annoying you (in which case I apologize): at this moment, technically the stock does not look good as a short. It has based roughly around 100 several times in the last couple of years but the bears could not push it much lower than that. And now a nascent and nasty short-squeeze seems to be taking place. FSLR is trading above its 200-day moving average as I'm typing this and the 200 resistance level is beckoning. <br /><br />That said, I don't know anything about your expected holding period or the way you got short or where you got short and thus how much heat you're willing to take, so my opinion is just that, an admittedly myopic opinion.Isam Larouihttps://www.blogger.com/profile/17824104233197250291noreply@blogger.comtag:blogger.com,1999:blog-4815867514277794362.post-7018767166845377882010-04-16T02:20:43.208+10:002010-04-16T02:20:43.208+10:00Here's some cost numbers: "Jiangsu Zhongn...Here's some cost numbers: "Jiangsu Zhongneng adopts the most advanced hydrochlorination technology and the<br />modified Siemens process to produce polysilicon. Through hydrochlorination,<br />by-products from the polysilicon production process – Silicon Tetrachloride and<br />Hydrogen Chloride – are effectively recycled back into TCS, one of the major raw<br />materials used in the polysilicon production process. Not only is this process<br />environmentally friendly but it also helps to significantly reduce unit production cost.<br />Jiangsu Zhongneng’s production cost decreased from US$66/kg in 2008 to US$36/kg<br />for the month of June 2009, which is highly competitive versus its global peers."<br />http://www.gcl-poly.com.hk/attachment/2009092121055117_en.pdf<br /><br />"Average selling price was US$65.4 per kg for polysilicon and US$0.83/W for wafers. It sells to Trina."<br />http://www.gcl-poly.com.hk/attachment/2010031723481317_en.pdf<br /><br />RegardsWai Cheehttps://www.blogger.com/profile/01146399854043442388noreply@blogger.com