explo

LifeTime Member
  • Content count

    9,102
  • Joined

  • Last visited

  • Days Won

    144

explo last won the day on April 20

explo had the most liked content!

Community Reputation

653 Excellent

2 Followers

About explo

  • Rank
    师傅

Profile Information

  • Location :
    Europe

Portfolio

  • Portofolio %
    FSLR 10%, DQ 4%, JKS 5%, JASO 4%
  1. LOL, I was thinking the same thing already regretting taking it back instead of putting all CN3 in JKS. Looking at the 3m chart its been the man though. That's why I dumped it. I can't believe I got tempted to momentum play all of a sudden and took it back. Momentum playing JASO seems like fighting the laws of nature. Jokes aside. Looking at fundamentals JASO had the best operating margin in Q4 and Mr. Market seems to be in the process of correcting its view on it.
  2. All very green except for panels: http://pvinsights.com/
  3. Jinko are betting on mono ingot capacity (where they have some properitary tech edge) to support their mono PERC capacity. My main worry is not this bet instead of multi plus black silicon, but rather how they are doing on DWS. DWS still adds most value to mono ingot slicing even though it can be used for multi too now. DWS seems more no brainer and critical transition to me, while multi vs mono is still open and Jinko's and JA's strategy to not go all in on multi historically could benefit them long-term. It gives them a tech diversification risk reduction at least. GCL and YGE are the ones who has taken most crystallization tech risks, TSL and CSIQ outsourced a good portion of that risk. CSIQ bet on DWS capacity seems spot on.
  4. I don't think it means much. Looking at a different time frame you get a different relative move. For example on March 8 I moved all my CSIQ at $13.49 to JKS at $15.69. I don't know how CSIQ will behave but I think there are some near-term uncertainty on their panel sales profitability. A while ago some people here had entry targets in single digits for it. For me a relative rise in my holding to CSIQ will make CSIQ re-entry look attractive. Especially a large rise in FSLR might result in a move of half FSLR to CSIQ. It's actually only DQ I'm excited about. I recommend reading their Q4 ER and 20-F. JKS, DQ and FSLR's annual reports have been out for a while. I think these look like good risk/reward here. JASO is still cheap and was strongest in the difficult Q4.
  5. Yes and then there's a lot left to drop down. Japan and China should be good for them.
  6. It's all about the bottom line and margin trend though. They'll be selling a billion or more worth (703 MW) of project assets in US, but their own comment on the margin outlook of that is "I think we can make money". If such a big revenue chunk is not making money it is important for the stock that investors don't look at GM trend in 2017 and apply it to revenue outlook for 2018. The J-REIT IPO is what I'm excited about for CSIQ and they have a decent chunk (198 MW) of project asserts in China to sell (70 MW to be sold in 1H17) and they say they have "very good margin" despite being quite reluctant to talk them up when they were developing them. This year China is where they will develop most projects (400 MW). The J-REIT IPO size seems a bit small though. 60 MW in operation end of 2016 and 105 MW to be developed this year, 70 MW in operation at IPO time.
  7. For once the other stocks should not react too much on FSLR's ER since FSLR should focus on ramping down panel production and reducing project assets to release cash, so basically they should focus on not be tempted to win too much new business. If other stocks dip on that it could present an opportunity.
  8. That's just how much I allocate to stock in genaral. My allocation to solars is 100% of my stocks.
  9. I moved half of my JKS to JASO. Net effect of the 3 latest strategic moves is that I moved CSIQ to DQ.
  10. I think BS can't be igored in valuation discussion. Adjusting PPS for net cash/debt per share FSLR is a $10 stock and CSIQ a $40 stock. To be fair CSIQ is intent to reduce its net debt per share this year and FSLR will likely need to reduce it's net cash per share which they might not fully be able to match with faster monetization of Systems business assets held on BS. Both these should also be considered as forces that will close the cash adjusted stock prices.
  11. You are right. Despite the 90m OPEX cut this year They are increasing third party panel shipments some 40% further in 2017 although revenue is only slightly up as ASP is as you correctly state down 27% from 52 cents in 2016 to 38 cents in 2017. I've noted though that there is discrepancy in MW shipped and sold in their reporting so 38 cents is to be considered rough estimate. The steep ASP decline in 2017 is why I think ASP will consolidate following years. Noted also that 38 cents for S4 corresponds to 42 cents for S6. I think the OPEX percent of revenue increasing is more an effect of falling ASP and revenues than increasing third party shipments. That's why the effort to cut OPEX with 23% in 2017 makes good sense. FSLR outlook is not rosy, but FSLR's outlook relative to market challenges is decent since sensible actions are made to address it. Think about how things will add up for CN4 with a couple of cents per watt in OPEX, a couple of cents per watt in outsourcing penalty and a couple of cents per watt in tariffs to add to in-house costs. Jinko is growing volumes a lot and have big exposure to highest ASP region (China) so they might fight the headwinds, but JASO they'll survive, but without glory.
  12. Their third party panel sales trippled in 2016 and SG&A was only up $6.8m. Since they, unlike CN4, include freight and warranty costs in cost of sales instead of OPEX their OPEX is much more fixed than it is for CN3.
  13. You are right. It brings $396m in gross profit from the Components segment in 2019, which is 17m more than the 379m in 2016. In 2017 they will reduce OPEX with 90m through 30m reduction of R&D and 60m reduction of SG&A. At least the latter is in full part of the restructuring of their Components segment and is thus likely to persists. You are also right that 2019 Component GP just barely covers total OPEX and their 2016 Components GP was just 8m short of doing the same. However the Components OPEX in 2015 was only 44% of total OPEX. The OPEX distribution was not disclosed in 2016 but this indicates that it was around 180m and half of that is to be cut this year, so the panel business is clearly profitable ($1.66 EPS before tax in 2015 and maybe $1.90 in 2016). But let's assume 2019 $396m Components GP more than covering all 295m OPEX in 2017 and that it will be similar OPEX in 2019. Then they have $1 EPS plus all GP from Systems and equity in earnings (171m reported equity in earnings net of tax in 2016 derived from unconsolidated revenues). To conclude. The issue isn't the EPS contribution from panel sales. It has be growing quickly after the 2012 crash and turned distinctly positive already in 2015. With S6 this growth is expected to continue after the re-tooling blip. The real question is what will happen to their big EPS contribution from the Systems segment that they have enjoyed since 2012. Attached chart shows how the role of hero segment changed in 2011 and is changing back again now.
  14. I've crunched their numbers now. I think their Components segment (internal and external panel sales) will contribute $1 more to EPS in 2019 than it did in 2016. 2016 non-GAAP EPS was a bit over $5. The question then becomes how much the Systems segment (everything else) contributes in 2019 compared to 2016. I don't have a good view of this. Assumptions about the Components business is that internal and external panels sales will be at fully ramped capacity (if this happens in 2019 or two months later is of less interest since the fully ramped capacity metrics is what is interesting going forward) which is assumed to be 3.3 GW based on guided gains over today's 3.0 GW from e.g. increased power per panel produced. Then the cost per watt is quite clearly guided as 0.23 $/W and the S6 ASP entitlement in the 2016 market as 0.58 $/W (6 cent higher than their S4 ASP in 2016 which was 0.52 $/W). I'm then assuming ASP will fall 40% from 2016 to 2019 to end up at 0.35 $/W (I know this seems high compared to today's spot price on standard multi c-Si panels, but there is a severe supply/demand imbalance for standard multi c-Si panels). 40% decline in 3 years is an average decline of more than 15% per year, which is not a very conservative assumption on the PV cost reduction pace in the march towards grid parity. Anyway this $0.12 GP on $0.35 ASP translates to a panel sales GM north of 30%. Already in 2015 and 2016 major S4 cost reduction were achieved to reach panel sales GM of 20.0% and 25.5% respectively after 3 years of single digit GM on panel sales. In that light 30% GM is not stretching it given the ASP entitlement and cost advantages S6 is expected to have over S4. If FSLR can't make decent money on their S6 panel sales, assuming the guided cost and ASP advantage over S4, I have a hard time seeing how their competitors can make money on their panel sales at all. So FSLR is a hedge based on likely better profitability than competitors on panel sales once S6 is launched, especially in markets with climate where their technology has an energy yield advantage over c-Si. This will be a position of strength in the market to motivate growth investments. It's not based on FSLR making a lot of money in 2019. It is a hedge based on the fact that if FSLR have problems its compentitors could have even worse problems (BS strength, revenue stream diversification considered) and I don't want to risk being concentrated in deep problems. That said I know that panel profits won't suffice as return for FSLR. They need to continue to add Systems profits to that. At least the Components profits are growing now and will likely continue to grow after S6 launch and FSLR shareholders are a bit starved for clear signs of a growth segment. I still have a 50% higher bet on c-Si than FSLR, but I think there are good risk/reward reasons to spread my bets like this.