In general, we are looking at a pretty substandard set of Q2 earnings and guidance from the 11 Chinese solars. In the midst of all the mayhem, there were encouraging metrics and comments. For these solars, it appears that Q3 will be the earnings trough for seven of the 11 solars. Below, we provide notes summarizing some of the important points helping to shape our Q3 and Q4 EPS estimates. We will not go through the headline performance because those numbers have been well documented. For further information on the solar metrics, please refer to our research website.
Although many of the points might seem dry and technical, earnings releases and conference calls provide very important building blocks for modeling EPS estimates and ultimately forecasting stock prices.
This article will be the first step to providing a series of articles for Seeking Alpha:
- The Earnings Summary for Q2 (This article)
- The Q3 EPS Estimates for the 11 Solars
- The Full Year Estimates for the 11 Solars for 2011
- The Full Year Estimates for the 11 Solars for 2012
- Stock Price Forecasts Based on the Full Year Estimates
The 11 solars discussed in this article along with a good, bad or ugly are listed below:
- Canadian Solar Inc. (CSIQ) Bad
- China Sunergy Co., Ltd. (CSUN) Ugly
- Daqo New Energy Corp. (DQ) Good
- Hanwha Solarone Co., Ltd. (HSOL) Bad
- JA Solar Holdings Co., Ltd.(JASO) Bad
- Jinko Solar Holding Company Limited (JKS) Good
- LDK Solar Co. Inc. (LDK) Ugly
- Renesola LTD (SOL) Bad
- Suntech Power Holdings Co., Ltd. (STP) Ugly
- Trina Solar Limited (TSL) Bad
- Yingli Green Energy Holding Co. Ltd. (YGE) Good
Just to summarize, of the 11 solars, JKS continues to shine. YGE was a surprise in Q2 results. CSUN and STP need a lot of help.
This is a company in a bit of trouble. In good times, the weaknesses are somewhat hidden but under stress conditions, the weaknesses become pronounced. They made wholesale changes in guidance for shipments. They appear to continue to sustain losses for Q3 and Q4. As well, they will most likely post a loss for the full year 2011.
The new Quasar modules will produce 2 to 4 cents to gross profits compared to current modules. However, in the medium term, the company noted that shareholders would have to be patient. On a somewhat positive vein, they have accounted for most of the projected Q3 shipments and about 25% of the Q4 projections. They expect ASPs for modules to range between $1.22 to $1.25 for Q3 and $1.15 to $1.20 for Q4.
The CC was very confusing in terms of processing costs. Management stated 25 cents for modules by end of Q3 yet later they stated 29 cents. The 29 cents makes much more sense. Cell processing costs should be 21 cents for end of Q3. Wafer processing costs were fairly inefficient for Q2 (25 to 30 cents) but they expect 23 cents for Q3. I am wondering if the 25 cent number was meant for YE 2012. Again, as with other solars, it was frustrating getting an idea of a point estimate for many metrics.
ASP guideline for July ranged between $1.30 to 1.40/watt. ASP guideline for Aug was:$1.25 to 1.30/watt (they mentioned possibly under 1.30 so I assume about 1.25). They were hesitant to give a range for September but thought it might be similar to August. The CSIQ ASP numbers are a bit more hopeful than many of the other companies.
For 2012, they are looking to Japan for 2 to 3 GW of installations. They already have 200MW shipped or firm ordered for Q3. It appears that Q3 wafer ASPs are around 50 cents and cell ASPs are around 80 cents. GM will fall from 13.20% to 9 to 12% for Q3. They might have difficulty turning a profit for Q3. Right now my spreadsheets are showing a Q3 profit.
GCL will co-locate 600MW to the CSIQ campus. This project will be expandable to 1.2 GW.
DQ is an interesting company because they are doing very well so far this year. Sky high gross margins in the 50% range make them attractive on paper. The CC revealed a lot and it appears that each successive quarter this year will see a drop in EPS. This is not a good sign. Even more alarming is the fact that they will most likely be facing a 2012 with a declining 2012 EPS over 2011.
They have secured a $153 million loan from the Bank of China for the completion of the phase II poly plant. The interest rate will be around 7.5% (seems pretty high).
As expected, they have been running at 100% capacity. They have projected $50 to $55 poly for Q3.. The wafer and module business has not been very impressive to date. Surprisingly, they have brought the costs down significantly to 25 cents for wafers and 30 cents for modules.
They feel demand in China will be around 500 to 900 MW for 2011.
Like a few other companies (such as CSUN and STP), HSOL ER was not very encouraging. Not only did they lose money in Q2 but it looks like they will continue to be in a loss position for Q3. It is possible that by Q4 they will show a profit. This is a company in transition from an inverted pyramid model to a more integrated model. Right now it is hard to tell what they will look like in a few years from now.
Poly prices should reach $50 exiting Q3 and then in Q4 should be just under $50. Their Q3 blended poly cost will be around $52. Not that this means a lot but they mentioned they have not signed any agreement for purchase of poly from Hanwha.
The mother company, Hanwha, will contribute some valuable downstream customers. This should result in 60 to 120 MW of orders for this year. They indicated that the parent company should provide even more orders for 2012. This is one major advantage for HSOL. However, we should keep in mind that Hanwha, the parent company, will most likely not be doing this for free. So a follow-up question during the CC should have asked about fees paid to the parent.
Management agreed that with Hanwha as the parent company, HSOL does not have to remain public to function in the solar business. However, they quickly mentioned that there are no plans to change the HSOL structure at this time.
ASPs for Q3 will range from $1.25 to $1.30 and for Q4, from $1.15 to $1.20
Non silicon processing cost for Q2 was 86 cents and for Q3, the cost will be about 75 cents. The drop is fairly significant and was explained by higher depreciation per watt due to under-utilization for Q2, silver paste and other cost reduction actions.
The JASO ER for Q2 was not good. The CC was brutal. We had estimated a loss of 15 cents. They reported a 22 cent loss. That was ok. It was the downward revisions to forward shipments. In their defence, it appears that for them, Q2 will be the low point for this year. As in the past, they refused to share any important information on metrics. They gave vague rough percentages as answers to many questions. So we are left to estimate metrics often based on metrics shared by their peers.
On a preliminary basis, for Q3 we now estimate an EPS of 10 cents. However this could change with the uncertain pricing of cells at this time. For the full year we had to revise our EPS estimate of 80 cents down to 43 cents.
The big news with LDK was the revisions to Q2 guidance a few weeks ago. The ER simply confirmed much of the revisions. They should record a small profit for Q3 and a bigger profit for Q4. Q1 will be peak earnings for this year.
18.6 m shares have been bought back. This was a startling admission and positive in light of LDK's current series of management blunders.
They have guided significant growth in wafers, cells and modules for Q3 and Q4. Wafer cost is 26 cents versus 21 cents for GCL. Current blended poly was $61.8/KG. Q3 module ASP is in a range of $1.20 to $1.25. They feel ASPs have stabilised recently which is confirmed by peers. 50% of Q3 sales will be in China.
They have a $1.1 billion LC for operating purposes and the $9 billion CDB credit facility for project financing.
Although they have a $30/KG target for poly, we are wondering when this might happen as they seem to be greatly behind on the processing cost roadmap. Even for Q2 we had an estimate of $37 processing cost but they reported $40/KG.
Of significant note, the CDB has funded non-China solar projects. This is very encouraging not only for LDK but also for the other solars. Over the next few years, they should all be able to unlock a very valuable asset.
Just from the poly stats provided, we are beginning to wonder if they will achieve full capacity of 25,000 MT by the end of this year. As well, this would make it difficult to have an IPO by the end of the year if they cannot produce poly at a competitive price to peers.
Although the JKS results were very good considering the overall solar market conditions, the CC was a bit sobering in terms of H2.
Non-silicon processing costs fell from 73 cents/watt to 70 cents/watt. They expect to reach 67 cents by the end of Q4. They feel that they could achieve 60 to 62 cents by Q4 2012. A large part of this reduction would be due to reductions in material prices. 51 cents of the current 70 cents is in materials. I believe they mentioned that they could purchase even more of the material portion from Chinese sources.
Poly costs on a signed contract basis appear to be hovering from between $51 to $53 per kilogram. Although a bit hard to understand it appears that spot prices would range from between $40 to $50 per kilogram. Together with outstanding inventory, we will assume the average cost of poly will be about $53.
For Q4, depending upon module pricing, they assume poly pricing could fall to $35. Although logical, it seems prudent to model $4o to $45 for Q4.
They mentioned a lot of the non listed Chinese manufacturers were either going out of business or reducing production. JKS has been at 100% production over the past two quarters.
For Q3, they expect a module ASP range of $1.20 to $1.30 per watt for the quarter. This seems very low. Are they being overly conservative? We are now at the midpoint of Q3 and I was hoping that they would have provided a point estimate for Q3. Guess one of the analysts should have been more precise. For Q4 they are projecting around $1.15 to $1.25.
I was surprised that no one mentioned the fairly large growth in accounts receivable.
The Q3 and full year guidance do not really reconcile well (in a good way). I can only assume that the 250 mw for modules is understated and that Q4 will see around 360 mw of sales. Just with rough calculations, total revenue would then reach the high end of revenue guidance of $1.5 billion.
They gave a pretty interesting picture of the solar landscape during the CC. Not a very rosy picture. If we use $1.28 ASP, then we are looking at an EPS of $1.80 for Q3. However at $1.22 ASP for modules, the EPS drops to about $1.21.
The SOL ER was not optimistic for Q3 and Q4. SOL indicates that gross margins will only be about 7% for Q3 and 10% for Q4. This means, they would report a loss for Q3 and break-even for Q4.
Wafer processing cost rose 2 cents/watt due to conversion to Virtus wafer production. They expect wafer processing cost to fall to 19 cents by year-end. In-house poly costs were $40/KG. The blended cost was $54. Poly processing costs should fall to $35 by year-end.
Poly expansion to 8,500 MT has been delayed to the end of Q2 2012. Gross profit margin will fall to about 6 to 8% for Q3. Current wafer price is about 52 to 55 cents. Current module price is about $1 to $1.20 per watt. Virtus premium is about 5 to 10% over normal wafers.
8500 MT of poly should cover 80% of internal needs. Including tolling, they might be able to cover 100% of poly needs. Second phase is hydro-chlorination for poly.
Q3 module should be $1.20 and wafer about 55 cents per watt. Steel wire production will be in full production exiting 2011. Wafer prices for 2012 could go to as low as 45 cents per watt.
They expect China demand at 3 GW at $1.10/watt and for 2012, about 3 to 5 GW at $1 per watt. The SOL pricing seems too low.
Q4 gross margin might only be 10% which is kind of low. Q4 poly production could be around 850MT.
This was a pretty poor showing by one of the solar tier one companies. Bottom line is that they will most likely lose money in Q3 and Q4 as well as the loss announced today for Q2. There 2011 year is also headed for a loss so together with CSUN, two of the 11 solars are projected to lose money this year.
Since this year is a write-off for them, can they turn things around for next year? It is hard to tell but if they can clean up some of their metrics, they could make 50 cents next year.
They indicated a 22 cent wafer processing cost for Q3. For this year, they feel Germany will account for 7 GW and Italy, 2 to 3 GW.
They feel that the current production capacity in China is 20 to 35 GW. However, their take is that many tier 2 and tier 3 lines are idle or have closed down. Practical current capacity capable of producing at current prices might be between 15 to 20 GW.
Average interest rate for STP is about 6% on a variable basis. They suggested the following metrics for China FIT projects: 1.15 RMB for FIT, 1700-2000 sunshine hours and IRR of 8 to 10%. They mentioned non-silicon cost of 54 cents (I assume non-wafer cost). They felt this had a potential to go down to 45 cents.
Well, not much to say about TSL. The bad news was already out with the revised guidance a few weeks ago. From all the information provided, it looks like they will make around 40 cents (plus or minus) for Q3 and somewhere around $2.10 for the year.
They indicated 73 cent non poly processing cost for Q2. If we add 43 cents poly, the total in-house processing cost was $1.16/watt. For outsourced wafers, they added 5 cents to the total for $1.21/watt. For Q3, they expect to reach 70 cents per watt PC sometime in July. This supports the statement that they will exit 2011 under 70 cents for non-poly costs.
On the GCL rumour that they would start a 10GW module plant with Foxconn, obviously Gao was pretty upset because he phoned the CEO of GCL. It would have been special to listen to that conversation. The CEO of GCL assured Gao that they would not be making modules.
I have believed for a long time that GCL could be a huge threat to Chinese solars. I am sure that TSL regrets allowing GCL to co-locate on their campus to produce poly and wafers. To me, that was a huge tactical error on TSL's part. TSL unfortunately will pay for many years to come. Just as an example, for 2012, we are modeling one GW outsourcing of wafers. At five cents per watt, this represents $50 million to gross margins or potentially 70 cents in EPS. Just rough numbers, full integration would add about 25% more to the TSL bottom line for 2012.
YGE was a pleasant surprise in the midst of our solar turmoil. With an EPS of 36 cents, they managed to beat the Street (29 cents) as well as my estimate (25 cents). With the steep declines in module ASPs we feel that Q3 will see a decline in earnings to about 11 cents.
They are producing pretty well at 100% capacity. It appears that they will be sold out to the end of the year. The big problem right now are the spot prices. It was a bit irritating that they did not specify specific range metrics for ASPs and future shipments but instead chose to use words like high 20s or low teens (in terms of percentages).
Their processing cost roadmap seems to have worsened over the next few quarters (compared to my spreadsheet numbers). They now appear to be 73 cents for Q2 and going down to 70 cents entering 2012. Poly per watt will fall from 5.8 g/watt to 5.5 g/watt. It appears that poly pricing will be around $50 for Q3 and $47 for Q4.
For the Panda line, YGE will have a 10% higher processing cost at this time and 15% premium in pricing. By next year, they expect with economies of scale to reduce the processing cost to 5% higher than standard processing. They are equipped to produce 300MW at this time.
In terms of the poly plant update, for 2012, the plant will be either neutral or a drag on earnings. They expect processing costs in the range of $45 (Q1 2012) to $35 (Q4). They expect to reach $25 by YE 2013. As a separate cost center, we estimate that they could lose around $15 million in 2012 bringing the poly plant up to full efficiency. The plant could end up being earnings neutral by 2013 if they only reach $30 per kg because the spot price of poly might fall to $30 by 2013. GCL already has poly costs down to $20 per kg.