Ed Job - Director of Investor Relations Shawn Qu - Chairman, Chief Executive Officer and President Michael G. Potter - Chief Financial Officer and Senior Vice President

Philip Shen - Roth Capital Partners, LLC, Research Division Colin W. Rusch - Northland Capital Markets, Research Division Brandon Heiken - Crédit Suisse AG, Research Division James Medvedeff - Cowen and Company, LLC, Research Division

Good day, ladies and gentlemen, and welcome to the Canadian Solar Third Quarter 2013 Earnings Conference Call. My name is Lacie, and I will be your operator for today. [Operator Instructions] As a reminder, this conference is being recorded for replay purposes. I would now like to hand the conference over to Mr. Ed Job, Canadian Solar's Investor Relations Director. Please go ahead.

Thank you, operator. Welcome to Canadian Solar third quarter earnings conference call. Joining us on the call today are Dr. Shawn Qu, our Chairman and Chief Executive Officer; and Mr. Michael G. Potter, Senior Vice President and Chief Financial Officer. Before we begin, may I remind our listeners that in this -- in today's call, management's prepared remarks will contain forward-looking statements, which are subject to risks and uncertainties, and management may make additional forward-looking statements in response to your questions. Therefore, the company claims the protection of the Safe Harbor for forward-looking statements that is contained in the Private Securities Litigation Reform Act of 1995. Actual results may differ from management's prepared remarks and, therefore, we refer you for more detailed discussion of the risks and uncertainties in the company's Annual Report on Form 20-F filed with the Securities and Exchange Commission. In addition, any projections as to the company's future performance represents management's estimates as of today, November 13, 2013. Canadian Solar assumes no obligation to update these projections in the future unless otherwise required by applicable law. At this time, I would like to turn the call over to Dr. Shawn Qu. Shawn, Please go ahead.

Thank you, Ed, and thank you all for joining us on the call today. The third quarter was a very strong quarter for us. Revenue and gross margin exceeded our guidance, and we achieved a gross margin of 20.4% with net income of $27.7 million. This allowed us to reach our goal of returning Canadian Solar to profitability for the first 9 months of the year. Our success has been mainly driven by our total solutions business, which accounted for over 41% of the total net revenue in Q3, up from about 26% in Q2. Our pure module business also made a solid improvement, delivering higher margin on the back of solid demand and improved pricing power. Our powerful momentum is clearly evidenced by the project sales we announced, new project investors secured, continuous progress on the development of our existing pipeline, including the expansion of our late-stage project pipeline in Japan. It is also evidenced by large module contracts the company announced in China and emerging new markets. In short, Canadian Solar continues to strengthen its industry leadership position, and we continue to build value for our shareholders. In Q3, we shipped 478 megawatts, comfortably exceeding our shipment guidance of 410 to 430 megawatts for the quarter. From a geographic standpoint, we saw strong demand from Japan, China, U.S., Thailand and India as well as other emerging markets. Our module sales to Japan continued to be strong this quarter, putting us on target to ship over 500 megawatts for the full year, representing around 8% to 9% market share in Japan. These sales are in support of both residential of commercial opportunities. Our brand is strong in Japan and carried a lot of positive quality in a market with a reputation for quality and performance. We were one of the first foreign module suppliers to establish our brand in Japan, and our timing was ideal. We are confident we can continue to leverage our first mover advantage in this important growth market as we pursue larger opportunities on the commercial and on the utility-scale project market segment. Our sales to Europe represented 9.5% of the total revenue. The demand in Europe has been weak due to the negative impact of the resolution of the trade dispute between EU and China. We have carefully reduced our exposure to Europe since late last year while diversified into other markets. This effort apparently paid off as shown by the strong continuing growth of total shipments in the past 3 quarters. Let me now provide the highlights of the progress we made on our total solutions business. During the third quarter, we completed the sale of 2 utility-scale solar power plants in Canada valued at over CAD 95 million to TransCanada. We signed an agreement to sell 2 other power plants totaling 20 megawatt AC for over CAD 120 million to BlackRock. In addition, we started the construction of the 130 megawatt DC solar power plant in Ontario, Canada on behalf of Samsung. As mentioned previously, this EPC contract represents approximately CAD 310 million of revenue for Canadian Solar. This EPC contract may be expanded to include phase 2 and phase 3, each expected to be 130 megawatt DC once these 2 phases receive full permit from the Ontario Power Authority. Our backlog of late-stage solar projects and EPC contracts in Canada now stands at around 500 megawatts with an estimated revenue opportunity of over CAD 1.7 billion once the projects are built and connected to the grid. In Japan, we expanded our pipeline of late-stage projects to 278 megawatt DC. All of these projects have the [indiscernible] feed-in tariff approval and are in Stage 1, 2 or 3 of the utility approval process. We expect to start building our first solar power plant there in the beginning of 2014. In U.S. and China, our pipeline of utility-scale projects at the end of Q3 totaled 198 megawatts and 40 megawatts, respectively. In total, our global utility-scale cellphone projects and EPC contract pipeline now stands at over 1 gigawatt. Clearly, we are making major progress in our business transformation from a module manufacturer into a one-stop shop provider of solar power solutions with a global reach. The total solutions business requires a combination of composite elements, including special talents in project development, permitting, EPC and banking, as well as strong and long-time track record with infrastructure investors and banks. As a result, we believe that this space is not yet -- a lot less crowded than the manufacturing space and provides us with the right platform to power our growth in the years ahead. As we look into the future, we continue to view our total solutions team and geographic reach to grow our project pipeline and extend our visibility. As we disclosed previously, our late-stage backlog is over 1 gigawatt and our soft pipeline of early-stage projects exceeds 3 gigawatts. This extends the visibility into our business beyond the next 4 to 5 years. As an example of the potential opportunities ahead, you may have noticed in a recent news report that Canadian Solar recently signed a memorandum of agreement with the government of the provincial of Punjab to build a solar power plant in Pakistan. At this time, we are still at the early stage of assessment and due diligence in Pakistan, and there's still a lot of work ahead of us. But we view this as a sign of how strong the project business for Canadian Solar can be in many areas of the world. In support of our planned growth, we raised approximately USD 50 million in gross proceeds during the third quarter using an at-the-market offering of the company's common shares. Clearly, our strategic position in the solar power industry is differentiated and strong. Our focus going forward is to continue executing the key elements of our strategy. That is, to grow and expand our module business by entering new emerging markets, continue to drive our manufacturing costs down and to differentiate our business model by expanding the total solutions business in order to capture additional margins and position the company for sustainable and profitable growth. Now let me comment on our guidance for Q4 and the full year 2013. We expect Q4 shipments will be in a range of approximately 480 to 500 megawatts with volume mainly driven by China, Japan, Canada, the U.S. as well as other emerging markets in Asia. Our gross margin in Q4 is expected to be in a range of 13% to 15%. This does not include the potential positive impact of project sales, which I will further elaborate. Two of our solar power projects in Ontario, William Rutley and Mississippi Mills, are in commercial operations. These projects are connected to the grid and have been generating stable and higher-than-expected feed-in tariff income for us. We expect to reach approximately USD 60 million additional revenue with 20% to 25% gross margin for each of these 2 projects once the transactions are completed. As I noted a second ago, our current Q4 guidance does not include the potential margin contribution from these projects due to the uncertain timing of final closing of the transaction. The winter weather condition in Canada and the Christmas season are expected to be 2 special factors in the progress of acceptance test with customers in this quarter. For the full year 2013, we adjust our module shipment guidance up to approximately 1.75 gigawatts to 1.77 gigawatts from previous 1.6 to 1.8 gigawatts. Our guidance includes modules used in our projects. As we indicated previously, our forecast shipment volume for the year maintains our track record of continuous annual shipment growth. But our main focus for 2013 and beyond is to maintain our focus on profitable growth by leveraging our differentiated business model. Now let me turn the call over to our CFO, Michael Potter, for a more detailed review of our financials. Michael, please go ahead. Michael G. Potter: Thank you, Shawn. Net revenue for the third quarter of 2013 was $490.9 million, up 29.1% sequentially and up 50.6% compared to the year-ago period. Gross profit in Q3 was $100.2 million compared to $48.7 million in Q2 and $7.3 million in the comparable period of last year. The sequential increase in gross profit was primarily due to increased revenue contribution from our higher-margin total solutions business as well as higher gross margin from our pure module business. Our gross profit in Q3 is net of a $2.6 million depreciation charge, representing approximately 0.5% of revenue related to the unutilization of our manufacturing assets. Gross margin in Q3 was 20.4% compared to 12.8% in Q2 and 2.2% in the third quarter of 2012. This comfortably exceeded our guidance of 10% to 12%. The gross margin improvement in Q3 reflects the contribution of our high-margin total solutions business in Canada, the benefit of our geographic exposure to higher-value markets and our continuous and relentless efforts to reduce our manufacturing costs. Operating expenses were $44.9 million in Q3 compared to $36.4 million in Q2 and $41.8 million in the comparable period of last year. Interest expense in Q3 was $11.8 million compared to $9.9 million in Q2 and $15.2 million in the comparable period last year. The sequential increase in interest expense was a result of higher bank charges. The year-over-year decline in interest expenses due mainly to higher level of interest expense capitalized for our solar power projects. Interest income was $2.7 million in Q3 compared to $3.2 million in Q2 and $3.6 million in the year-ago period. In Q3, we recorded a loss on the change in fair value of derivatives of $1.6 million compared to a gain of $1.8 million in Q2 and a loss of $5.3 million in Q3 of last year. Net foreign exchange gain in Q3 was $2.3 million compared to a net foreign exchange loss of $20.5 million in Q2 and a net gain of $7 million in Q3 of last year. Income tax expense in the third quarter of 2013 was $12.4 million compared to income tax benefit of $5.1 million in Q2 and income tax benefit of $1.8 million in Q3 of 2012. Net income attributable to Canadian Solar shareholders for Q3 2013 was $27.7 million, or $0.56 per diluted share, compared to a net loss of $12.6 million, or $0.29 per diluted share, in Q2 2013 and net loss of $43.7 million or $1.01 per diluted share in Q3 of last year. Moving on to the balance sheet. In Q3, cash and cash equivalents increased to $273.7 million at the end of Q3 compared to $141.4 million at the end of Q2. The restricted cash balance was $408 million at the end of Q3, up from $399.2 million at the end of Q2. Our accounts receivable balance, net of allowance for doubtful accounts, was $271.8 million at the end of Q3, up from $262.9 million at the end of Q2, driven by our higher sales volume. Inventories increased to $220.6 million at the end of Q3 compared to $218.5 million at the end of Q2. Short-term borrowings at the end of Q3 totaled $801.6 million, down from $813.6 million at the end of Q2. Long-term debt at the end of Q3 was $190.5 million compared to $254.6 million at the end of Q2. Total net debt at the end of Q3 was down over $200 million to $310.5 million compared to a total shareholder equity of $373 million. Short-term borrowings and long-term debt directly related to utility-scale solar power projects totaled $228.3 million. Let me now take a minute to provide additional detail on our total solutions business. In Q3, we completed the sale of Brockville 2 and Burritts Rapids to TransCanada valued at over CAD 95 million. We have now completed the sale of 3 out of the 9 projects we agreed to sell to TransCanada. Our pipeline of self-developed or acquired projects on Ontario now totals 24 projects. Two projects, William Rutley and Mississippi Mills, are in commercial operations and earning the feed-in tariff. These projects are undergoing customer acceptance testing before we can close the sale and recognize revenue. In addition, 16 projects are expected to reach commercial operation in 2014 and 6 in 2015. At the time of this call, 13 of the remaining 24 owned projects in Canada are under construction. Turning to our EPC business, we started the construction of our 130-megawatt DC solar power plant on behalf of Samsung. As we mentioned in other occasions, we're also building and plan to operate 3 solar power plants totaling 39 megawatts DC as EPC contractor on behalf of Penn Energy Renewables. These 3 EPC projects are expected to be completed in the first half of 2014. In addition, as we transferred title of the Demorestville and Taylor Kidd projects, these projects are now part of our EPC pipeline, which allows us to book revenue on a percentage-of-completion basis. The construction of these 2 BlackRock projects started in the first quarter and third quarter of 2013, respectively, with commercial operation expected to be fourth quarter of 2013 and second quarter of 2014, respectively. In Japan, during the third quarter of 2013, we have been able to expand our late-stage utility-scale pipeline to approximately 278 megawatts. We're making feasibility assessments on more than 200 megawatts of additional utility-scale power projects. We plan to begin construction of our first Japanese power project in early 2014. In the United States, during the third quarter of 2013, we completed construction of 3 solar power plants totaling approximately 35 megawatts [Audio Gap] [Technical Difficulty] Michael G. Potter: I apologize to the listeners of the call. And we're not sure exactly where my voice cut out, so I'm going to start again on the balance sheet section. Moving on to the balance sheet. In Q3, cash and cash equivalents increased to $273.7 million at the end of Q3 compared to $141.4 million at the end of Q2. The restricted cash balance was $408 million at the end of Q3, up from $399.2 million at the end of Q2. Our accounts receivable balance, net of allowance for doubtful accounts, was $271.8 million at the end of Q3, up from $262.9 million at the end of Q2, driven by higher sales volume. Inventories increased to $220.6 million at the end of Q3 compared to $218.5 million at the end of Q2. Short-term borrowings at the end of Q3 totaled $801.6 million, down from $813.6 million at the end of Q2. Long-term debt at the end of Q3 was $190.5 million compared to $254.6 million at the end of Q2. Total net debt at the end of Q3 was down over $200 million to $310.5 million compared to total shareholder equity of $373 million. Short-term borrowings and long-term debt directly related to utility-scale power projects totaled $228.3 million. Let me now take a minute to provide additional detail on our total solutions business. In Q3, we completed the sale of Brockville 2 and Burritts Rapids to TransCanada valued at over CAD 95 million. We have now completed the sale of 3 of the 9 projects we agreed to sell to TransCanada. Our pipeline of self-developed or acquired projects in Ontario now totals 24 projects. Two projects, William Rutley and Mississippi Mills, are in commercial operation and earning the feed-in tariff. These projects are undergoing customer acceptance testing before we can close the sale and recognize revenue. In addition, 16 projects are expected to reach commercial operation in 2014 and 6 in 2015. At the time of this call, 13 of the remaining 24 owned projects in Canada are under construction. Turning to our EPC business. We started the construction of 130-megawatt DC solar power plant on behalf of Samsung. As we mentioned in other occasions, we're also building and plan to operate 3 solar power plants totaling 39 megawatts as EPC contractor on behalf of Penn Energy Renewables. These 3 EPC projects are expected to be completed in the first half of 2014. In addition, as we transferred title of Demorestville and Taylor Kidd projects, these projects are now part of our EPC pipeline, which allows us to book revenue on a percentage-of-completion basis. The construction of these 2 BlackRock project -- plants started in the first quarter and third quarter of 2013, respectively, with commercial operation expected to be fourth quarter of 2013 and second quarter of 2014, respectively. In Japan, during the third quarter of 2013, we have been able to expand our late-stage utility-scale pipeline to approximately 278 megawatts. We're making feasibility assessments on more than 200 additional megawatts of utility-scale solar power projects, and we plan to begin construction of our first Japanese solar power project in early 2014. In the United States, during the third quarter of 2013, we completed construction of 3 solar power plants totaling approximately 35 megawatts DC. Currently, we're building 64 megawatts DC and remain on target to complete approximately 91 megawatts DC in the U.S. during 2013. In China, our initial pipeline in China includes 2 solar power projects we are building totaling 40 megawatts as well as the 30 megawatts of Golden Sun projects we completed earlier this year. We expect to complete construction of the additional 40 megawatts in the fourth quarter of 2013. Although we generally do not detail early-stage projects as most do not become realizable projects, we are very active developing opportunities in Japan, China, United States, South Africa, Chile and Pakistan. Our model remains as a build-and-transfer model, but we have the ability to set up our own yield company or to work with existing yield companies if it becomes the best model for the future. In summary, we are very pleased with the progress we've made in the transformation of our business from a pure module supplier to a provider of solar energy solutions. During the third quarter, we returned Canadian Solar to profitability, generated $152 million in cash from operations, raised $48 million through a follow-on offering of shares and significantly strengthened our balance sheet, reducing our net debt by over $200 million to approximately $310 million at the end of Q3. Clearly, Canadian Solar is coming out of the recent downturn in a strong position to deliver profitable growth in the quarters ahead. As you've probably learned from our press release issued earlier today, with the agreement of the board, I've resigned from Canadian Solar's Board of Directors in an effort to enhance board independence and allow for the appointment of another independent director. This was as a result of us considering and agreeing with the recommendation of the leading shareholder advisory company during the recent annual proxy voting. With that, I'd now like to open the call to your questions. Operator?

[Operator Instructions] And our first question will come from the line of Philip Shen with Roth Capital. Philip Shen - Roth Capital Partners, LLC, Research Division: I know William Rutley in Mississippi Mills, there's some lack of visibility there for Q4. I was wondering if we could explore some of the other projects. So in terms of Demorestville and Taylor, how much contribution could we see from those projects in Q4 revenues? And how much do they actually have in Q3? Michael G. Potter: They were completed in terms of the sale process in Q3 and we got just over $60 million of revenue from those projects in Q3. The bulk of the Taylor Kidd revenue will be in next year. There won't be as much construction activity and the panels will be installed on the racks next year, and that's when most of the revenue is recognized. We recognized a good portion of the Demorestville revenue already in Q3 and we'll get some in Q4. So we tend not to be specific on revenue forecasts for dollar amount. But out of the remaining 100% of the project, maybe 1/3 will be recognized this year and the rest next year. Philip Shen - Roth Capital Partners, LLC, Research Division: Okay, that's helpful. And then in terms of China, you talked about maybe 40 megawatts in Q4. Can you talk to us about the size of the early and late-stage pipelines in China? Michael G. Potter: It's a variety of sizes that ranges anywhere from 10 megawatts to 100 megawatts per project, with some projects potentially larger than 100 megawatts. These are all pretty early stage, so we tend not to get too excited until we're much more advanced in our development progress.

Yes, Phil, this is Shawn speaking. As you know, in China, the project development typically start with the MOU with the local government, then we proceed to the feasibility study and then the final permitting. Some of those in the pipelines in China. All of the pipelines in China which we mentioned already have the MOU signed with the corresponding local government, and some of them have already passed the feasibility stage and moving into the full permitting stage. At this point, as Michael said, there are projects which are the total size over 100 megawatts, and some are small. And there are some projects at 6 megawatts. Now typically, we focus on small to medium scale. Small means 4, 5 megawatts, up to a medium to about 14, 15 megawatts because the project of those size can be -- the permitting process is easier. And also, it's easy to view those projects and get permitted in Eastern China, where the great accessibility is better. Philip Shen - Roth Capital Partners, LLC, Research Division: Great. And actually, I was looking to see if we could get the aggregate size of those pipelines, early and late-stage. And correct me if I heard wrong, but does it sound like the China early-stage pipeline is 100 megawatts or is it -- did I mishear?

No. As I said, for the early stage which we signed MOU with the local government, is around 3 gigawatt. Now -- but -- however, since those are also -- still at an early stage, we cannot focus how many of those projects will move into the late stage. Now I want to emphasize on our strategy of being globally diversified. So we don't just focus on one country. Rather, we have projects spread out in different countries. So for that purpose, our late-stage project has reached over 1 gigawatt, as we mentioned in this press release. Philip Shen - Roth Capital Partners, LLC, Research Division: Great. Stepping back a bit away from projects, can we talk about what you see for 2014? I know you haven't released official '14 guidance, but I was wondering if you could comment on '14. What do you expect for shipments? And we can probably -- probably say what do you expect for projects? And then in terms of geography, where do you expect those shipments to go to?

Yes, this is Shawn again. As you said, we are not giving 2014 guidance yet. Now for module shipment, geographic wise, we see China continue to be one of the key market in terms of volume. And for Japan, we believe we'll keep the same momentum and maintain roughly the same shipment and market share. And U.S. is a growing market, so we see some growth in the U.S. market as well. In Canada, Canada is -- will be a very busy year for us since most of the -- we are -- we'll be in construction with almost all of our projects in Canada next year. And we still -- I expect a strong market in Southeast Eastern Asia and South Asian countries. The examples there are Thailand and also India. Now turning to the project business, Michael just mentioned that we are already in construction of 13 projects out of the remaining 24. And also, we have started construction on the Samsung project, the 130-megawatt Samsung project. And the -- for the remaining projects, we are at the late stage of permitting. So we expect most of the Canadian projects will reach the revenue recognition step next year in 2014. Now for the Japanese -- the Japan project, and we mentioned 275 -- 278-megawatt late-stage, some of the those projects already passed the so-called stage 3 of the utility approval. So we can start building very soon. And the schedule shows that we will have some project, some Japan project, completed next year, while the majority will happen in 2014, some in 2015. And by the way, this is the preferred way for us because in this way, we can have a managed sustainable growth on solar total solution project in 2014, 2015 and 2016. Now so I hope I answered your question, Phil. Michael G. Potter: Shawn meant to say, for the Japanese projects, some in 2014, there's a lot more in 2015 into 2016.

Yes, great. Philip Shen - Roth Capital Partners, LLC, Research Division: Great. I have one more question, I'll jump back in queue. I mean, you mentioned something interesting in your prepared remarks about the potential of a yieldco and potentially working -- or working with a yieldco. Can you expand on that? And perhaps talk about what that could look like and perhaps what the timing might be as well? Michael G. Potter: As an operating company, the cost of capital that we have, it doesn't make the greatest of sense to hold and operate projects within the current Canadian Solar structure. One of the things that some people in the market are doing today, and that we have been actually studying for quite awhile, is setting up a vehicle, a yieldco or a growth company to hold and operate projects that's separate from the main operating company. So this is something that would happen the future if it looked like it would be in the best interest for our shareholders. Philip Shen - Roth Capital Partners, LLC, Research Division: Great. So could it take the form of a potential kind of spinoff IPO for the yieldco if you decide to do it internally? Michael G. Potter: That's one possibility. Another possibility would be forming our close alliance with one or several existing yieldcos and marrying our pipeline to them to have a steady end buyer at good prices, removing some of the uncertainty of financing, and the end buyer process as well. So we have several different options in front of us and we've actually been working on this for awhile. So I just mentioned it to let people know that it's something longer-term strategy that we are looking at.

Our next question will come from the line of Colin Rusch with Northland Capital Markets. Colin W. Rusch - Northland Capital Markets, Research Division: Could we get the manufacturing margin for 3Q comparable to the 13% to 15% gross margin guidance? Michael G. Potter: I'm not sure I understand your question. You mean the module margin in Q3? Colin W. Rusch - Northland Capital Markets, Research Division: Yes. Michael G. Potter: Without the projects mixed in? Because that was a little bit over 13%. Colin W. Rusch - Northland Capital Markets, Research Division: Okay. So we're at 13% moving a little bit higher in the fourth quarter? Michael G. Potter: Basically, yes, on a very similar basis quarter-over-quarter. The guidance does not include the impact of selling any projects in Q4. Colin W. Rusch - Northland Capital Markets, Research Division: Okay. But the module margins are going from just a touch over 13% to 13% to 15%. That's what I wanted to get at. And then, can you talk a little bit about terms for modules in China in terms of payment terms? Are you getting prepayments at this point? And can you talk a little bit about trajectory and prices through the balance of the quarter and into early next year?

Colin, as you know, we are very conservative in terms of project control. So we have been very -- as a result, have been conservative in terms of sales into China market. Now moving into Q4, the -- both the price and the payment term conditions start to improve. Now in our case, most of our Q4 module agreements in China were actually assigned in Q4, in October and November. So we do benefit from the good payment term. Typically -- not typically, but all of the accounts that we signed so far have roughly 30% prepayment than some project. Some customer will pay us 100% before delivery, some will pay up to 80% or 90% around delivery. And so this is the payment term we have seen. Colin W. Rusch - Northland Capital Markets, Research Division: Okay, perfect. And then, just a couple of housekeeping questions. You've had the flexible manufacturing model for a long time. Could you just give us a sense of where you're at in terms of total capacity right now and what you might be doing in terms of thinking about CapEx for next year?

Yes. Our module capacity in China, it has been around 2.1 gigawatts. Now through debottlenecking, we actually run it at somewhere around 2.3 -- 2.2 to 2.3 gigawatt. And in Canada, we have 300-megawatt capacity. We also try to debottleneck it. So next year, we are planning -- we'll try to stretch it to over 400 megawatt. And we think we can do it without adding too much new equipment. So that's on the module side, the current capacity. And on the solar cell part. Solar cell part, our internal capacity is on -- is around 1.4 to 1.5 gigawatt. And then, we typically purchase around 400 to 500 megawatts of solar cells in Taiwan in order to fulfill the shipment request in U.S. And therefore, the solar wafer, we only have about 200 megawatts of wafer capacity. We buy the remaining of the wafers from our strategic wafer suppliers. Moving into 2014, I do see strong demand for our module products and from -- we see the growth potential, strong growth opportunity from both pure module sales and from the EPC project demand, the demand of our own EPC -- no, our own project and EPC business. So responding to that situation, we are looking to different options. And the options, including some limited expansion of owned capacity or including OEM and job [indiscernible] from other -- with other partners, we haven't made the final decision yet. The -- but we want to make a balance of capturing growth and, same time, be conservative on CapEx spending. Colin W. Rusch - Northland Capital Markets, Research Division: Okay, great. And then just a final question on allocation. How are you guys making decisions in terms of where to allocate product right now? And do you feel like you could end up increasing the volume in the fourth quarter if there is incremental demand that comes through?

Well, Q4, not much. We pretty much are running our capacity flat out. If we recognize the 2 projects in Canada, as we mentioned, then we will -- that will increase our module shipment. But those projects are already viewed. It's just that the modules asset, they are balance sheet rather than as sales because we use the fully accrued basis for the revenue recognition. Other than that, I don't see too much. Well, basically, our capacity is very much flat out and also, we are very close to the end of the year.

And our next question will come from the line of Brandon Heiken with Crédit Suisse. Brandon Heiken - Crédit Suisse AG, Research Division: So about a month ago, there was an announcement from your team that you're going to offer residential financing. Do you have any updates on your expectations on the initial interest level from that program? Michael G. Potter: It's meant more mainly for the installer base within the U.S. to help sales into residential customers. And there's been some interest -- and as we rollout the training and the marketing materials to more installers, we expect it to pick up. It's not something that we make a profit on. It's just something to help the installers sell more modules. Brandon Heiken - Crédit Suisse AG, Research Division: I see, okay. And then, the Japanese market, can you -- do you have a sense of the volume of projects that -- where they may have used Canadian Solar's, the label, for their module supplier in trying to get permitting? And what do you think that the Meti-approved projects are so much higher than installation rate? How does that bottleneck open up so that the installation rate really increases in Japan?

Those are very good questions. I only know that there are always customers who come to us out of blue and they already -- and told us that they are penciling the Canadian Solar brand name in their project applications. But I can't estimate how many. I only -- can only say that every quarter, we see a few customers come to us like that. And -- but overall, the module demand in Japan has been, for us, has been steady, just like the Japanese market itself, has been pretty stable and steady. And in terms of the project approval, there are -- there seems to be, indeed, much more METI approval than a project actually getting built. And that's why, in our press release, we made it very clear that in Japan, the project approval typically have several stages. The first is the METI approval, then 3 utility approval stats, and then the final negotiation with the landowners for the final land purchasing. So there are quite a few hurdles to jump before you can actually view the projects. Now when we give you the -- a number, the guidance of late-stage projects, those projects all have METI approval and also, well into the utility approval process.

[Operator Instructions] And our next question will come from the line of James Medvedeff with Cowen and Company. James Medvedeff - Cowen and Company, LLC, Research Division: Question, can you give us the breakdown of your module costs this quarter in the Q3? How much was silicon, how much was cell processing, how much was module processing?

Yes, this is Shawn speaking. See, we typically buy wafers rather than buy silicon. So the silicon is -- the cost of silicon in itself is not that -- we are not sensitive to the polysilicon price itself. For wafers, the typical price for the polycyclic -- for the multicrystalline wafers is around USD 0.21 per watt. And then for the cell processing, all workshop is around $0.13, $0.14, around that order. And so the cell cost is our own -- from our own self factory, is around $0.34, $0.35. Module is way -- module is below $0.20. So that's the breakdown for our own cell module productions. Now we also buy the modules from Taiwan for some -- no, sorry, we also buy solar cells from Taiwan and -- to make modules for the shipments to U.S., for example, and those solar cells are typically $0.05 higher than our internal purchased solar cell. James Medvedeff - Cowen and Company, LLC, Research Division: Okay. What is -- just the internal total is what?

The internal total solar cell capacity is around 1.4 gigawatt. James Medvedeff - Cowen and Company, LLC, Research Division: No, no, not capacity, or not cell capacity. The internal cost or total cost of module production.

Total cost of module production will be around $0.53. Michael G. Potter: Yes, we're at $0.53 end of the quarter, and the blended average during the quarter was about $0.55. And we're targeting to get it as low as $0.51 by the end of the year. James Medvedeff - Cowen and Company, LLC, Research Division: Right. So that target has not changed? Michael G. Potter: It hasn't changed, no. James Medvedeff - Cowen and Company, LLC, Research Division: When you look at your order book, I think you said you're pretty running flat out this quarter. How far does that visibility extend into next year?

We have now -- we do have our full quarter estimate for next year. We are quite confident in our continued growth in 2014. And we have been giving module shipment guidance for the past few years. Pretty much every year, we either meet or hit our guidance. James Medvedeff - Cowen and Company, LLC, Research Division: What might the mix of internal and external module sales be next year? In other words, sales to third party versus sales to your own projects business? Michael G. Potter: Well, sales to our own project business is sort of a misnomer because we sell the projects onto third parties. So this year, we were hoping to do 300-plus megawatts of project sales in terms of recognized megawatts, and we certainly would expect that to increase next year with the type of numbers that Shawn and I were talking about for Canada alone.

I'll give you some colors here. For this year, our module shipment will be somewhere over 1.7 gigawatts. All of that, probably about 1.5, 1.6 -- 1.5 gigawatts or so is from the pure module sales. I do expect that number to grow and grow quite reasonably in next year. Now for the solar module -- no, project side, it's -- you can work out as following: On the Canada side, the module will be -- we'll need about 300 megawatt modules into our own project, over 300 megawatts; then the Samsung project is 130 megawatt; and we also have a few other OEM projects. So add them all together, you will get somewhere close to 450 or so, somewhere there. And in U.S., typically, every year, we view it somewhere close to 100 megawatt. And in China, this year, we viewed 40-megawatt -- no, actually, this year we viewed around 70 megawatt. Next year, we do expect to build more in China. So that give you a rough picture of the total project business for 2014. Does that help? James Medvedeff - Cowen and Company, LLC, Research Division: Yes. Michael G. Potter: And for the shift in the metric report, that shift didn't recognize, right? So again, we could sell into our project or install in the project in Q1 and then sell the project in Q2 and the module recognition would be in Q2 when the revenue is recognized. James Medvedeff - Cowen and Company, LLC, Research Division: Okay. So in order to deliver that many -- I was kind of penciling those down, it's 600 to 650 megawatts to projects next year, do you have to shift some production away from modules that would ordinarily go to third-party or just be pure module sale?

No, we don't. We make an estimate of the demand from both sides and we try to meet those demand as long as those are good orders. Good order means reasonable margins and good credit terms. We still have some time to plan the -- our operation for next year. James Medvedeff - Cowen and Company, LLC, Research Division: And then the final question on this is, it looks like underutilization was about $0.05 a watt this quarter. Should -- may we assume that, that $0.05 per watt penalty would be not there next year, given all this opportunity? Michael G. Potter: It was about $2.6 million in total. So if you divide that into the total... James Medvedeff - Cowen and Company, LLC, Research Division: I'm sorry, I had $0.26, so it's only $0.005? Michael G. Potter: About $2.6 million, yes. James Medvedeff - Cowen and Company, LLC, Research Division: Okay, so it's only $0.005. So it's not a big factor? Michael G. Potter: No, no. I mean, it's been shrinking as the year goes on. We're utilizing more and more of our equipment.

Ladies and gentlemen, we are now at the top of the hour. I would now like to turn the conference back over to you, Shawn Qu, Chairman and CEO, for closing comments. Please proceed.

Thank you, operator, and thank you, everyone, for joining the call today, and thank you for your continued support. If you have any further follow-up questions after today's call, please contact us. Have a great day.

Thank you for your participation in today's conference. This concludes the presentation. You may all disconnect. Good day, everyone.