CSIQ 阿特斯太阳能

Operator

Ladies and gentlemen, thank you for standing by. Welcome to Canadian Solar’s Fourth Quarter and Full Year 2015 Earnings Conference Call. My name is Sheila and I will be your operator for today. [Operator Instructions] As a reminder, this conference is being recorded for replay purposes. And now, I would like to turn the call over to Ed Job, Director of IR. Please go ahead.

Ed Job

Thank you, Sheila and welcome everyone to Canadian Solar’s fourth quarter and full year 2015 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 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 current expectations and therefore we refer you to a more detailed discussion of the risks and uncertainties in the company’s Annual Report on 20-F filed with the Securities and Exchange Commission. In addition, any projections as to the company’s future performance represent management’s estimates as of today, March 10, 2016. 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.

Shawn Qu

Thank you, Ed and thank you all for joining us on the call today. We are very pleased with our results for the quarter and the full year. Our record performance reflects solid execution of our business plan at both the module and solar energy business units. We have delivered solid results, while also making progress on key elements of our strategy and positioning the company to create lasting value for our shareholders in the years ahead. Our notable record highlights for 2015 include record revenue of $3.47 billion; record module shipments totaling 4.7 gigawatts; record cash flow from operations totaling $397 million; record 568 megawatts of solar power plants built and connected to the grid in 2015 alone; and also record 10.3 gigawatts of solar project pipeline, including 2 gigawatts in late stage of development and construction as we exit the year. In addition to reaching all of these milestones, we delivered net income of $172 million for the full year, equivalent to $2.93 per diluted share, while retaining close to 400 megawatts of our premium solar power plants with a retail value exceeding $315 million on our balance sheet. These accomplishments serve to reinforce our position as a global leader in solar energy industry. Moving on to the Q4 performance, our revenue and gross margins again exceeded our guidance reflecting strong demand for PV modules and stable ASPs, the sales of three solar power plants in Canada totaling 42.9 megawatts as well as the partial sales of our Roserock and Garland solar power project to Southern Company in U.S. Our module sales, was again led by strong demand in U.S., China, India and Japan with each of these markets making a significant contribution to total shipments. Shipments to U.S., China, in particular, exceeded 300 megawatts in the quarter respectively, while shipment to India exceeded 200 megawatts. Our energy business also delivered positive results. In Canada, we completed the sales of three solar power plants totaling 42.9 megawatts and valued at over CAD197 million. In addition, we started commercial operation of our [indiscernible] project, which we currently own and operate. In the U.S., in addition to our solar project sales to Southern Company, we also secured tax and tax equity commitment to support the construction of our other utility scale solar power plants, which will all be connected to the grid during 2016. In Japan, we energized three solar power plants totaling 6.2 megawatts, bringing total solar power plants in operation under the Japanese FIT program to 21 megawatts. In addition to sales and partial sales of certain solar power plants – operating solar power plants and our ownership now totals approximately 400 megawatts, up from 257 megawatts at the end of Q3. The total retail value of these 400 megawatt power plants is estimated at over $815 million, with expected resale gross profit of over $170 million. Meanwhile, our pipeline of late-stage solar project remains large at approximately 2.0 gigawatts. We estimated the resale value of our late-stage solar project pipeline at over $4.5 billion, with expected gross profit contribution in excess of $850 million. Now, let me comment on our guidance of Q1 and full year 2016. As we enter 2016, we expect to continue to grow both our annual module shipment volume and our downstream solar power plant business. We currently expect total Q1 module shipment to be in the range of approximately 1.085 to 1.135 gigawatts, including 50 megawatt of shipments to our own utility-scale solar project. Revenue for the first quarter of 2016 is expected to be in the range of $645 million to $695 million, with gross margin expected to be between 12% to 14%. As we do not expect to sell or partially sell any of our solar power plants, the revenue in Q1 mainly comes from the sales of solar modules and the electricity income from the operating solar power plants under our ownership. For the full year 2016, we expect total module shipment of 5.4 to 5.5 gigawatts with approximately 5 gigawatts to the third-party customers and therefore recognized into revenue. We expect the total revenue for the year to be in the range of $2.9 billion to $3.1 billion. With the above guidance, we plan to grow our annual megawatt module shipment by 15% to 17% from the 2015 level. However, different from some of our competitors, our focus is really on the upgrade of our technology and improvement of our cost structure through selected capacity investment, especially in the midstream solar cell part of the value chain. In particular, we plan to aggressively adopt monocrystalline PERC technology, multicrystalline black silicon technology and diamond wire cell in our production in the next 2 years as these technologies will allow us to significantly increase the sales and module upload while significantly reduce the production cost at the same time. We also plan to start production at a new manufacturing site in Southeastern Asia in Q3 of 2016, with nominal cell and module capacity at 700 megawatts and 500 megawatts, respectively. These efforts once completed will better prepare us for future competitions. Our revenue guidance for 2016 does not include the potential sales of solar power plants that we currently plan to own and operate in OECD markets, which we expect to reach 1.1 gigawatt by the end of 2016. We estimate the resale value of these assets at approximately $2.5 billion, with gross profit contribution of approximately $355 million. We will remain flexible on how to monetize these assets in order to maximize shareholder return and may consider selling some of these OECD solar power plants, in which case the revenue for the full year is expected to be in the range of $3.2 billion to $3.6 billion, an increase of $300 million to $500 million over our base forecast. Meanwhile, we estimate the electricity revenue for these solar power plants on an annualized run-rate basis to reach approximately $160 million to $170 million by the end of 2016. We have been preparing for the launch of YieldCo with some of our high-quality solar power assets if the market condition allows. However, we remain flexible and are actively considering several other options to monetize our solar power plant assets. These options include, for example, sales of solar power plants, our asset-backed securitization. We believe that our solar power plant assets in low risk OECD countries are valuable and highly liquidable. Our goal is to maximize the long-term return to our shareholders. Let me now turn the call over to our CFO, Michael Potter, for a more detailed review of our results for the fourth quarter. Michael, please go ahead.

Michael Potter

Thank you, Shawn. Net revenue for the fourth quarter of 2015 was $1.1 billion, up 31.8% sequentially and up 17.2% compared to the year ago period. Q4 revenue came in well above our guidance as a result of better-than-expected pricing environment as well as the partial sale of the Roserock and Garland projects. Gross profit in Q4 was $200.5 million compared to $126.8 million in Q3. Gross margin in Q4 was 17.9% compared to 14.9% in Q3. The sequential increase in gross margin is in part due to stronger-than-expected shipment volume, stable ASPs and strong gross margin contribution from the U.S. project sales. Total operating expenses were $95.2 million in Q4 compared to $95.9 million in Q3. Income from operations was $105.3 million in Q4 compared to $30.9 million in Q3. Operating margin was 9.4% in Q4 compared to 3.6% in Q3. Net foreign exchange gain in Q4 was $10.4 million compared to net foreign exchange gain of $4.8 million in Q3. We recorded a loss from fair value of warrants of $8.9 million in the fourth quarter of 2015. The warrants were issued in conjunction with the $180 million in financing arranged by Credit Suisse in Q4 of 2015. These warrants can be settled in cash at the discretion of the holder. And as a result, they are recorded as a derivatives liability with a fair value to issuance and will be subsequently mark-to-market at the end of reporting period. As a housekeeping item, there is approximately a $1 million change in value of these warrants with a $1 change in our share price. Income tax expense in Q4 was $31 million compared to income tax benefit of $3.9 million in Q3. Net income attributable to Canadian Solar shareholders for Q4 was $62.3 million or $1.05 per diluted share compared to a net income of $30.4 million or $0.53 per diluted share in Q3. Moving on to the balance sheet, in Q4, cash and cash equivalents were $553.1 million compared to $345.8 million at the end of Q3. The restricted cash balance was $581.6 million at the end of Q4 compared to $656.2 million at the end of Q3. Our trade accounts receivable balance was $426.8 million at the end of Q4, up from 431 – down at the end of Q3. Inventories decreased to $334.5 million at the end of Q4 compared to $426.4 million at the end of Q3. Short-term borrowings at the end of Q4 totaled $1.2 billion compared to $1.1 billion at the end of Q3. Long-term debt at the end of Q4 was $606.6 million compared to $514.3 million at the end of Q3. Senior convertible notes outstanding totaled $150 million at the end of Q4. Short-term borrowings and long-term debt directly related to utility-scale solar power projects totaled $560.6 million at the end of Q4. As this is both the year end summary and the introduction of our outlook for 2016, I want to touch on a few topics before we open the call for questions. The first topic is our ability to finance our plans. We have been very careful not to over-commit and keep within our financial means. We are able to access the global financing markets and our close financings as we have planned and as we have needed over the past several years. We are deliberately keeping our equity requirements down when we build projects and using amortizing debt and higher levels of debt at the project level. We maintain flexibility, so we could reduce that debt with more equity if equity costs are reasonable, but our goal is profitably balanced capital stacks, not short-term cash flow metrics for advertising purposes. We also recently closed an asset-based security structure for our project in Japan that financed close to 85% of the value at a 1.4% coupon rate. This is repeatable for other Japanese projects. It is an example of how we can recycle capital without selling projects. I can say that we have good confidence that we have financing available for both the capacity expansion we plan and the project pipeline to be built in 2016 and beyond. The liquidity pool is very deep if you are a strong company with good projects. I believe that the plan that Shawn laid out in his remarks are well supported by our financing ability and the ability of the market to provide liquidity. I expect our absolute debt level will continue to rise, especially as we get close to the completion of our USA projects. This debt is all non-recourse and fully supported by the expected cash flow of the project. At the same time, we are using our cash and other financings to pay down higher interest debt or to lengthen the maturity of shorter-term debt. As an example, so far we have bought from the market and retired $15 million of our convertible bonds in 2016, the bonds trading below par and represented a good opportunity to efficiently reduce debt. We plan on paying off or reducing several other loans in Q1 and we will continue to do so in the future. We plan on slowly reducing the non-project debt burden over time. You can be assured that our management does not plan to overreach and that we have the discipline to effectively manage our balance sheet. If the cost of equity is reasonable, we are willing to raise some equity financing as well. The recent ATM re-launch is an example of that. These are more opportunistic financings and we suspended the program when the share price moved below our target price. Up to that date, we had issued 500,000 common shares at the weighted average price of $27.73, with net proceeds of $13.6 million after discounts and commissions but before operating expenses. Unless we believe that market conditions are very favorable, this program will remain suspended. The second topic is the status of our YieldCo. As you know from our press release several months ago, we have done a confidential filing of an S-1 with the SEC. The initial filing has been reviewed and we have answered the questions and have received some further clarifying questions. Based on my experience, I would say this process is fairly advanced, but I also know from experience, it’s not done until the SEC accepts the filing. So I can say the mechanical process is going as I would have expected. The mechanical process is one thing, but market reality is another. The market in general has not performed well in recent times. In addition, several yield cosponsors are in some distress, and we are fully aware that YieldCo valuations are challenged now as is the IPO market as a whole. The good news is that we can build our pipeline regardless of the status of the YieldCo market. As indicated in our outlook for 2016, if the market remains challenged, we will sell some projects and recycle the cash into our pipeline. With the recent extension of the ITC in the USA, we have years of development ahead of us and our global development capability is growing stronger than ever before. We expect that we can both maintain the option of a YieldCo and sell projects at a profit. We continue to believe that the best long-term value for our shareholders is to hold projects and that a lower cost of capital vehicle provided by such as the YieldCo is needed to execute on that strategy. The third topic is our ability to control costs and execute on both on our module and energy businesses. Over the last 2 years, we have been studying the various trade cases and also developing and evaluating technology for our module business. We reached a point where we have started our expansion both inside and outside of China. There are concerns about overcapacity in the module business being a risk later this year, but it’s important to understand that our main investments are closing the gap between our module capacity and our wafer and cell capacity. We expect this to improve margins and we will still have less wafer and cell capacity than module capacity, so we should be able to keep them full. I am close to hitting my 5-year anniversary as CFO here and I have 9 years associated with the company. Every year, I get more confident of our ability to succeed even with all the challenges the industry faces. This is because our global team, including newer team member such as the former recurrent staff are very good and can find opportunities to overcome and exceed the challenges in front of us. With that, I would like to stop for our prepared remarks and turn the call over for questions. Operator, please go ahead.

Operator

Thank you. [Operator Instructions] And your first question comes from the line of Philip Shen of ROTH Capital Partners. Please proceed.

Philip Shen

Hi, everyone. Thank you for taking my questions.

Shawn Qu

Hi.

Philip Shen

Hi, Shawn. In your release I think you mentioned that some of your late-stage projects may not receive the appropriate permits, it may not reach completion, can you quantify for us how many megawatts that might be for the year?

Michael Potter

I believe it’s meant to be more for the earlier stage projects and that’s the just saying that if we have a 10 gigawatt pipeline, we are cautioning people that you don’t necessarily reach permits and complete everything that’s in your pipeline.

Philip Shen

Okay. I think you guys mentioned it for late-stage...?

Shawn Qu

No, we didn’t mention. Mike do you want…

Michael Potter

Yes. The late stage, we have all of the permits and they are in construction, so we are not worried about anything that’s in construction or about to enter construction.

Philip Shen

Great, okay, just wanted to clarify. In terms of ABS, can you provide some additional color on your views there, especially from markets outside of Japan and for example, could we see efforts to issue an ABS here in the U.S.?

Michael Potter

It’s possible we could do it in the U.S. market. It’s more effective and efficient in Japan right now because of the interest rate environment. But we do have several financings that have been proposed to us for the U.S. market as well. There are certainly ways to increase the loan to value financing, particularly on a portfolio effect in markets like the U.S. and that would allow us to recycle some capital back to the company. I think our primary strategy in the U.S. would be focused on selling projects if the YieldCo market remains closed rather than doing an ABS link strategy. But we could also do ABS in the U.S.

Philip Shen

Okay, great. One more for me here, in terms of your Southeast Asian facility, it looks like you guys increased capacity goals there from 400 megawatts to 700 megawatts in cell and then from 400 megawatts to 500 megawatts in module. Just talk just about the rationale there, I am guessing it’s to target the U.S., but if you can speak to that expansion there, that would be helpful?

Shawn Qu

Yes. The solar cell, I believe we always targeted 700 megawatt for the baseline. As for the module side, 400 megawatt to 500 megawatt, that’s a little different, it’s only a little bit technical. Technically we measure the space and we realized we can put in a little bit more equipment. And so once we finish that project, we will have a nominal 700 megawatt of solar cell production in Southeastern Asia. And remember, we already have a 300 megawatt module plant in Vietnam. So our cell and module capacity in Southeastern Asia will be at 700 megawatts and 800 megawatts. So it’s pretty balanced internally.

Philip Shen

Great. Thank you, Shawn and Michael.

Michael Potter

Thank you.

Operator

And your next question comes from the line of Colin Rusch of Oppenheimer. Please proceed.

Colin Rusch

Thanks so much guys. Can you talk a little bit about the transition from the assets that are in your pipeline into backlog and how that pacing is working now that we have got some clarity on the ITC in the U.S. and how should we think about this 2 gigawatt number getting replenished as we go forward?

Michael Potter

So the 2 gigawatt number is actually somewhat of a lower estimate for the possibilities we could do in the U.S. That’s shorter-term projects that we have good visibility of, that with the ITC extension, we certainly got easier ability to monetize those more quickly. Over the entire 5-year extension period and then the commenced construction period afterwards, we believe we will do better than that and they will follow this significantly better than that.

Colin Rusch

Okay. And then as you think about your options in terms of the financing, can you just give us some metrics in terms of what you are looking at. And then also if you could just comment on this $160 million to $170 million in sales from electricity, that run rate, at the end of the year, how that would breakdown into cash flow metrics and so that we can have a sense of what the options are for what’s the actual cash flow and what your options are for monetizing them?

Shawn Qu

Yes. The electricity revenue is an estimate and it’s just based on the OECD projects. If we added in all the projects we are operating, it would be slightly higher than that as well. For the U.S. projects, most of them are tax equity financed and we are planning and using amortizing debt. So, the actual cash flow produced from the U.S. projects for the first few years is lower. Canadian projects, UK projects and Japanese projects, the cash flow per dollar of electricity revenue is a lot higher. So, we have said in the past, we believe that we will have between $50 million to $60 million with the captive for the YieldCo and the YieldCo is made up of our OECD projects. So, that should give you a rough indication of what we think we could produce if we didn’t sell any of those projects.

Colin Rusch

Okay. I will hop back into queue and take a few things offline. Thanks.

Michael Potter

Okay.

Operator

And your next question comes from the line of Paul Coster of JPMorgan. Please proceed.

Paul Coster

Yes, thanks for taking my questions. A few quick ones. I mean, the first one is a little bit rhetorical, but why wait on the decision around the YieldCo, why not postpone it with for definite period of time right now? I mean, the stock is down a few percentage points this morning presumably because the guidance continues to withhold the revenues associated with that strategy?

Michael Potter

I would say the answer to that is that the market has suspended any YieldCo plans for a launch. There has not even been any IPOs worth mentioning outside of the solar space. So, we have always made sure we had a plan B and plan C available and we are certainly executing on those. So, I mean, we have the optionality that the market window does open up, the SEC process would have moved along. And if for some reason, it does open up, we can go on it. If not, we have the ability to sell projects, recycle capital and our pipeline is big enough, we can just replenish that and keep the option available for the future if it does get better.

Paul Coster

Okay. So we should just assume this is going to be a continuous state of affairs now where we have got this steady state held-back portfolio?

Michael Potter

Unless for some reason it’s conclusive that some form of securitization like a YieldCo is just not in the best interest of the shareholders, in which case we will probably go to selling the projects.

Paul Coster

Okay. It sounds like you are investing in monocrystalline PERC capacity. Can you just talk to us a little bit more about that? What end markets is it destined for? Why choose that particular technology? And what kind of cell efficiency are we talking about in commercial volume?

Shawn Qu

Yes, Paul. This is Shawn speaking. For the mono PERC, at this moment, our cell efficiency is approaching 21% and we believe that while as we continue the learning process, eventually the cell efficiency using mono PERC can go over 21%. And in a few years, if we are looking towards 2020, there is a possibility for the mono PERC to go to 22% to 23% cell efficiency once we combine that with other technology, for example, the five busbar technology or even more busbars and also the improvement of the solar wafers. So, there is a significant output increase, output improvement within the mono PERC. And the cost of mono wafer has declined significantly in the past 6 months. We also see more capacity of the mono wafers coming online, especially with the wide adoption of the diamond wire cell technologies. And for the market, talking about market, we will probably launch our mono PERC-based modules to some of the high premium market first, while albeit it will be the residential market in Japan. And I think the mono PERC product will also have an interesting market in China as the Chinese government has been promoting and promoting and permitting some solar FIT in China with high-efficiency technologies.

Paul Coster

Okay. Last question, can you just talk a little bit about the performance of your China portfolio, is it cash collections meeting your expectations?

Shawn Qu

As you will see that, we remain conservative in our own – in the developing and operating our own solar power plant in China. And our power plants, in general, don’t have the curtailment issue other than one we have one project in Xinjiang Province, that one does have some curtailment issue, other projects in the Inland and East Coast. So, we don’t have curtailment issue and all the projects still facing the delay on the feed-in tariff payment from the Ministry of Finance. And so but because of that, we will continue to be conservative in building and owning solar project in China. However, the overall solar project market in China continued to be strong. So, we still expect to sell high-efficiency solar modules to third-party customers in China.

Michael Potter

Ironically, the one project we got early approval of the FIT is the one in Xinjiang which is being curtailed.

Paul Coster

Thank you.

Operator

[Operator Instructions] And the next question comes from the line of Sheng Zhong of Morgan Stanley. Please proceed.

Sheng Zhong

Thank you for taking my questions. Just I first want to follow-up with our technology, can you please give a little more color on your detailed plan on the capacity ramp up of the new technology like the PERC how much capacity will you have in 2016?

Shawn Qu

Well, we haven’t released that specific number yet, but we do expect to increase our PERC capacity fast. Right now, we have – we are working on a one line, one PERC line in our capacity. And by the middle of the year, we expect to add more capacity.

Sheng Zhong

I see. Thanks. And my next question is about our projects profit margin, I see the guidance in 2016, the resale of project gross margin is about $355 million and that translate to around ‘14 gross margin. But if we look our 2 gigawatt total pipeline I think the margin, resale margin could be much higher. So, just want to give a little more sense on our look of the market if that – does this mean that the project margin will improve in future years?

Michael Potter

I think that the stuff that’s operating now is actually higher value projects in Japan and Canada and the UK. And if we sold those, we are likely to get a higher gross profit – gross margin percentage. The U.S. projects, although profitable, are lower. We estimate between 10% to 15% gross margin for those projects. So, as those projects come online and get included in the bigger metric we talked about at the end of this year for the OECD operating projects, the total gross margin percent would go down. Globally, the margins you can realize on projects are somewhere between the 10% to 15% range in most developed countries right now, particularly newer projects with newer PPA. So, it’s been more of an equalization and not quite such a dramatic drop in the cost as there was in prior years which led to higher margins.

Sheng Zhong

Very helpful. And my last question is about our project sales target, you mentioned that the YieldCo plan and you will sell some projects. But if I look at your 2016 full year revenue guidance and if I minus the module income, that indicate that your project sales revenue – project sales revenue is not very high. So this is because you are conservative on your revenue guidance or because you want to maintain more on your balance sheet?

Shawn Qu

Our base guidance of $2.9 billion to $3.1 billion sales does not include much of the project sales, as you said. So you are correct. They include some China sales, but not many OECD country project sales. We did our cash flow planning. We believe that because our cash flow in the module business is expected to be healthy. And also therefore, all together we may only have to securitize or monetize $200 million to $300 million additional project value one way or another. That’s why our guidance we mentioned that if we choose to monetize, to recycle this much cash through project sales, we may increase the project, increase the total revenue by $300 million to $500 million. And we put these numbers simply because, at this moment it seems like we only need to recycle this much cash for 2016. And overall, we believe that because the low interest environment is probably going to sustain in most of the countries in the world, therefore our high quality OECD country project has very good value and that value may increase as time goes.

Michael Potter

I would just say, Japan has negative interest rates. They just lowered interest rates and increased stimulus in Europe. And there are certainly lots of questions around the U.S. fed if they are still going to be as aggressive as they said last year for any interest rate raises. So the financing market for long-term high quality cash producing assets like solar projects is quite good right now.

Sheng Zhong

Thank you very much. That’s all for me. Thanks.

Operator

And your next question comes from the line of Jeff Osborne of Cowen & Company. Please proceed. Jeff please check your phone is not on mute and go ahead, you are in the call. Okay. So we will move on to next question, which comes from the line of Carter Driscoll of FBR. Please proceed.

Carter Driscoll

Good morning. Just to clarify, so in terms of the timing of some of the projects that you may potentially sell, I mean is it fair to say that it’s directly related to the plans or the cooperation of the equity markets in terms of when you may or may not do a YieldCo, I mean is that the way we should be thinking about it, if it looks like the capital markets aren’t cooperating, we could anticipate maybe increasing sales on a quarterly basis or maybe just giving your additional thoughts on how the two are not directly related?

Michael Potter

Yes. It would likely be in the second half of the year. And I don’t think most people expect the capital markets to improve, particularly dramatically, very quickly this year. So we are working on selling projects is one of our main plans. We have several other alternatives such as asset backed securitization or other forms of recycling capital. So the $300 million to $500 million would be sort of a second half of the year indication and it really depends on people’s feelings of what’s going to happen to the capital market then.

Carter Driscoll

Got it. Okay.

Michael Potter

If the Capital Market remains closed to do a YieldCo IPO, we will be selling projects.

Carter Driscoll

Got it, okay. So it’s between your base and your extended guidance is really where we should be thinking if we have a viewpoint that the YieldCo is not going to open up in two halve ’16, is that fair to say?

Michael Potter

Yes. And we are actively working to sell these projects now. It’s not like we think we will work on it later. I mean, we are actually doing what we need to do, so when we get there, we can sell them relatively quickly.

Carter Driscoll

Understood.

Michael Potter

And just some PowerPoint plan, we are actually executing on.

Carter Driscoll

Understood. And then just maybe a quick follow-up, so obviously some of your competitors, you directly addressed it in your prepared remarks, talked about module overcapacity in the second half of the year. Obviously, you guys feel confident because of the internal mismatch that shouldn’t be a problem. But one of your other competitors talked about some project push-out related I am assuming directly to the ITC expansion as people were not quite in a rush. Can you maybe characterize what you are seeing maybe for the second half of the year, is that a fair characterization, you are not seeing quite the rush you did before. And then, would that be directly related to what maybe your competitor was talking about in terms of module pressure?

Shawn Qu

We think, as a whole, 2016 is still going to be a strong year. Now for 2017, as some of the third party market research firm indicated, there may be a slowdown, some kind of slowdown in the U.S. market, although the ITC expansion has resulted in some project being pushed into 2017. That would probably help to smooth out this year-to-year change a little bit. And so when we do the 2016 planning, we can certainly sell more modules. We guided 5.4 gigawatt to 5.5 gigawatt on module shipment. And I guess if we really want to sell more than 6 gigawatt as some of the competitors, we can do that. However, I don’t think that should be our focus. So our focus for 2016 is to maintain a reasonable and steady growth on the shipment. That growth will see us grow from the 4.7 gigawatt of 2015 shipment to 5.4 gigawatt to 5.5 gigawatt, so about 15% to 17% of the shipment growth. Now meanwhile, we want to emphasize, we want to put focus on the upgrade of the technology and also some expansion of our cell and wafer internal capacity rather than focus too much on the module capacity. And I believe by doing so, we can – like patent immune us from any market change. So basically, we are not that aggressive in the total module shipment growth, but would rather put an emphasize to increase our internal wafer and cell capacities to increase the level of internal vertical integration. Therefore, we will have a better margin structure. So even if some of the overcapacity does happen at some of other solar peer company indicated, our strategy will make us better preparing for that than our competitors.

Carter Driscoll

And just lastly, in terms of the timing, all of the expansions are in the same timeframe as you discussed with this last quarter, is that fair, has any thing slipped or been pulled in?

Shawn Qu

Well, it’s very much on schedule.

Carter Driscoll

Got it. Thank you. I will get back in the queue, I appreciate your time.

Shawn Qu

Thank you.

Operator

[Operator Instructions] And the next question comes from the line of Paul Coster of JPMorgan. Please proceed.

Paul Coster

Yes, I am sorry to jump back on, but Mike I didn’t really understand the answer to the question that was posted by the Morgan Stanley analyst about the late-stage pipeline, which seems to have a sort of 19% gross margin and the held-back projects which have a 14% gross margin. Why is it that the margins are increasing again? I didn’t understand the answer.

Michael Potter

The projects that are in operation at the end of 2015 are Canadian projects, Japanese projects and UK projects. Those projects are higher gross margin on average than the U.S. projects. When you get to the total number of over 1 gigawatt, which is used to represent the projects we are holding that includes the U.S. projects and those have lower gross margin percent and they would bring the overall margin down.

Paul Coster

Got it. But why does the margin then jump up to the 2 gigawatts?

Shawn Qu

Yes. Paul, our current holding project do include significant portfolio in China. And because of the feed-in tariff, the delays on feed-in tariff payment, we expect a lower retail margin for those projects. Out of about 400 megawatts of current operating projects, over 200 megawatts are in China. And these ones we assume a lower retail margin. And in the new pipeline, pipeline to be added in this year, we have the project from Japan and also from UK. This project have higher gross margin. Now, meanwhile, I would also have some sub-project in U.S. The U.S. project doesn’t have very high gross margins. The U.S. project – retail gross margin will be 10% to 15%, but the Japan project and UK project do have a higher retail gross margin. So, then in average the retail gross margin for the future pipeline is higher than what we are holding right now.

Paul Coster

Right, thank you.

Shawn Qu

Did I answer your question?

Paul Coster

Not really. I am so sorry to labor the point, but say $150 million of gross margin on $4.5 billion associated with that $2 billion of late-stage – 2 gigawatts of late-stage pipeline, that’s 19% gross margin. It doesn’t – the narrative doesn’t looking forward then…

Michael Potter

If I just take a look at Japan, it’s 580 megawatts there. So, if I look at the numbers on the press release, it says that the retail value of the operating power plant is exceeding 20%. They are currently operating power plants have a lot of Japanese and Canadian and UK and no U.S. As a percent, it goes down in the grand total number, but we do have more Japanese projects when you build all of them out for the end of 2020 I believe is what our chart says. If you include all of the Japanese projects and we don’t have COD for any U.S. projects past 2016 in this number. So, you can calculate a higher gross margin on that because of the high Japanese contribution. If you start adding in U.S. projects, our gross margin percent is going to go down.

Paul Coster

Hey, just one last question, I am so sorry. You are looking at growth for both module and the power plant business this year, is there a relative emphasis on one versus the other?

Michael Potter

I am sorry I want to make sure I understand your question. Are you talking about growth...

Paul Coster

You are looking for growth in modules and the module business and in the power plant business this year, is there a relative emphasis on one versus the other moving forward?

Michael Potter

I think the emphasis on the module business is steady growth and augmenting the technology and increasing the investment in wafer and cell capacity. We are falling well below 50% cell capacity compared to module capacity in our module business and that’s mainly been because we have been waiting until we had the technology roadmap ready and able to go into mass production. And now we are starting to invest into that. So, the energy business, we certainly are going to be investing a lot into growing our U.S. pipeline. It won’t be a 2016 item, but ‘17 and onwards it will be quite a big growth in our U.S. pipeline as the development efforts start bringing that into closer focus in the PPA start lining up. Japan, we are going to be focusing on executing on our current pipeline. And for most other regions of the world, it will be built in cell. For example, Brazil, there will be construction activity in ‘16, but most of the selling will happen in ‘17.

Paul Coster

Okay, thank you.

Operator

I would now like to turn the call over to Canadian Solar’s Chairman and CEO, Dr. Shawn Qu for closing remarks.

Shawn Qu

Thank you, operator and thank you everyone for joining the call today and thank you for your continued support. In summary, 2015 is a year of record for Canadian Solar. We remain a leader in the global energy business by both revenue, profit and megawatt shipment. Because of our consistent focus on execution of our strategy, we are positive on our prospects for 2016 and our ability to build long-term value for our company and shareholders. If you have any further follow-up questions after today’s call, please contact us. Have a great day.