第一太陽能 (FSLR) 2016 Q2 法說會逐字稿

Good afternoon, everyone, and welcome to First Solar's second quarter 2016 earnings call. This call is being webcast live on the investors section of First Solar's website at firstsolar.com. At this time all participants are in listen-only mode. As a reminder, today's call is being recorded.

I would now like to turn the call over to Steve Haymore from First Solar investor relations. Mr. Haymore, you may begin.

Thank you. Good afternoon, everyone, and thank you for joining us. Today the Company issued a press release announcing its financial results for the second quarter of 2016. A copy of the press release and the presentation are available on the investors section of First Solar's website at firstsolar.com.

With me today are Mark Widmar, Chief Executive Officer; and Alex Bradley, interim Chief Financial Officer. Mark will provide a business and technology update, then Alex will discuss our second-quarter financial results and provide updated guidance for 2016. We will then open up the call for questions.

Most of the financial numbers reported and discussed on today's call are based on US generally accepted accounting principles. In the few cases where we report non-GAAP measures such as free cash flow or non-GAAP earnings per share, we have reconciled the non-GAAP measures to GAAP measures at the back of our presentation.

Please note this call will include forward-looking statements that involve risks and uncertainties that could cause actual results to differ materially from management's current expectations. We encourage you to review the Safe Harbor statements contained in today's press release and presentations for a more complete description.

It is now my pleasure to introduce Mark Widmar, Chief Executive Officer. Mark?

Thanks, Steve. Good afternoon and thank you for joining us today. Before reviewing the results for the quarter, I want to provide a brief update on some key strategic priorities and how I intend to lead the organization in my new role as CEO.

In our decision-making we will continue to apply the same disciplined approach to long-term strategy that has served us extremely well to date. At various times over the past several years, some have questioned our measured approach to key strategic decisions, including capacity expansion, asset ownership and balance sheet utilization. However, we continue to see the benefits of our disciplined strategy and are confident this approach will optimize shareholder value over the long term. This approach also allows us to be nimble and navigate an industry that is cyclical, high-growth and dynamic.

The updates I will provide today do not constitute any major strategic changes from what we discussed earlier this year at our Analyst Day in April but are further refinements aimed at simplifying and focusing our business. The core of our overall strategy continues to be our differentiated CadTel technology and its tremendous potential for further conversion efficiency improvements.

Closely aligned to our module efficiency roadmap is the evolution of our module form factor, which as we indicated in April will begin next year with the introduction of our Series 5 module. Our Series 4 technology is an outstanding product. With a superior temperature coefficient, spectral gain and shading response, Series 4 holds an energy density advantage over multi-crystalline silicon products. Series 5 retains the energy density advantage of our CadTel technology, further boosts efficiency and is optimized for cost-effective installations. To put this into perspective, in a typical market, the reduction in Series 5 BOS cost relative to Series 4 would be equivalent to boosting module efficiency by approximately 100 basis points. The combination of these factors make this a powerful new product, which is supported by positive initial feedback from our customers and partners.

Given the exciting potential of this product, we are now planning to invest in additional efficiency improvements over the next couple of years that will take the Series 5 roadmap to a 390-watt module versus the 365-watt module we introduced at Analyst Day. The competitive position of a Series 5 390-watt module with a similar form factor to crystalline silicon, but with the inherent energy advantage of CadTel, is very exciting. Moreover, the substantial increase to the Series 5 roadmap has an extremely compelling return on investment. The estimated $60 million of capital expenditures that will be required to enable this efficiency gain across our current 3 gigawatts of production will result in anticipated payback time of less than a year.

So focusing on increasing Series 5 efficiency does not change the priority we are placing on Series 6. We continue to push the pace of Series 6 development as fast as possible, with pilot plant production beginning in 2018. It should be noted that we intend to prioritize long-term investment in the production and development of Series 5 and Series 6 technologies over adding new manufacturing capacity in the short term. In light of the current supply/demand dynamics in the global solar market, we will continue to evaluate the point in time at which adding additional capacity makes sense.

As we eagerly look forward to these expanding technologies, some strategic realignments are necessary to ensure focused and seamless execution. Firstly, as we recently announced, we have decided to discontinue TetraSun production in order to maximize our Series 5 assembly capabilities at our facility in Malaysia. This not only enables a more rapid rollout of Series 5 product next year, but also allows us to focus more resources and capital on our CadTel roadmap.

Secondly, the new Series 5 form factor will significantly broaden the spectrum of third-party structures and tracker technologies, available for use with our modules, without requiring First Solar specific components. Previously, the different form factor of our Series 4 technology led us to develop optimized in-house solutions. The new interoperability of Series 5, with the broad global ecosystem of existing third-party structured products, is expected to facilitate increased adoption of our module by new customers and increase market opportunities.

The evolution of our module form factor allows us to tap into a broader and deeper supply chain for structures and trackers and we have therefore decided to discontinue substantial internal development of fixed-tilt and tracker structures. This decision allows us to refocus resources towards developing closer relationships with industry-leading structure and tracker technology providers and at the same time eliminate some of the channel conflicts that existed in certain situations today.

Thirdly, the progression of the interoperability of our modules also impacts our strategy as it relates to EPC capabilities. Historically, we have self-performed EPC on the majority of our systems projects as well as provided EPC services to third-party developers using First Solar module technology. The evolution of our module form factor. combined with the lower volume of system projects in 2017, implies a reduced need for internal EPC services and requires a reduction in our EPC workforce. While these decisions are difficult and will impact a number of dedicated associates, this is a necessary step.

We will continue to maintain sufficient internal EPC resources to support the construction of our self-developed projects. This allows us to continue to provide comprehensive solutions to our customers and also enables continued development of next-generation BOS technologies such as MVDC, which we introduced at our Analyst Day. It should be clear that this decision around EPC in no way impacts our project development business, and in fact allows us to allocate more capital and focus more resources on this critical aspect of our business model.

Lastly, as part of our refocusing efforts we have also decided to explore options for the disposition of skytron energy. Our original intent in acquiring skytron was to expand our global O&M footprint into Europe and supplement our existing capabilities. Over the past two years competition for O&M contracts in Europe, combined with the progression of First Solar's internal O&M capability, particularly as it relates to smaller-scale solar plants, has changed how skytron fits into our strategy going forward.

All together we expect charges of $105 million to $120 million related to TetraSun and the actions we are announcing today. On an annual basis going forward, the total reductions in operating expense and cost of sales is expected to be $60 million to $80 million. Alex will provide more detail on these charges and associated savings.

While these decisions are difficult for part of our organization, these actions will best position First Solar to increase its competitive position over the next several years. We are excited about the potential of our Series 5 and Series 6 technologies and are well positioned with the resources and capabilities necessary to execute on our plan.

With that let's turn to our results. Our financial results for the quarter were strong with net sales of over $930 million and non-GAAP earnings per share, excluding restructuring charges, of $0.87. These results were underpinned by continuing execution across both our module manufacturing and system project portfolio.

Our module lead-line efficiency is currently running at 16.7%, and we continue to lower our cost per watt. We completed our Silver State South and McCoy projects in Q2 and are nearing completion of the Stateline project.

We recognized significance cost savings on these projects in the quarter as a result of our outstanding execution. Alex will discuss the financial results for the quarter in more detail.

Moving on to slide 5, I'll provide an update on our latest bookings activity. In the second quarter we shipped close to 715 megawatts of modules, bringing our year-to-date total as of June 30 to 1.6 gigawatts. Against these shipments we have booked approximately 1.4 gigawatts year-to-date including the period up to today's call.

I will now provide some more details on the approximately 800 megawatts of bookings since our last earnings call. In the US we have signed PPAs for two projects totaling over 100 megawatts DC -- 180 megawatts DC, with a leading Western utility. These projects have CODs in 2019 and 2020, which now brings our total projects with executed PPAs with CODs in 2019 or later to over 1.2 gigawatts DC, which are ideally suited for the competitively advantaged Series 5, Series 6, and MVDC. As evidenced by these bookings, we continue to see many encouraging signs for solar demand over the long-term horizon.

Beyond these project bookings we continue to have active discussions with both utilities and C&I customers seeking large-scale solar power supply. The interest from C&I customers is particularly strong, with these customers typically looking for renewables to meet up to 100% of their energy supply needs. We see this opportunity as a key market focus and have a dedicated team specifically working with these customers to provide comprehensive solutions that meet their needs.

In the US we are also encouraged by the increasing demand for community solar solutions, a market segment that we continue to view as having tremendous potential to meet residential and commercial customers' needs at a dramatically lower cost than rooftop solar. Underscoring the growth potential for community solar, a third-party research firm recently estimated that community solar's addressable market is more than 7 times as large as rooftop solar.

Thus far in 2016 we have booked over 120 megawatts DC of volume that will supply more than 25 different community solar projects across multiple states. The majority of this volume is with a single customer where we continue to see opportunities to collaborate on future community solar projects. An initial 41-megawatt project is underway with M+W as the EPC contractor.

To put the 120 megawatts of bookings in perspective, once this volume has been installed it will effectively double the amount of cumulative community solar installations in the US. We are pleased with the recent progress and continue to work with various customers to capture the growing community solar opportunity in the US.

Elsewhere in the United States we booked modest supply agreements for over 80 megawatts DC in diverse solar markets such as Idaho and North Carolina.

Internationally we have bookings in India and have been strong with over 280 megawatts DC of new bookings. The volume booked was a combination of both module supply agreements and 16-megawatt AC of PPAs.

These latest PPAs bring our current development pipeline in India to 200 megawatts AC. Altogether we have booked approximately 250 megawatts DC of systems projects this quarter, with approximately one-third of the volume scheduled for 2017 deliveries.

Relative to our expectations for the 1-gigawatt of systems business in 2017, we have approximately 400 megawatts DC of contracted projects with additional projects added. We have visibility to a number of additional potential booking opportunities that we continue to pursue towards the 1-gigawatt total.

As we bid on future PPAs, we will continue to maintain our discipline and leverage our strengths in bidding opportunities. In situations where we engage early in the process with key partners and leverage the strength of our new Series 5 technologies into the offering, we feel that we are extremely competitive.

In addition to the 800 megawatts of bookings in the past quarter, we have also been awarded opportunities of over 410 megawatts DC which we expect to convert into bookings during the remainder of this year.

Firstly in Zambia along with our partner Neoen, we were awarded 53 megawatts DC power plant which is expected to deliver the lowest cost solar electricity in sub-Sahara Africa. The energy density advantages of our CadTel modules is particularly strong in this type of hot and humid climate and allows us to provide extremely competitive solar solutions.

The remaining awarded volume of approximately 357 megawatts DC pertains to module supply agreements in France, Israel, India, and Thailand. In France we were pleased with the results of the most recent tender process, where we were awarded over 100 megawatts of volume. As these awarded projects are converted into bookings, we will discuss them in more detail on future earnings calls.

Moving on to slide 6, our bookings in expected revenue terms stands at $6 billion, a decrease of $900 million from the end of last year. The decrease in expected revenue is the result of the higher mix of module-only business booked as compared to a higher mix of system revenue recognized year-to-date.

In addition it is also driven by lower ASPs associated with some of the longer-dated PPAs included in our bookings. While ASPs on these longer-dated PPAs are naturally lower than on a project we are currently constructing, there is still an expectation of attractive margins on these projects, given our efficiency and cost roadmaps discussed at our Analyst Day earlier this year.

Slide 7 provides an update of potential bookings opportunities, which has grown to 24 megawatts DC, an increase of approximately 700 megawatts from the prior quarter. Our mid- to late-stage booking opportunities are 1.8 gigawatts, with international opportunities again comprising close to 90% of this number.

Note that over 400 megawatts of awarded but not booked volume discussed earlier in this presentation is included in this 1.8 gigawatts. Also it is important to note, as the number of third-party module sales opportunities in our potential bookings number has increased over time, we are beginning to see an increase in the velocity at which these projects can move from early-stage opportunities to confirmed bookings. For instance, more than half of our bookings during the past quarter were in early-stage opportunities at the time of our Q1 call.

Turning to slide 8, while North America remains our largest geography, at over 40% of the total, we continue to have a diversified geographic mix of potential booking opportunities. The largest regional increases since the last call were in the United States and India, where we saw significant growth in opportunities from the prior quarter. We are encouraged by the expanding pipeline in both of these regions, given our historical win rate, reflective of our energy density and the mix of system and module sales.

One last point to keep in mind is the timing of bookings can be very uneven. For example, in the first six months of the year our bookings total was only 800 megawatts; however, in the past month alone we have added close to 600 megawatts. We continue to aggressively pursue new opportunities while working diligently to close those projects nearing the finish line, including the over 400 megawatts AC awarded volume previously mentioned, but the timing of when this occurs can be lumpy.

I will now turn it over to Alex who will provide more detail on our second-quarter financial results and discuss updated guidance for 2016.

Thanks, Mark and good afternoon. I will begin with some highlights of our operational performance during the past quarter.

In Q2 we produced 785 megawatts DC or an increase of 1% from the prior quarter, resulting from higher module efficiencies and increased throughput. Compared to the second quarter of 2015 production was 39% higher, as a result of higher efficiencies, improved throughput, and the addition of new capacity.

Our factory capacity utilization is unchanged at 100% in Q2 versus the prior quarter, but increased by 15 percentage points versus the same period in 2015. The higher year-over-year capacity utilization was due to fewer efficiency upgrade activities in Q2 as we focused on maximizing output to meet demand in the first half of the year.

The fleet average module conversion efficiency in Q2 was 16.2%, which was unchanged from the prior quarter but increased 80 basis points year-over-year.

Our best line conversion efficiency was also unchanged from Q1 at 16.4%, but increased by 20 basis points compared to the second quarter of 2015. Most recently our lead line has been running at 16.7% efficiency, and we're encouraged by our progress towards our target of 17% lead-line efficiency at the end of the year.

Turning to slide 11, I'll discuss the P&L results for the second quarter. Net sales of $934 million for the quarter compared to $848 million in Q1. The sales increase resulted from higher third-party module sales, the sale of our Kingbird project, and higher revenue recognition across various systems projects.

Partially offsetting the higher sales was lower revenue recognized on our Silver State South and Stateline projects which reached or neared completion in Q2. The higher module sales were driven by shipments to projects in Dubai and the Southeastern US.

In relation to our Kingbird project it should be noted that this project sale was structured with a tax equity partner with the residual cash interest sold to 8point3 Energy Partners, which resulted in a different accounting treatment as compared to our historical project sales. As a result of the transaction structure, whereby First Solar provides an indemnity relating to the tax equity structuring, only a portion of the revenue from the transaction was recognized and the entire profit of $20 million has been deferred on the balance sheet.

If we had been able to recognize the entire transaction in Q2, our gross margin would have improved by approximately 80 basis points and earnings per share would have increased by approximately $0.16. It's important to keep in mind that the underlying economic substance of the transaction was very favorable; but under current real estate accounting it is not reflected in the current period results. Also, note that all cash related to the project sale, both from tax equity and 8point3 was received.

Continuing on, as a percentage of total quarterly net sales our solar power systems revenue, which includes both our EPC revenue and solar modules used in systems projects, decreased to 83% from 93% in the prior quarter as a result of the higher mix of module-only sales.

Gross margin for the quarter was 20% compared to 31% in the first quarter. The decrease in gross margin percentage resulted primarily from the mix of systems projects recognized between the quarters. As a reminder, in Q1 we sold an additional interest in the Stateline project, leading to a significantly higher mix of revenue from development projects in Q1.

Q2 gross margin was also impacted by an $8.5 million charge associated with the write-down of TetraSun inventory. Adjusted for this item, second-quarter gross margin would have improved by approximately 90 basis points.

Our component segment gross margin decreased slightly to 24% in the second quarter compared to 29% in Q1. The decrease was primarily due to the geographical mix of sales, partially offset by a decrease in our module cost per watt. We are very encouraged by our progress in reducing our module cost per watt through efficiency improvements, increased throughput, and bill of material reductions.

Operating expenses, excluding restructuring asset impairment charges, were $97 million in Q2, a decrease of $1 million from the prior quarter. The decrease was due to lower employee-related costs, partially offset by higher R&D expenses for development of our Series 5 technology.

Restructuring and asset impairment charges totaled $86 million in the quarter, primarily related to TetraSun. As indicated in our recent press release we anticipate total charges of up to $100 million, substantially all of which is expected to be non-cash. We expect operating expense savings of $2 million to $4 million this year, and $8 million to $10 million on an annual basis going forward.

Operating income was $9 million for the second quarter or $94 million adjusted for restructuring. Prior-quarter operating income was $165 million.

Other income was $7 million in the second quarter compared to $36 million in Q1. The current-quarter income was primarily related to a reversal of our ending contingent consideration associated with the TetraSun acquisition. Other income in the prior quarter was primarily due to a $38 million gain from the sale and rebalancing of certain restricted investments associated with our module end-of-life program.

Tax expense for Q2 was $9 million, and the effective tax rate was unusually high due to the impact of restructuring. Adjusted for the restructuring and asset impairment charges, the tax rate for the quarter would have been more in line with the full-year guidance tax range.

Also, note that while not in our Q2 results, during July we received a favorable ruling from a foreign tax authority which will result in a tax benefit of $35 million in our Q3 results.

Q2 earnings were $0.13 per fully diluted share on a GAAP basis and $0.87 on a non-GAAP basis. This compares to earnings of $1.66 in the prior quarter.

Non-GAAP EPS does not include the $0.16 of deferred Kingbird earnings. Please refer to the appendix of the earnings presentation for the GAAP to non-GAAP EPS reconciliation.

Nonrecurring charges aside, which were primarily from TetraSun, Q2 was another strong quarter of financial performance and we're pleased with our progress through the first half of the year.

Continuing on to slide 12, I will discuss select balance sheet items and summary cash flow information. Cash and marketable securities decreased $213 million to an ending balance of $1.7 billion.

Our net cash position also decreased to $1.4 billion. Cash decreased from the prior quarter due to ongoing construction of projects on balance sheet, capital expenditures, and debt repayments.

For the quarter, net working capital, including the change in noncurrent project assets and excluding cash and marketable securities, was virtually unchanged. The decrease in project-related assets due to the sale of Kingbird and the reclassification of certain projects in Chile and India to PV solar power systems was largely offset by an increase in unbilled AR. The Luz del Norte project was reclassified, as it has been placed in service; and we anticipate holding the asset for several years pending transmission upgrades in Chile.

The increase in unbilled AR is expected to decline in Q3 as billing milestones are reached.

Total debt decreased $66 million from the prior quarter: $233 million, primarily due to the $70 million repayment of the remaining balance on our Malaysia debt facilities. The remaining debt outstanding is comprised of $163 million of limited or nonrecourse project-level debt and $70 million of recourse project-level debt, the majority of which is a short-term VAT loan associated with our Luz del Norte asset.

Cash flows used in operations of $75 million compared to cash flows from operations of $50 million in the first quarter. Free cash flow was negative $140 million compared to positive free cash flow of $13 million last quarter.

Capital expenditures were $78 million as compared to $52 million in the prior quarter, as we increased investments in our Series 5 technology. Depreciation for the quarter was $53 million, or approximately $2 million lower than the prior quarter.

Before turning to our updated guidance for the year we have summarized on slide 13 the expected impact of restructuring charges and the associated future benefits from these actions. Firstly, as indicated, total expected charges associated with both the TetraSun and other restructuring announcements is between $105 million and $120 million, the majority of which is non-cash. Of the $90 million to $100 million of TetraSun charges, we incurred approximately $86 million in Q2 with the remainder expected to be incurred in the second half of 2016.

The charges associated with our EPC and skytron restructuring are expected to be in the range of $15 million to $20 million and are expected to occur primarily in the third and fourth quarters. Going forward we expect total annual savings related to these actions of $60 million to $80 million, of which $55 million to $65 million are cash savings.

Annualized OpEx savings associated with these actions are expected to be between $35 million and $45 million. Note that due to the timing of the restructuring we expect limited OpEx savings in 2016.

Turning to slide 14 I will now discuss the updates to our full-year 2016 guidance. Overall, the changes to our full-year non-GAAP guidance, which excludes restructuring, are relatively minor. While the first half of the year results have been strong we continue to take a measured approach to guidance for the time being, given uncertainties in the second half of the year. In particular, we anticipate closing the sales of our Moapa asset, our California Flats asset, and our remaining 34% interest in the Stateline asset.

As we progress further into the third quarter we will have greater visibility into these various project sale dynamics.

Relative to Stateline, our guidance assumption continues to anticipate dropping our remaining 34% interest in the project to 8point3, subject to market conditions. To the extent that we do sell all 34% of our interest in 2016, this would likely result in earnings above the midpoint of our guidance.

We will also continue to evaluate the potential to sell a portion of our remaining interest in Stateline in 2017, dependent upon developments in our own business during the second half of the year, as well as based on market conditions relating to 8point3.

Turning to our guidance ranges, our net sales forecast is unchanged at $3.8 billion to $4 billion. We are bringing up the low end of our gross margin guidance to a revised range of 18.5% to 19%, compared to the prior range of 18% to 19%. This update is a result of the significant project cost savings achieved in the second quarter.

Note that as previously discussed, the net sales range and gross margin guidance include only a portion of the revenue and none of the deferred profit from the Kingbird sale.

Our GAAP operating expenses are projected in the range of $485 million to $520 million for the year. This includes the $105 million to $120 million of restructuring associated charges.

Our non-GAAP operating expense guidance, excluding those restructuring and asset impairment charges, is unchanged at $380 million to $400 million. We are leaving the range unchanged as 2016 OpEx reductions are expected to be limited, given the timing of the announced restructuring, and are offset by earlier-than-planned production startup expenses from the Series 5 acceleration and from incremental R&D for Series 6.

Improvement in the gross margin has been flowed through to the low end of our non-GAAP operating income guidance, which has been raised to $310 million. The non-GAAP effective tax rate is expected to be in the range of 16% to 18%. This projected tax range excludes the impact of all restructuring actions and also excludes the impact of the $35 million tax benefit mentioned previously.

Moving on to EPS guidance, on a GAAP basis we expect earnings in the range of $3.65 to $3.90. Excluding restructuring charges the low end of our non-GAAP EPS guidance is a range to $4.20 with the upper bound unchanged at $4.50.

The impact to EPS from the expected sale of our remaining interest in Stateline and First Solar's share of 8point3's earnings remains unchanged at approximately $145 million net of tax. Additionally approximately $0.20 of other income net of tax is included in the earnings guidance range from the previously discussed Q1 sale of the restricted investments. Also keep in mind that the EPS guidance does not include the $0.16 of deferred Kingbird earnings.

The distribution of earnings for the remainder of the year is anticipated to be slightly more weighted towards the fourth quarter due to the timing of expected project sales.

Operating cash flow has been reduced to $500 million to $650 million from the prior range of $500 million to $700 million, reflecting the expected impact of cash restructuring charges. Again, the operating cash flow range does not include approximately $320 million from the expected sale of the remaining interest in Stateline, which we expect to treat as an investing cash flow.

Our capital expenditure guidance has been revised to a range of $275 million to $325 million from the previous range of $300 million to $400 million, reflecting a change in timing of expected spend. The adjustment to timing has no impact on our planned efficiency roadmap and Series 5 product launch.

The net result of these changes leaves our net cash balance guidance unchanged at $1.9 billion to $2.2 billion.

Turning to slide 15, I will now summarize our progress during the past quarter. Firstly, our financial results for the second quarter were strong with revenue of $934 million and non-GAAP EPS of $0.87. We updated our 2016 non-GAAP earnings guidance range to $4.20 to $4.50 range.

Our technology performance is progressing towards the goals we outlined in our Analyst Day, with a best-line efficiency currently running at 16.7%. And our year-to-date bookings are now 1.4 gigawatts, with over 800 megawatts booked since our last earnings call. The number of potential bookings opportunities continues to grow and is now over 24 gigawatts.

With this we conclude our prepared remarks and open the call for questions. Operator?

One thing I guess before we do that. Apparently there was a technical issue I think on the webcast, that the individuals who were logged in on the webcast I think did not get the first part of the prepared remarks. There will be obviously as always a link on our website that you can hear the replay; and that replay would be available later today for any of those on the webcast who were unfortunately unable to catch the first portion of today's prepared remarks.

With that, operator, we will open it up for questions.

(Operator Instructions) Vishal Shah, Deutsche Bank.

Hi, thanks for taking my question. Mark, can you maybe talk about the profitability of the bookings that you are seeing right now, especially in markets like India? And can you just talk about how we should think about capacity, especially as you get into 2017? Thank you.

Yes. I guess just general color around the profitability on bookings, we're still very pleased with the margin realization that we're capturing. We are being very disciplined though and engaging in making sure that we are capturing full value of our technology.

If you use India as an example, it's a market that our technology is competitively advantaged. In a hot humid climate we have a superior spectral response, and that can command an energy advantage over crystalline silicon.

Plus it's mainly fixed tilt structures in India, which is again more advantaged for our Series 4 product than, say, a tracker solution. Plus we have a number of self-developed projects in India, and the returns on what we are capturing against the sell-down -- our anticipated sell-down of those projects are in line with what we would normally expect and even in some cases on the upper end of our expectation. We have been very pleased with the ultimate performance of those assets.

Having said that, is the market becoming more competitive? It clearly is. We are seeing a lot of very aggressive pricing behavior in the market, whether it's at the PPA level or whether it's at the module level.

And that's why we have to be disciplined and selective, and engaging with customers and developing relationships to ensure we can capture the full value of the offers and the value creation that we provide to our partners and our customers. So we are being disciplined in that regard, and we are seeing a very aggressive marketing environment.

There's no doubt about that. That's what's also informed our view around capacity expansion and why we are being disciplined as we think about when we'll add capacity. As we indicated in our prepared remarks, we will continue to assess that as we move forward.

As it relates to near term, our focus is converting all of our capacity into Series 5. We want all of our Series 4 product into series 5 as quickly as possible, and our current timeline would indicate that that would happen in the end of 2017.

We will continue then also to validate the business case around Series 6. And as we get more informed and have a better understanding and realization of the current views around Series 6, we will start to think about how do we add capacity and the associated timing around that.

So I would say for now, again: near-term focus, moving Series 4 to Series 5; as we get better information and validate Series 6 that will determine the timing around when do we start adding incremental capacity.

Ben Kallo, Robert Baird.

Hi, thanks for taking my question. I've got two. First of all, when you guys talked about your margin I guess at Analyst Day, you have your target margin -- and maybe this loops onto Vishal's question and a lot of concerns out there in the marketplace. As we look to next year, is that a valid range for us to think about as of this year and what you guys have talked about as your target margin?

I guess as you look at projects like Moapa which you can move around a little bit, what type -- and I think you guys already said that you're at the high end of your guidance if everything goes okay. So it seems like there's some flexibility into next year as well.

So how do you guys think about that, about 2017, and being able to move projects into that if you don't need to finish them this year, as well as the margin question?

I will do the margin question. I will let Alex take the question around how we're thinking through timing of activities between 2017 and 2016. As it relates to the margin realization, similar to what we I think highlighted in the Analyst Day, especially when you start looking at our contracted pipeline in particular, the margin profile will start on the lower end of the range in 2017 and then continue to improve as we move through the balance of this decade. Primarily because as we transition into Series 5 it's a much more competitive product. And what we look at right now for our Series 5 capacity in 2017 it's probably going to be right around a gigawatt, which means we have about 2 gigawatts of Series 4 product that we're going to have to sell through.

We also indicated in the comments today that the value of that delta in form factor as it relates to lower labor cost, fewer connectors, better wire management, elimination of clips is worth over 100 basis points of equivalent efficiency. So what you should think about is our margin profile in 2017 is going to be on the lower end. It's mainly because two-thirds of our production volume is still going to be Series 4.

As we transition that into Series 5 and then really get to a position of an equivalent form factor and really capturing full value of our energy advantages, you will see margin expansion as we go into 2018 and 2019 and 2020.

Yes, talking about the projects, it really relates to our Cal Flats, Moapa, and then our residual interest in the Stateline asset. We would expect to complete the sale of the Cal Flats asset in either late Q3 or early Q4, and then Moapa in Q4 of this year.

We've guided to selling our remaining 34% interest in Stateline this year, and that's still in our base case. I'd say along with both Moapo and Cal Flats there is still significant uncertainty around the structure and the ultimate value of those sales. So in terms of our guidance we are remaining conservative as we continue to develop those structures.

The Stateline sale is really dependent on the other two deals as well as market conditions relating to the yield curve. So we will make that decision in Q3 or Q4. We do have the flexibility with Stateline to push some of that into 2017 if needed.

Brian Lee, Goldman Sachs.

Hey guys, thanks for taking the questions. I had two. First off, last quarter, Mark, you mentioned roughly 950 megawatts of potential systems bookings as you move through the back half of the year. I think you had mentioned those were international.

Can you update us on if that pipeline is still intact, or how much has already been booked and vice versa how much maybe has come off?

Then the second question, just on some specific projects. On Stateline if you can update us on how much revenue is left to be recognized and what percent of the project has been completed.

And then separately, I know you alluded to this. But on Moapo it sounds like the expectation is that you do sell the project in its entirety this year in Q4. But are you exploring different options of monetization similar to what you've done with Stateline now, that you've taken it off the 8point3 ROFO list? Thank you.

Relative to the contracted or systems business pipeline for 2017, one of the things we said in our prepared remarks -- with our incremental booking that we now have, India which is a little less than 100 megawatts DC, we have about 400 megawatts of systems business already positioned for 2017. We have a pipeline that would move us towards closing that out and getting to a gigawatt expectation for 2017 for the other 600 megawatts or so.

We're still working through those. There are a couple in there that I would say -- not around the economics. These economics are very compelling. There are some issues around structuring of the PPAs and overall bankability of those PPAs.

As you know, as you indicated already, some of these are in international markets where we're trying to work through in making sure they are clearly financeable and bankable. So there are some challenges in that regard.

But what I would say is there's sufficient pipeline of projects that would enable us to get to our goal of having a gigawatt of systems business in 2017. The other aspect that we continue to be very mindful of and very disciplined on is making sure that we get a proper return on capital, given the risk profile of doing international development business.

In some markets pricing has gotten very aggressive, so we have to be careful in that regard. So we'll be balanced, making sure that the return on capital is commensurate with the underlying risk associated with those opportunities.

But what I would say in aggregate, we're still very happy with the robustness of our pipeline to get to our goal next year of a gigawatt of systems business.

On the Stateline discussion, I will take that one and let Alex take the question on Moapa. When you get the Q, which will be available tomorrow, you'll see -- I think it will say effectively 96% or somewhere in the upper 90%s Stateline has been completed through the second quarter. So the vast majority of the revenue that was related to the sale for Southern has now been recognized through Q2. We have about 4% at that left, and the project will achieve COD in the third quarter.

With relation to Moapa, yes, it's no longer on the ROFO list for 8point3. That gives us a lot more flexibility in the structuring.

I say that we're still in early stages, so we're not going to go into the detail of the structure now. But we would expect to monetize the entire asset in Q4 of this year.

Paul Coster, JPMorgan.

Thanks. Mark, in the prepared remarks you talked of how the expected revenue per expected module of shipment is coming down a little bit, owing to the shift towards longer-dated PPAs. Is that a trend? And is it a trend that in any way relates to your customers' awareness of your product transition? Thank you.

You know, I don't -- look, clearly it's a trend. PPA prices are coming down, and you can see that each and every day when somebody else announces a new PPA price, wherever that may be around the world. PPA prices have become pretty competitive in a number of markets.

Some of those, I would argue in some cases it's questionable whether or not they are going to be viable or able to be achieved. But there's a pretty long runway to delivering against some of those assets that go out into the 2019, 2020 time frame or potentially even later than that. So clearly PPA prices have come down.

I don't know if it's directly related to -- I wouldn't position them as being directly related to our advantages and our advancements in our roadmap and our form factor and efficiency. The other day we were competing against crystalline silicon, so it's really relative to the alternative technology that's in the marketplace. So to the extent that we create separation relative to crystalline silicon, that value should accrete back to First Solar.

I would also say that there is a very strong bias in the marketplace for more and more customers and developers and partners coming to First Solar first. We've got a number of parties who historically over the years have moved away from First Solar that are coming back to First Solar. We have also got a number of other partners that are starting to come to discussions with First Solar that we historically have never had conversations with.

I do think that is related to our roadmap and our technology, and understanding where First Solar is going to go long term with our capabilities, not only on the module side but, as we indicated before, with MVDC and the advantages that that creates for the balance-of-system cost and just general learnings around BoS and optimization and wire management and a lot of things that we're doing to drive down a lot of cost.

We're getting more and more interest from other parties who want to do more business with First Solar. So clearly a trend to a lower PPA, but not to the extent that when I look at where the PPA pricing is going that I'm concerned relative to what we can personally do with our overall cost-reduction roadmap.

Just as a relevant data point, just on the module alone year-on-year our reduction on our module cost has been in the mid to upper teens. So in the last 12 months we've taken mid to upper teen type of cost out of our module, which is extremely impactful and competitive and ultimately creates an opportunity to have robust margins against the contracted pipeline that we've created.

Phil Shen, ROTH Capital Partners.

Hey, thanks for taking my questions. I think you've touched on this in part already, but I wanted to ask a more pointed question. Module ASPs globally are falling sharply on overcapacity issues. The acceleration to the downside really started at the beginning of July.

How is this impacting your module-only and module-plus business? And how do you expect to respond to this downside price move?

You talked about moving up the capacity rollout of Series 5. Could we start to see some meaningful volumes of the Series 5 module? And if so, can you quantify what that might be in 2017?

Like I mentioned previously, Phil, our volume right now for Series 5 in 2017 will be about a gigawatt. So about a third of our production will be Series 5.

So there still will be a meaningful amount of Series 4 product that we obviously will have to sell through in the next year. Some of that is already contracted, but there is more volume that we have to make sure that we sell through in that regard.

You're right, module pricing has become more competitive. We are still able to command a premium for our technology. That's why we try to focus on markets where we're more significantly advanced.

We even referenced in our prepared remarks Zambia, right? There was an opportunity in Zambia for 50 megawatts. It's a hot humid climate, which is advantaged for our technology.

So what we are doing globally is looking at pockets of strength. We're trying to manage with constraining ourself on available capacity at least at this point in time, especially as it relates to Series 4. We are looking for pockets of strength where we can sell through and capture the highest-margin opportunity entitlement that we have relative to what's going on in the marketplace. But we're very well aware of how aggressive some of the pricing has become.

I would also say that not every customer does see the value, but there are clearly some that differentiate. Some see the value in our balance sheet, our performance.

It depends a lot on the buyer. So if you are selling to developers who are looking to flip assets, they may not value the full entitlement of the module, and they may not be the sellers that we choose to sell to.

While we are capacity constrained, as Mark said, that allows us to pick and choose those buyers a little more and look for those long-term owners that will fully value the benefits of the module.

Julien Dumoulin-Smith, UBS.

Hi, good afternoon. I wanted to ask just a couple quick questions. Maybe to first start off, following up with the last, can you comment a bit on the duration of the cycle that we're looking at? What inning do you think we are, to use the analogy, in terms of the inventory cycle and the buildout and the supply?

Then perhaps secondly, going back to the 2017 discussion on margins, can you elaborate a little bit? As you ramp up from that 400 megawatts of development systems to that gig, when you are thinking about margins and being at the lower end of that range, is that specific to the development, inclusive of that 1 gigawatt playing out?

I don't know. We always get that question: What inning are we in? I don't know.

Look, this industry is so dynamic and can change so quickly, to be able to articulate and have a well-thought-out view of what the inning is is difficult, because it can change. The game seems like it can change very quickly and events can happen very quickly that become disruptive or undermine, or events could happen that become very positive.

Clearly on the positive side there's a clear understanding globally of the demand and elasticity around solar and the competitiveness of solar in many different markets. And that's driving a demand profile that obviously is very encouraging.

There's also -- the offsetting side of that is there is quite a bit of capacity that is being added -- or plan to be added, maybe is a better way to say that. But that could change very quickly as well.

As people think about the market opportunities and what they are willing to sell through and what ultimately is sustainable, the planned announcements that have been communicated may or may not happen. There is still a lot of uncertainty as to whether or not that production supply comes to market or not.

So I don't want to get into an analogy of an inning because I think it's very, very difficult from that standpoint.

On the margin piece I'd say 2017 is still pretty uncertain. We will have more clarity at the end of the year on the guidance call.

What I would say though is that we always knew that 2017 would be a challenging year, especially as we saw margins decline from our more lucrative legacy contracts as they rolled off. The ITC extension in 2016 came too late to influence 2017. That's why we think it's incredibly important that we see both a transition towards higher module business in 2017 but also the product, from Series 4 to Series 5.

The other thing I would refer you back to is the Analyst Day when we talked about what our long-term contracted pipeline looks like. Back then we spoke about having over 2.5 gigawatts of contracted assets after 2017, with north of $1 billion of contracted margin entitlement. I think if you look at that, combine that with -- even without capacity expansion another 7 or so gigawatts of modules for sale, I would expect some of that volume to be incremental systems as well as modules. When you combine that with executing on the cost-reduction roadmap and our OpEx coming down I think we look to see beyond 2017 having acceptable and robust margins across the new products.

2017 itself is somewhat opaque today, and we will give better guidance around that towards the end of the year.

Patrick Jobin, Credit Suisse.

Hi, thanks for taking my question. I have three of them here. First, on 2017 are you still comfortable with the 1-gigawatt system business?

I guess this morning Dominion came out and doubled their outlook for solar in 2017, and Georgia Power clearly with their 1.6 gigawatt business. I guess I'm trying to understand your comfort level on the 1 gigawatt, if that has increased or decreased relative to your Q1 view. That's the first question.

Second question, can you just flesh out maybe that low end margin guidance? There was a few ranges at the Analyst Day; I'm curious.

Third point, the component gross margins of 24% -- I guess more of an accounting question -- but is that driven by spot market pricing or the pricing locked in during contracting? Thanks.

Yes, on the gigawatt for 2017, we're 400 megawatts in right now. If you asked me the other 600, I have a really high confidence on 300 of the 600.

The other 300 I would say there's still a number of moving pieces. And some of the announcements that have come out here recently and now people are thinking about, especially in the US, opportunity to procure in in 2017 that gets me more encouraging.

As I look at some of these international opportunities -- and especially the challenge in some cases to these PPAs, as I indicated, to get them to be financeable -- that sort of sways me the other way.

But I would say on balance we feel that we will get to with -- whether it's 850 or 900 it's going to be within a ZIP Code that's going to be close to that gigawatt of volume business for next year. So I'd say there's reasonable level of comfort from that standpoint.

On the margin guidance I think the best way to handle it, as Alex already said, is we're not going to get into the specifics of that right now. There's so many moving pieces that we would rather wait until we do our guidance call in December.

What I would also continue to point you to is that whatever that view is when we communicate that in December, I don't think we should look at this as a discrete data point, as indicative of the long-term earnings potential of First Solar. The 2017 margin profile, especially as we indicated with Series 4 still being the predominant product in our platform, it's going to drive a little bit of margin pressure. There's no doubt about that.

But that's a temporary issue that corrects itself in the second half of 2017 and then is completely fixed by the end of 2017, where we got a form factor that's consistent with crystalline silicon and driving towards a 390-watt panel with all of the energy advantages that we've talked about around CadTel. So when we look at it through that horizon, we feel very confident about where we're going with our technology and ultimately the competitiveness of our products.

Around the component gross margin, there is an element of it's spot and it's also an indication of contracted volume. So it's a combination of the two.

As we indicated this quarter, as an example, over half of our bookings that we recorded of our 800 megawatts or so were not in our late-stage development pipeline, which means that those were transactions since the last earnings call that we've gone out and we've negotiated around pricing in a very short period of time, and then closed those orders.

And a high percentage of that volume will ship this year, and those margins that will be realized against those shipments will be in line with what we're currently reflecting in our components segment today.

As I also indicated, if you just look at our module cost delta year-on-year, we've seen high teen, upper teen cost reduction on our module just in the last 12 months. And the team continues to make tremendous progress on that side of the house.

So we're able to as best we can manage that price/cost battle. But we also do it in a way that we're trying to be selective of where we're selling through and to make sure we're targeting markets that we get the highest value for our products.

Krish Sankar, Bank of America Merrill Lynch.

Hi, thanks for taking my question. I had a couple of them. One is, Mark, I understand you are not giving next-year guidance. But I'm just trying to think: Is the thought process right that when you move to Series 5 capacity more next year and you decide to add more capacity, your cash balance should come down exiting 2017?

Then the second question I had was, of your total shipments in Q2 or of the booking opportunity that you have, how much of that booking opportunity is modules-only? And of the total shipments in Q2, how much is direct versus indirect module? Thank you.

On the Series 5 comment and impact to cash balance, one thing I think you also ought to think about around Series 5 is it's really relatively CapEx-light, relative to if we had to start a new greenfield. All we are taking is our existing platform and we're converting it into a different product that is optimized around [LAR] management and some other things along those lines. So it's not really a significant CapEx.

We're going to finish this year somewhere around $2 billion of cash as we exit the year. Series 5 in and of itself is not going to drive a significant delta to that number, as we deploy that capacity or conversion of product -- Series 4 into Series 5 product -- in 2017.

I'm not sure I got your last question. Can you repeat your last question?

Yes. I was trying to find out: of your total shipment of modules, how much is direct versus indirect in Q2? And of the 24-plus gigawatt booking opportunity, how much is modules?

So when you say shipments in Q2, are you talking bookings in Q2 or true shipments?

True shipments.

Okay. What we disclosed in that regard was that a little bit more than 80% of the shipments -- or the revenue I guess is a better way to say it -- in Q2 was systems and the balance was modules. So I don't know if that gets to your question in that regard.

In terms of the 24 gigawatts and the breakout between systems and modules, we don't really provide that level of detail. But I think you clearly can anticipate that that profile and that mix of systems versus module is increasing. So there will be a higher percentage of module sales reflected in that pipeline than would have had historically.

I think that's our last question. The other thing I would just like to remind everyone on is, again, apologize for the issues on the webcast. There will be a replay available on the First Solar Investor Relations portion of our website, and that should be available later today. We thank everyone for their time.

Thank you. That does conclude today's conference call. We do thank you for your participation today.