第一太陽能 (FSLR) 2018 Q2 法說會逐字稿

Good afternoon, everyone, and welcome to First Solar's Second Quarter 2018 Earnings Call.

This call is being webcast live on the Investors section of First Solar's website at firstsolar.com.

(Operator Instructions) 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, Abby.

Good afternoon, everyone, and thank you for joining us.

Today, the company issued a press release announcing its second quarter financial results.

A copy of the press release and associated presentation are available on First Solar's website at investor.firstsolar.com.

With me today are Mark Widmar, Chief Executive Officer; and Alex Bradley, Chief Financial Officer.

Mark will begin by providing a business and technology update.

Alex will then discuss our financial results for the quarter and provide updated guidance for 2018.

Following their remarks, we'll then have time for questions.

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 presentation 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.

I would like to start by discussing the global PV market.

As you are aware, since our last earnings call, there has been significant developments in the global market, primarily stemming from policy decisions in China.

The near-term impact has been an almost immediate collapse in pricing across the crystalline silicon supply chain.

While we have seen planned maintenance pull forward or other actions taken to better align near-term supply with demand, there is still an oversupply across the value chain which is driving declining module ASPs in both China and certain international markets.

While it's too early to fully assess the long-term effect these decisions will have on the industry as a whole, the end result most likely will be more competitive PV power prices which will lead to demand elasticity, both in China, specifically, and the global market in general.

Additionally, we will likely see industry consolidation as uncompetitive technologies and financially unstable companies struggle to compete.

While we will continue to carefully monitor these recent developments, we remain focused on leveraging our competitive advantages and executing our differentiation strategy.

First and foremost, our cad tel technology and, specifically, our Series 6 product, is a competitive advantage.

In an industry that suffers from a lack of differentiation, Series 6 has the potential to achieve a distinctive combination of low cost and high efficiency.

While there is still a great deal of work ahead to realize its full potential, our unique technology is a key competitive advantage.

In addition, as we look over the horizon over -- of an oversupplied market, our nearly 11 gigawatt pipeline of future contracted shipments is a position of strength.

While I will talk more about this in a moment, nearly 80% of our available supply from now until the end of 2020 is booked.

This is a substantial pipeline of contracted volume that provides good visibility to future demand at an uncertain time in the market.

Finally, another competitive advantage unmatched in the industry is our balance sheet which enables us to invest in our business and be opportunistic at a time when greater stress is likely to be placed on competitors who are already highly levered.

Our net cash at the end of Q2 was a record $2.7 billion even after significant year-to-date investments in Series 6 capacity and project development activities.

While this is an industry that has experienced periods of overcapacity and difficult market conditions in the past, the long-term potential for solar energy still shines brightly.

Furthermore, as a company, we have never been better-positioned to deal with the current near-term challenges given our Series 6 product, contracted bookings and balance sheet strength.

Before providing an update on our progress related to Series 6, there are some important points to keep in mind pertaining to our financial results for Q2.

First is that when we began our transition to Series 6 1.5 years ago, we anticipated that 2018, and in particular the first half, will be the trough in our earnings power.

With Series 6 production largely slated for the second half of the year, we knew we would be at the low point of our module availability in a period with elevated levels of ramp and startup cost.

Additionally, the second quarter was significantly impacted by the timing of closing of certain project assets.

As we've seen in the past, there's a great deal of uncertainty associated with project sale timing, and the effects on a single quarter can be pronounced.

Given this quarterly variability, we provided financial guidance on an annual basis as we believe this is the most meaningful way to evaluate our performance.

Lastly, certain initial Series 6 production issues that we have experienced during Q2 impacted our results.

Lower-than-targeted throughput and yields resulted in fewer modules available at project sites and a higher module cost per watt.

While we see these ramp related impacts primarily as near-term issues rather than as long-term structural challenges, they, nonetheless, caused a delay in some project revenue recognition and resulted in a decrease in our full year margin outlook.

Alex will provide discussion around the financial results in more detail later and provide an update to guidance.

Now turning to Slide 4. I'll provide some more context related to our Series 6 ramp and the manufacturing issues mentioned.

Overall, we are very pleased with the progress we have made thus far and remain confident in the long-term capability of the Series 6 product from a cost and an efficiency perspective.

To give you a sense of the progress we are making, at the time of our last update in April, we had only recently started production and initial commercial shipments from our Ohio factory.

Since that time, we have commenced production at our second Series 6 factory in Malaysia, with our third Series 6 factory in Vietnam not far behind.

In Vietnam, we are completing factory acceptance test of the equipment and we expect the first module production in late Q3, with commercial shipments to follow in early Q4.

Construction of our second Series 6 factory in Vietnam is also progressing according to schedule with tool installations beginning later this quarter, followed by first production in 2019.

In the U.S., we are progressing with our second Series 6 factory that we announced on our previous earnings call.

We held the groundbreaking events for the new factory in early June, and the 1.2 gigawatt nameplate factory is anticipated to commence production in late 2019.

Overall, there has been significant progress made in the past 3 months, and the organization is intently focused on building out our Series 6 capacity.

As it pertains to the manufacturing ramp of our Ohio and Malaysia factories, we have made substantial progress over the past 90 days.

Production at the Ohio factory is now running at approximately 60% of nameplate capacity and our Malaysia factory has ramped very rapidly to over 40% of nameplate.

However, even with this progress, the planned 2 -- Q2 production was below our expectations.

Our biggest challenge remains the throughput on the back end of the line bussing through final pack out.

The layout at the back end of the line was built according to the tool set availability specification which resulted in few required buffers.

As we started to ramp the back end of the line with the toolset availability not yet at a mature state, we realized there were multiple single points of failure in the line that could shut down production.

Effectively, the line was not adequately buffered given the current performance of the toolset.

We are in the process of installing inventory accumulators to properly buffer the back end of the line.

Once completed in our Ohio factory, we'll use our copy smart approach to roll out to Malaysia and Vietnam.

Over time, as the toolset availability improves, while the inventory accumulators will remain in place, the need for inventory buffers will decline.

The impact of reevaluating the back end, identifying the required buffers, and installing the inventory accumulators across our manufacturing facilities, in addition to adversely impacting Q2, has resulted in a reduction of approximately 200 megawatts to the full year Series 6 production plan.

It is important to note that despite the 2018 volume reduction, with the actions we are taking, we anticipate to exit the year at the originally anticipated throughput levels and enter 2019 on track to our previously announced Series 6 production volume.

Keep in mind that the issues we are working through do not impact our long-term outlook for Series 6. Relative to our long-term expectations, I wanted to make a comment on the toolset, which incorporates approximately 200 tools across the front end and the back end.

Through our initial production, we have validated the cycle time and performance of each tool, and the capabilities meet or exceed our planned requirements.

This validation helps to critically inform our views on long-term expectations.

Module wattage continued to improve steadily and is currently averaging 415 watts per module, as compared to a production entitlement that yielded most modules near 400 watts at the end of our last call.

Our top bins are currently at 420 watts per module and are approaching 425 watts.

With the robust pipeline of advancement to come, we have line of sight to continued improvements in the fleet average efficiency.

Overall, our efficiency and watts per module are on track relative to our expectations for the year.

Series 6 product readiness has taken a significant step forward since last quarter with the completion of both UL and IEC certifications.

We indicated previously that our -- achieving these certifications was primarily a matter of time and completing both the certification is an important milestone.

We have not yet scored this metric green as a matter of due course given the relatively recent introduction of the Series 6 product.

As we gain more experience throughout this year, we expect to advance this metric to green.

Continuing on Slide 5 on higher bookings activity for the past quarter.

Since our last call, bookings will -- have continued to be solid as we have contracted nearly 900 megawatts of new business.

This brings our year-to-date net bookings to 4.1 gigawatts and our total future expected shipments to 10.9 gigawatts.

To put our future expected shipments into context, almost the entire 10.9 gigawatts is expected to ship between now and the end of 2020.

Given the same time period, our anticipated supply of both Series 4 and Series 6 modules is 13.8 gigawatts, which implies approximately 2.9 gigawatts of remaining bookings to fully contract through the end of 2020.

Factoring in the over 750 megawatts of contracts signed but not yet counted as bookings, the remaining bookings number to the end of 2020 drops to less than 2.2 gigawatts.

While we still have considerable work ahead to fully contract the remaining 2.9 gigawatts of supply, we currently have 5.1 gigawatts of mid to late-stage opportunities with shipment requirements before the end of 2020.

An important element to highlight relative to our year-to-date bookings is both the stability of our Series 6 pricing and the pricing advantage of Series 6 versus Series 4. For example, the average ASP of Series 6 modules booked this year is essentially consistent with the 2017 Series 6 bookings average module ASP.

Note, this is especially noteworthy given the current year bookings are for shipments through the end of 2020.

Furthermore, when compared to 2018 Series 6 and Series 4 bookings, Series 6 average ASP is 6% higher than Series 4.

We also continued to make progress in building our systems pipeline as we highlighted by 2 new PPAs we've booked in the U.S. One PPA for 75 megawatts AC was awarded by a utility in California and has an expected completion date in 2021.

A second PPA for 73 megawatts AC was obtained through a recent pipeline acquisition and will be our first project in South Carolina.

The PPA is with South Carolina Electric and Gas, and the project has an expected completion date of 2020.

We're excited by this entry into a new part of the southeast United States in a region that has excellent solar resource.

While this was the only project acquired with a signed PPA, the pipeline acquirer also includes a number of mid to late-stage opportunity development projects which, in aggregate, total approximately 600 megawatts.

As I mentioned earlier, our balance sheet strength allows us to be opportunistic.

And we will continue to evaluate other project or pipeline acquisition opportunities so long as they meet our return thresholds.

While not yet counted as a booking, we have also signed an approximately 60 megawatt AC PPA with the utility in western United States for a project that will provide power to a corporate customer.

We'll provide more details in the future, but this project is a prime example of our capabilities to address the renewable energy goals of corporations in partnership with utilities by leveraging efficient and reliable large-scale offsite generation.

Previously, we're able to bring these same capabilities to bear when we partnered with NV Energy to power data centers for Switch with clean, affordable electricity.

In addition to the signed PPA, we are in advanced discussions with utilities for 2 additional projects that would supply over 100 megawatts AC of power to corporate customers.

All 3 of these opportunities highlight this important growth opportunity as many companies increasingly commit to 100% clean energy.

Outside the U.S., we also continue to see growth in our systems portfolio this past quarter with approximately 30 megawatts AC of additional systems bookings in Australia.

In addition to new development project bookings, we also recently converted an additional 65 megawatts DC of previously contracted module volume to an EPC sale.

When we add EPC scope to previously booked module sales, we do not count these agreements as new bookings.

However, they do provide incremental future revenue and margin.

This is the fourth project that we will construct for Tampa Electric with an expected completion date of this project in 2019.

Year-to-date, our total net system bookings are now 1.3 gigawatts, which is -- comprised of over 750 megawatts of development project bookings and more than 500 megawatts of EPC contracts which we converted from module sales.

In addition to the system projects discussed, the remaining bookings for the quarter were module sales primarily to customers in the U.S. Other module agreements were also signed with customers in India, the Middle East and Europe.

Continuing on to Slide 6. I'll next discuss our mid to late-stage bookings opportunities which, on a net basis, is unchanged at 8.3 gigawatts DC.

When factoring in the bookings for the quarter, a number of which were included as opportunity from the prior quarter, our mid to late-stage pipeline actually grew.

On a geographical basis, North America increased with a roughly corresponding decrease in opportunities in Asia Pacific.

North America increased primarily as a result of acquiring the project development portfolio in the southeast U.S. mentioned previously.

Keep in mind with -- in addition to more -- to the more advanced project opportunities which are included in the 8.3 gigawatts, there is a robust portfolio of early-stage projects not reflected here.

Similar to last quarter, the total potential opportunities include deals that are signed but not yet counted as bookings until finance in our other CPs are closed.

As mentioned, there are over 750 megawatts of such projects including the PPA with the Western utility already discussed.

With respect to the expected shipments, timing of the mid to late-stage opportunities, we have 5.1 gigawatts of opportunities in 2019 and 2020 against the remaining supply in this time period of 2.9 gigawatts.

Early-stage projects not included in this number provide additional opportunity to sell the remaining volume.

Next, I'll provide an update on progress we are making on our systems project pipeline.

As we discussed in our Analyst Day last December, the systems business remains a core part of our strategy, and on average, we are targeting around 1 gigawatt per year of development business in the next few years.

As it pertains to the development, we remain focused on key markets such as the United States, Japan and Australia, where we can pursue a differentiated strategy that can lead to capturing value and compelling returns.

Select EPC opportunities in the U.S. will also remain a priority as these agreements enable greater customer engagement and we believe enhance our value proposition for utilities wanting to own their own generation.

As highlighted on Slide 7, we are making good progress towards achieving our annual development target.

Note that the timing of revenue recognition on system project will vary from the shipment timing shown on the slide; however, is a good indication of the current status.

As illustrated, we have nearly reached the 1 gigawatt target in 2019 with contracted development projects, and we have the potential to exceed that mark if we're able to close the mid- to late-stage opportunities shown.

Both EPC projects plus potential EPC conversion opportunities take that total even higher.

In 2020, we are more than halfway to our target with a potential to significantly exceed that mark.

Keep in mind that there are always contracting risk associated with mid to late-stage projects, and we know -- we do not expect that we'll ultimately book all of these mid to late-stage opportunities shown.

However, this does highlight our progress on building our systems business over the next few years.

I'll now turn the call over to Alex, who will provide more detail on our second quarter financial results and discuss updated guidance.

Thanks, Mark.

Before discussing the quarter in detail, there are some key points to note as it relates to the first half of the year.

From the outset of the Series 6 transition, we anticipated that this time frame would be the lowest point of our earnings power due to lower module production levels and elevated startup expenses and ramp costs.

In Q2, these expected elements combined with a quarter of unusually low sales.

This is due to both the aforementioned Series 6 throughput and yield issues which impacted module cost and availability for projects already sold as well as to a delay in closing certain new project sales.

Despite these challenges, we've been able to maintain our revenue and earnings per share guidance for the year while implementing plan to fully address these early-stage manufacturing ramp challenges.

It's also worth noting that on our February earnings call, we guided to an expectation of approximately 25% of our full year earnings being recognized in the first half of the year.

With year-to-date EPS of $0.32, we are tracking slightly behind our expected earnings distribution across the year but remain on track to achieve our full year earnings per share guidance.

Turning to Slide 9. I'll start by discussing selected income statement items for the quarter.

Q2 net sales were $309 million, a decrease of $258 million compared to the previous quarter.

Systems revenue as a percentage of total quarterly net sales decreased slightly to 66% in Q2 versus 72% in Q1.

As indicated, the lower net sales in Q2 resulted from certain project sales pushing out of the quarter, lower revenue recognition on projects already sold, and a decrease in third-party module sales due to shipment timing.

As it pertains to the timing of system sales, we indicated on the Q1 call there was a potential for a significant impact in Q2 earnings if the closing of certain project sales moved out of the quarter.

While we saw delays in the sale of 2 projects, in both cases, this is only a timing impact, and in neither case do we anticipate any impact to overall project economics as a result of this change in timing.

With regards to revenue recognition on projects already sold, various issues with the California Flats project were the main reason for lower-than-expected revenue in Q2.

First, Series 6 module availability constraints due to both throughput constraints and delays in the release of module shipments pending final product certifications limited the amount of work that could be completed on the project.

Secondly, there was a small decrease on the project size due to lower-than-planned module bin classes.

While we'll still be installing the same number of modules as initially planned the lower wattage per module results in a lower total DC capacity.

Thirdly, the cost planned for the project was increased due to both higher module costs and acceleration costs resulting from the timing of module shipments.

The combination of these factors served to decrease total expected project revenue and increase total expected costs, which reduced both the project percentage of completion from work performed in Q2 as well as leading to a Q2 adjustment of the project-to-date revenue recognized.

Second quarter gross margin was negative 3%.

The module segment was impacted negatively by low sales volume and ramp-related costs.

Bear in mind that the module segment sales is composed entirely of Series 4 volume as early Series 6 volume is entirely allocated to our systems business.

However, the module segment COGS is comprised of both Series 4 COGS and Series 6 ramp-related costs.

These are allocated to the module segment.

In Q2, the module segment was burdened by over $20 million of ramp costs as well as several million of scrap charges related to initial Series 6 production.

Note that for the full year, we still expect ramp costs to be approximately $60 million.

And during the period where we're ramping on new technology, we expect to see continued noise between our 2 reporting segments.

The system segment gross margin was affected by the change in estimate of the California Flats revenue and cost plan mentioned earlier, as well as the higher mix of revenue from EPC projects versus development assets.

Q2 operating expenses were $96 million, a decrease of $3 million compared to last quarter.

Plant startup expenses decreased by $13 million as a result of lower Series 6 preproduction activities in Ohio, partly offset by increases in the Malaysia and Vietnam factories.

The decrease in startup expense is partially offset by higher SG&A.

Our Q2 operating loss was $104 million compared to an operating profit of $74 million in the first quarter.

The Q2 loss is primarily the result of the unusually low sales, impact to gross margin for over $20 million of ramp cost and more than $24 million of plant startup expense, with an income tax benefit of $6 million in Q2 as compared to a tax expense of $12 million in Q1.

As it relates to U.S. tax reform enacted last December, we did not record any adjustments in Q2 related to our original estimates.

However, as a reminder, we continue to evaluate our provisional estimates until we file our 2017 federal tax return later this year.

The sale of our ownership interest in 8point3 closed in the second quarter, and we recorded a gain on the sale that resulted in Q2 equity and earnings net of tax of $40 million.

The combination of the aforementioned items resulted in a net loss for the second quarter of $0.46 per share compared to earnings per share of $0.78 in Q1.

Moving to Slide 10.

I'll next discuss select balance sheet items and summary cash flow information.

Our cash and marketable securities balance ended the quarter at $3.1 billion, an increase of $256 million from the prior quarter.

We had a record net cash position of $2.7 billion at the end of Q2, a sequential increase of $238 million.

The higher cash balance is primarily due to proceeds from the sale of our interest in 8point3, partially offset by capital expenditures to support our ongoing Series 6 capacity expansion.

Turning to the sale of 8point3.

We received net proceeds of $240 million after the payment of fees and other amounts.

In addition, we collected the remaining outstanding balance of $48 million associated with the promissory note that was issued when interest and the debt of Stateline project was sold to 8point3.

Second quarter net working capital, which includes the change in noncurrent project assets and excludes cash and marketable securities, decreased by approximately $230 million.

The change is primarily due to the collection of accounts receivables and an increase in deferred revenue from module prepayments, partially offset by an increase in inventories.

Total debt at the end of the second quarter was $456 million, a net increase of $18 million from the prior quarter.

The increase was primarily associated with issuing project level debt in Australia.

And as a reminder, essentially all of our outstanding debt is project-related and will come off our balance sheet when the projects are sold.

Cash flows from operations were $129 million due to the collection of accounts receivable and the receipt of module sales prepayments.

Cash received for the sale of our interest in 8point3 and the repayment of the Stateline promissory note were classified as investing cash flows.

Capital expenditures were $195 million in the second quarter compared to $178 million in the prior quarter.

The cumulative spend on Series 6 capacity is now approximately $800 million out of the total expected spend of around $1.8 billion to 6.6 gigawatts of capacity.

And lastly, depreciation and amortization expense is $30 million in Q2 versus $24 million last quarter.

Continuing on to Slide 11, I'll next discuss our updated 2018 guidance.

Before discussing the specific updates, there are some key points and assumptions to highlight.

Firstly, we have narrowed our sales guidance range to reflect the impact of some systems revenue recognition moving into 2019.

As mentioned previously, this timing of new project sales has no expected impact on the overall economics of these projects.

We're lowering our expected gross margin range to reflect 2018 cost impacts, including Series 6 cost per watt increases mostly associated with aluminum costs for the module frame, as well as increased BOS costs.

And we see offsetting nonoperating impacts result in maintaining full year EPS guidance.

Secondly, it's important to reiterate certain risks we highlighted on our last earnings call with regards to our Ishikawa project in Japan.

Our full year guidance continues to assume the project is sold in 2018.

As mentioned on our previous earnings call, the project experienced weather-related construction delays earlier this year from which it has not fully recovered.

We continue to work through a mitigation plan and to see progress in construction, but there remains substantial risks as to whether the sale will be completed this year.

Given the size of this project and the expectation of sale and therefore initial revenue recognition will occur near to or at COD of the project.

We believe it is prudent to highlight the risk.

If the project sale does move into 2019, we continue to expect there would not be any change in the anticipated project economics.

However, this change in timing to a 2019 sale would likely result in 2018 revenue and earnings near the low end of our guidance ranges.

And thirdly, as it relates to the distribution of earnings between the third and fourth quarters, we expect Q4 to be the strongest quarter of the year from a revenue and earnings standpoint.

We expect the remaining earnings for the year to be split approximately 1/3, 2/3 across Q3 and Q4.

Having discussed some of the key assumptions underlying our guidance, I'll now cover the specific updates to the ranges.

Starting with net sales, we're narrowing the range to a revised forecast of $2.5 billion to $2.6 billion in order to reflect the revised timing of revenue recognition on certain systems project.

Note this is not an overall reduction to expected systems revenue, but rather a shift in timing between 2018 and 2019.

Our expected gross margin has been lowered by 100 basis points to a revised range of 20.5% to 21.5%.

Reduction accounts for the increase in module cost per watt and changes to the California Flats revenue and cost plan discussed.

The operating expense forecast, which includes plant startup, has been lowered by $10 million to a revised range of $390 million to $400 million.

Plant startup expense is unchanged at $120 million, and the reduction is a reflection of our ongoing management of core operating expenses.

Our outlook for operating income has been brought down by $15 million at the midpoint to a new range of $120 million to $160 million as a result of the lower revenue and gross margin, partially offset by the reduction in operating expenses.

Below operating income, we've increased our forecast to net interest income, as well as increasing our forecast full year tax expense to approximately $35 million, a result of jurisdictional mix of income.

Our guidance also assumes minimal additional equity and earnings for the balance of the year.

Putting these revisions together, our earnings per share guidance remains unchanged at $1.50 to $1.90.

The operating cash flow range has been increased by $100 million as a result of the revised timing of project development spending and expected improvements in module accounts receivable collection.

As a reminder, both the structural project sales and the timing of the sale can have a meaningful impact on our operating cash flow guidance.

As we discussed last quarter, if we sell a project later in 2018 than anticipated and the project continues to draw down debt financing in intervening periods, operating cash flow proceeds will be lower assuming the debt is assumed by the buyer of the project.

With Ishikawa and other international project sale expected in the second half of this year, we could have some revisions to our operating cash flow expectations even when the economic substance of transactions are unchanged.

Capital expenditures have been reduced by $50 million to a revised range of $800 million to $900 million, primarily due to timing of Series 6 spend and reductions in non-Series 6 CapEx.

As a result of the higher operating cash flow and lower capital expenditures, we're raising our net cash guidance by $200 million to a range of $2.2 billion to $2.4 billion.

And our shipment guidance range has been lowered by 100 megawatts to a revised range of 2.8 to 2.9 gigawatts to reflect the 200-megawatt reduction Series 6 shipments, partially offset by an increase in Series 4.

And finally, turning to Slide 12.

I'll summarize the key messages from our call today.

Firstly, while there have been immediate impacts to module pricing in international markets from the recent policy decisions in China, we remain focused on executing our strategy.

Our differentiated technology in Series 6 products, our strong contracted bookings and our unique financial strength allows us to thrive even in market conditions that may prove challenging for competitors.

Secondly, whilst our second quarter results were impacted by Series 6 ramp-related issues and module availability, factory throughput and higher cost per watt, we've maintained our earnings guidance for 2018 and do not foresee these issues having longer-term impacts to Series 6 cost efficiency or capacity.

With module wattages currently at 420 watts on our top end, improving throughput levels and a third factory that is expected to start production later this quarter, we're encouraged by the positive Series 6 momentum.

And lastly, we continued to make solid progress in booking new business, as evidenced by the approximately 900 megawatts of new volume contracted since our prior earnings call and total future contracted shipments of 10.9 gigawatts.

In particular, with recent PPA awards, we continue to make good headway in building a systems portfolio that we expect to average approximately 1 gigawatt per year over the next few years.

And with that, we conclude our prepared remarks and open the call to questions.

Operator?

(Operator Instructions) And we will take our first question from Paul Coster with JPMorgan.

This is Mark Strouse on for Paul.

So if I could start with -- so you disclosed you booked a little less than 1 gig since the last earnings call.

But are you able to say kind of what the bookings have been since the Chinese policy announcement change?

And maybe if you can't give details on that, just kind of anything high level you can say regarding customers potentially holding off on projects, just kind of waiting to see what the floor in pricing ultimately will be.

Yes.

So if you look at what we highlighted on the earnings presentation, just from the end of the quarter since June, right, so we're 26 days into we booked 400 megawatts of volume.

So the 900, 400 was booked in the month of July.

The 500 before that, a high percentage of it also was booked in the month of June.

So there was only about a month between our last earnings call and when policy decision was made, which effectively was May 31.

So we have been continuing to see good momentum actually with dealing with our head of sales -- Chief Commercial Officer today.

And he's got customers in and given us a list of opportunities to which they need modules for, and we're actively engaging in that conversation for hundreds of megawatts for this particular customer.

So I haven't seen, at least at this point in time yet, a meaningful slowdown.

You can also see it in our mid- to late-stage opportunities that we've highlighted, we still have over 8 gigawatts of opportunities sitting there.

So that hasn't come down, and we're seeing a lot of opportunity on the PPA side.

So we're bidding actively.

We're seeing a number of PPA, particularly here in the U.S., relatively successful on what we've seen so far.

Obviously, it's a very competitive environment for PPAs on the development side and you're not going to have a very high hit rate, but relatively pleased with that activity and we are putting some points on the board as we highlighted on the call with 750 megawatts or so, so far on the development side.

And we've got a number that we've been shortlisted on, and we're in active negotiations with customers to finalize some PPAs that we'll hopefully be able to report on the next earnings call.

So generally, it's still been pretty good.

Now that's the U.S. market.

As you get outside the U.S. market, I would say there's probably more of a pause of wait-and-see maybe a little bit.

Clearly, when -- after the announcement was made, we saw ASPs drop very quickly.

Started seeing them stabilize a little bit.

But clearly, there are some customers now that probably will wait and sort of see what plays out over the next couple of quarters and see where PPA prices go.

The nice thing about us, we don't necessarily have to engage.

If we have an opportunity with the module and the environment that we're very well positioned because of energy advantage with our temperature and spectral response advantages, we'll engage opportunistically and selectively, and we'll make sure we get the right ASPs.

But we're in a good position right now relative to the uncertainty of the market.

Our next question will come from Philip Shen with Roth Capital Partners.

In your prepared remarks, you guys had talked about some issues for throughput and yield.

Wanted to see if you could provide a little more color on each.

So as it relates to throughput, you talked about Ohio I think being at 60% and Malaysia being at 40%.

And I know you plan to be at 100% by year-end, or at least that's what it sounded like based on what you had been saying.

But could Ohio or Malaysia be at 100% earlier?

So could we see that perhaps in Q3?

And it sounds like it's a framing back end issue there.

As it relates to yield, sorry, Mark, you had mentioned that the fleet average is 415 watts now.

Do you expect the fleet average -- what do you expect it to be in Q3 and in Q4?

And how do you expect your efficiencies to progress?

Because we had been, in some of our checks, seeing that you're actually perhaps improving faster than expected, but wanted to get a feel for -- if there's a step function change that we could see ahead?

Or should we expect a more kind of continuous, kind of gradual improvement here?

Yes.

I'll take the throughput question first.

And again, the issue that we're having is really around the back end and it's related to the availability of the toolset, right?

So when you look at the toolset, and if you say, what are the most critical components to ensure full entitlement of the nameplate capacity that we need to achieve, there's really 3 components.

And when I look at it from the standpoint of order of importance, the first 2 and most important is really going to be cycle time on the tool and performance on the tool.

Those are critical.

So if we aren't getting cycle time and if we aren't getting the performance out of the tool, there isn't a lot that we can do to try to help enable that, right, other than redesign the tools or other issues that we have to think about that sort of address both of those.

Cycle time and performance around the toolset, from the front-end all the way through the back-end, is at or better than our expectation.

So that's extremely important.

The third component that we look to for the toolset is the overall availability.

And the overall availability at mature state thus will specify capability around the tool.

We believe the overall ability will still enable us to get to where we need to, but we're not at a mature state yet with the toolset.

And the way we manage through that is we'll have -- we're putting buffers into the back-end of the manufacturing process.

So think about it as we have a single point of failure on the production line today.

If that tool goes down, everything upstream is shutting down and we're starving the downstream processing, right?

So that tool is critical because at a single point of failure, we have to have the availability, and we've identified where those points are, and what we're going to do is we'll buffer it with inventory.

So if a particular tool goes down, we aren't starving the balance of the production and we can continue to run the production as an example, right?

Now when I look at the tool capability, if I look at Perrysburg as an example, we've had a number of times where we are running at, effectively, 90% to close to 100% of nameplate capacity at that particular time, and we measure it in one hour increments.

So we'll look at our increments of production and we'll say what is the output that we achieved during that particular hour.

We can have multiple hours, 2, 3 even 4 hours where we're running effectively at nameplate capacity but then something goes down.

And as soon as that single point of failure occurs, you'll see the production go from almost 100% of capacity down to 20% of capacity, right?

And so that's what we're working through, and we redesigned and we put some buffers in the back end of the line that will help address that and will enable us when we do have an event occur as we continue to ramp up the toolset availability to its full entitlement, that we won't have the adverse impact that we're starting -- that we're seeing right now as it relates to throughput.

And we're working through that right now.

We won't have all of the buffers and the inventory in place until probably end of Q3 more or less, beginning of Q4.

So there's -- we will continue to see a little bit of a headwind, and that's why we've reflected that in the reduction to the production plan.

And then once we do, Perrysburg will replicate all that as we roll out into KLM and Vietnam.

So that's the story about throughput.

And we will get to where we need to be at the end of the year.

It's just a matter of addressing the issues that I highlighted.

As it relates to efficiency, we're starting to touch 425 watt bin, which is really important.

And we will continue to see from the average right now of 415, we'll see that step up and then to 420, and then, ultimately, up to 425 to get closer to the average.

I will say one thing though.

The impact of the throughput -- we have not prioritized.

We do -- we refer to them as ETAs, so engineering tests that we do.

We have prioritized throughput to ETAs, and ETAs will also be helpful as we optimize and drive efficiency up.

So we haven't got the full foot on the gas pedal yet on all the activities that we need to do to drive the efficiency side of the equation, because, again, we're prioritizing the throughput over running some ETAs that will help us on the efficiency side.

But again, steady progression.

Real happy with 415 resting towards 420 as we get into the next quarter.

Love the fact we're starting to touch 425, but I think that's how we ought to think about it as we move through the balance of the year.

Our next question will come from Brian Lee with Goldman Sachs.

Maybe the first one is just on the manufacturing.

Given the pricing collapse you referred to, Mark, and the fact that Series 6 is pricing higher than Series 4, just wondering if it makes sense or if you're contemplating any shifts, specifically to Malaysia, one -- the one facility you haven't committed to a timeframe for shifting from Series 4 to Series 6. Does that potentially get accelerated and come offline sooner given the cyclical dynamic we have here?

And then just a follow-up would be around just cost competitiveness here.

Again, on that same topic, we're seeing global module ASPs trending toward the mid-$0.20 per watt range.

I think there's a general assumption in the marketplace that your targeted cost per watt for Series 6 will be in the low $0.20 per watt when fully ramped in mid to late 2019.

Correct me if I'm wrong on the timing.

But what are you kind of thinking in terms of what your cost advantage versus peers looks like given real-time pricing?

Has that potentially shrunk versus your original base case assumptions?

Yes.

So [let's talk, too] on prioritization, how we think about Series 4 production and KLM 1, 2. We are looking -- we are continuing to reassess and evaluate, not for the horizon through 2020.

What we're looking and spending time on is how do we best position the most competitive posture that we can have as we enter into 2021.

And so what I'm thinking -- what we're thinking through is what are all the critical dependencies knowing the uncertainty that 2021 could have it.

As we continue to book our volumes up through 2020, we're looking across that horizon beyond 2020 into '21 in particular.

And ideally, we're going to want as much Series 6 production as possible.

Scale is going to be important.

We got to get to the efficiency and the cost entitlements where we need to be, and Series 6 is going to be a critical enabler of that.

So as we think through those various levers, we'll continue to evaluate our current commitment around Series 4 and the time line in which we'll run production, and particularly KLM 1, 2, relative to opportunity to drive more Series 6, volume into 2021.

So still to be determined.

We haven't made any conclusions.

We're happy with what we have booked for that business right now, but -- for that volumes, but we'll continue to evaluate that.

Again more from how do we best position ourselves for success long term into 2021 and beyond.

As it relates to cost competitiveness, what we assumed when we did our analysis, and again, I also want to make sure that when you think about that $0.25 number that you referenced, you got to add $0.02 or so, compare that to our $0.20 or what people think the numbers are in the low 20s, right?

So let's -- or you take $0.02 off my number.

I don't care how you look at it, right?

If it's $0.25 for them, we're at $0.18; or if you're targeting us at $0.20 and their $0.25 becomes $0.27.

So make sure that's into your math.

And a lot of times, people don't always include the logistics cost and the warranty cost.

That number is relatively in line, maybe $0.01 or $0.02 lower than what we had assumed when we did our business case around Series 6 and where we thought we could get in the competitive position, that was great for us.

And the other side of that equation you got to keep in mind is the energy upside.

So we'll capture it and we'll be cost advantaged, and we'll have the energy upside, and we'll capture the value on that side of the ledger as well.

So yes, the markets (inaudible) and we're positioning ourselves for long-term success, and we're very happy with what Series 6 will enable for us.

Our next question will come from David Katter with Baird.

This is Ben for David.

I just want to make sure I heard something correctly.

So you said you signed the new bookings which were small.

Maybe you could talk why there's -- or I guess because they're so far out in the future, but they were at the same ASP as booked before Series 6. Is that what you said?

Ben, it's really hard to hear.

What we said -- the comments I made in my prepared remarks were the bookings that we're seeing right now on year-to-date on Series 6 is about 6% higher than what we have booked on Series 4. So think about it as adding $0.02 or so, right?

If you think about where the ASPs are.

The ASPs on Series 6 is about $0.02 or so higher than Series 4. And remember, Series 6 is about 40% lower cost.

So when you combine the 2 and you look at the margin entitlement, the fact we're capturing the upside on ASP, which we knew we would, with a larger form factor and quality of the product and still enable that lower cost of entitlement, I think we're very happy with what we're seeing so far.

The other thing I said is that the ASPs that we're recognizing in '18 are essentially consistent with what we recognized in '17.

But the volume that we're booking this year is carrying us further out.

We're booking '18 volumes that are carrying us into '20.

We didn't book as much volume in '17 for '20 shipments.

And so as you would've normally expected, you look further out in the horizon, ASPs would churn down a little bit.

But I'm happy that we got a relatively consistent ASP in '18, similar with what we have in the '17, and we've carried that profile of shipments all the way through 2020.

Our next question will come from Jeff Osborne with Cowen and Company.

Two questions.

One is on if you can just address the impact of the IRS extension for both your pipeline and your customers?

And then any comments on either debookings or contract renegotiations that your customers have with you?

Yes.

So I guess on the ITC extension, look we think it's probably a positive in the long run for the system business.

I think in the longer-term, it may delay some of the utility ownership of solar as you're going to see a continued competitiveness of PPAs with ITC versus utility-owned generation and rate basing, and that's where you're going to have a delay of a transition which may actually delay -- could be a longer, more optimal capital structure moving away from tax equity which is less efficient and more expensive over to a more traditional infrastructure financing option.

But I think in the short term, yes, we see it as a positive.

As it relates to the customer and the contracts, nothing new there in terms of what we said before.

There are obligations between both parties, there are security associated with it, there's termination penalties associated with it.

I think we look at it again, I think over 90-plus-percent, 95%, maybe higher than that, of our contracted pipeline of module sales, which I think is around 8 gigawatts, all has security associated with it.

Again, I think the sphere in which we negotiated these contracts with our customers was, again, somewhat a risk-sharing and an understanding of fair economics that would enable their projects to be successful.

And that's the way we're moving forward, and customers are honoring those obligations as well.

And our final question will come from Colin Rusch with Oppenheimer.

Could you talk a little bit about the competitive dynamics in the development business, particularly as it relates to integration of energy storage and what you're seeing in terms of pricing from your competitors and how difficult it is to close projects at this point?

Yes.

So development is, obviously -- it can be -- it depends on where you are; it depends on how the RFP is structured; it depends on what bid bonds have to be posted in order to bid, what dollars are at risk; the tenor of the PPA, a 15-year versus 20 or 25.

I mean, every opportunity on the development side is -- can differ relative to its attractiveness and relative to its competitiveness, relative to how aggressive things can be.

And we've got to be very selective in that regard, because you can get into certain opportunities where you've got developers -- especially if it's a free option and you got developers that are just going to make crazy assumptions around the installed cost or they're going to assume some huge hockey stick on our merchant curve.

And our merchant curve may have an assumption of a carbon tax embedded in it, or it may have an assumption of storage already incorporated, even though the asset doesn't have the capability to create firm power.

I mean, there's some -- all kinds of dimensions and flavors that can happen that are out there that can really drive some really aggressive assumptions.

What developers are only worried about is capturing that PPA and flipping it to somebody else, and then let them worry about it over the long run.

They'd make their money and they move on, right?

So it can be very competitive, and we've got to be very selective with where we play and how we play, and site selection and interconnection positions are -- can be critical at times.

And then -- so we have been selective.

And that's also why I've said in my comments that I don't want a very high win rate on development.

If we're winning a very high percentage of development, we're probably taking on a lot of risk, and we don't want to do that, right?

As it relates to storage, again, given what's going on and what you hear, like similar to dealing with the deal with APS, other deals that are happening with Xcel in Colorado and some of these views of PV-plus-storage and the capability of time shift incurring firm power, it's becoming more and more mainstream in most of the utility RFPs that we're doing.

Whether it's PPA or whether it's rate base or whatever else it may be, it's becoming more and more commonplace and we would expect that to happen.

And that's a good thing, because now, it just further expands the market opportunity for storage.

Now we still believe that there can be an interim step that with creating flexible storage using design reserves capabilities that we have, that you can actually have a much higher penetration of PV before you get into serious issues and need for storage, but we'll give the utilities a choice.

We can demonstrate flexibility, and we've done a work recently with a consultant with one of our large utilities, and we've demonstrated to them the full capabilities of flexible PV and how it can honestly drive down operating cost.

And I think that was an insightful study that was done.

We're hopeful we can actually make that a public announcement here near term.

And I think other utilities will open up their perspectives around PV and whether they go straight to storage or going more to a flexible storage platform with line reserves, the options will be there.

And ladies and gentlemen, this does conclude today's conference.

Thank you all for your participation.

You may now disconnect.