第一太陽能 (FSLR) 2018 Q4 法說會逐字稿

Good afternoon, everyone, and welcome to First Solar's Q4 2018 Earnings Call.

This call is being webcast live on the Investors section of the 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.

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

Today, the company issued a press release announcing its fourth quarter and full year 2018 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 full year and provide the latest updates around 2019 guidance.

Following their remarks, we will 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 briefly discussing our EPS results for 2018.

EPS of $1.36 came in slightly below the low end of the guidance range we provided at the time of our Q3 earnings call.

While Alex will provide a more comprehensive overview, I wanted to highlight 2 items that had a material impact on the quarter.

Firstly, late in the year, we incurred increased EPC costs in order to meet deadlines for certain U.S. projects.

Inclement weather and delayed shipments of materials to the site adversely impacted planned construction and project commissioning schedules.

The potential of project completion delay was particularly acute at one of our projects in California.

To ensure the project's capital structure proceeded as planned, we incurred significant acceleration costs to meet key schedule milestones.

While the project owner shared in a portion of these costs, acceleration costs impacted Q4 results by more than $10 million.

Maintaining the strong relationship was a key priority, and therefore, we made an investment in our partnership and long-term relationship with this customer.

Secondly, in Q4, we continued to make good progress with our Series 6 factory construction, start-up and ramp.

As a result, we started production at our second Vietnam factory the first week of this year, 3 months ahead of our original plan and 45 days ahead of our latest expectation.

The continued factory ramp across all sites, combined with the earlier-than-planned start-up of our second Vietnam factory, put pressure on our supply chain to support the accelerated schedule.

To maintain continuous operations across the entire fleet, we decided to airfreight certain raw materials to our factories, which adversely impacted the fourth quarter by more than $10 million.

Accelerating the Vietnam start date helps to provide resiliency to our 2019 Series 6 production plan.

The production could lead to additional revenue, but more importantly, it creates optionality for downtime investments to increase throughput via tool upgrades or production buffers or to run engineering test articles to increase module efficiency.

Turning to Slide 4, I'll provide additional comments on 2018.

Despite a year where the solar market faced excess capacity and pressure on module pricing, primarily as a result of policy changes in China, we were able to make steady progress and strengthened First Solar's competitive position.

In 2018, we added to our contracted pipeline with strong net bookings of 5.6-gigawatt DC, a greater than 2:1 book-to-ship ratio, which provides improved future visibility as we grow our Series 6 production over the coming years.

Business projects were a significant portion of these bookings, and we signed 1.3-gigawatt DC of new PPAs last year.

In addition, we added EPC scope to 500 megawatts of previously booked module sales, which, combined with our development bookings, positions us to meet or exceed our targeted 1 gigawatt per year of systems business.

Our 2018 bookings were -- also highlighted the strong demand for utility-scale solar from C&I customers.

Approximately 500 megawatts of our total 1.3 gigawatts of development project bookings were PPAs signed with utilities, where corporate customers are the intended consumers of the energy to be generated by these projects.

Additionally, this trend has continued into 2019 with our recent booking of a nearly 150-megawatt PPA with a corporate customer.

We expect corporate demand for solar projects to continue to grow in coming years, and we believe that our strong reputation and ability to offer turnkey solutions will position us to compete effectively for future opportunities.

International wins were a meaningful portion of our 2018 bookings with more than 700 megawatts booked, primarily in Europe.

While strong domestic demand for our Series 6 product has limited our ability to support international market opportunities, we expect international bookings to grow as we continue to invest in our regional sales team and add planned Series 6 capacity.

2018 was a record year for O&M bookings as we added nearly 3.5 gigawatts of new business, bringing our total O&M fleet under contract to over 11 gigawatts at the end of the year.

We remain encouraged by the opportunities to continue growing O&M and to leverage the fixed costs associated with this business.

From a manufacturing perspective, we made progress starting and ramping Series 6 capacity over the course of 2018.

During the year, we started production at 3 Series 6 factories, which collectively manufactured a combined 0.7-gigawatt DC of modules.

The production run rate at these factories at the end of 2018 was over 2 gigawatts, which is a significant achievement considering initial production did not begin until April.

Construction of our fourth Series 6 factory was completed in late 2018 and recently started production.

Lastly, a fifth factory is under construction and progressing according to plan with an anticipated start of production in January 2020.

To concurrently manage all the activities related to the construction, start-up and ramp of the 5 different factories was a major undertaking that has positioned us to meet our strong demand for Series 6 in 2019.

Also of note is, in late 2018, we reached the 20-gigawatt shipment milestone.

This reflects cumulative shipments since the founding of First Solar and highlights the extensive deployment of our cad tel technology worldwide.

Overall, our operational and financial results in 2018 have created a solid platform as we move into 2019.

Turning to Slide 5, I'll next discuss our most recent bookings in greater detail.

In total, our net bookings since the prior earnings call in late October were 1.6 gigawatts, including 1.3 gigawatts which were booked since the beginning of January.

After accounting for shipments of approximately 900 megawatts during the fourth quarter, our future expected shipments, which now extend into 2023, are 12.1 gigawatts.

Our most recent bookings include 2 PPAs that were signed, totaling more than 300-megawatt DC.

The first of these PPAs was signed with MCE for the expansion of the Little Bear project in California.

The second PPA was signed with a major utility customer in the Western United States, and the project will support collaboration between the utility and its corporate buyers to meet their renewable energy objectives.

Included in our new module bookings is a greater than 1-gigawatt agreement with a major customer in the United States for shipments in 2021-and-beyond time frame.

This booking highlights the continued strong demand for Series 6 in the United States, particularly as certain customers look for opportunities to safe-harbor modules to preserve the higher ITC.

While we are pleased with our 2018 bookings of 5.6 gigawatts and the greater than 2:1 book-to-ship ratio, it is important to put our 2019 bookings expectations into perspective.

Relative to our module competitors, we are in an extremely favorable position essentially being sold out over the next 8 quarters.

Generally, our customers, particularly in international markets, do not contract for module supply multiple years in advance given the project development cycle and the time horizon in which they have project certainty.

While we're encouraged by our bookings year-to-date and target a 1:1 book-to-ship ratio in 2019, our bookings may be more back-end loaded given our available supply is in the 2021-and-beyond period.

On the O&M side, as we highlighted earlier, in 2018, we added nearly 3.5 gigawatts of new projects.

A high percentage of these bookings was attributed to third-party wins, defined as projects where we are not the developer but, in which many cases, include our module technology.

Third-party O&M not only expands our addressable market but also helps to create economies of scale for our O&M business.

Some of the reasons for our continuing success in winning third-party business are highlighted by an example of how we were able to leverage our O&M expertise to address a customer's need in a way our competition was not able.

In 2018, we were approached by a customer seeking help with 2 large utility-scale solar power plants in its portfolio that were under contract with a competing O&M provider and were underperforming.

These projects utilized a competing module technology and were not constructed by First Solar EPC.

Based on the customer's experience with our O&M services, they asked us to investigate the cause of the underperformance.

By leveraging our industry-leading expertise of our O&M team, we identified the root cause of the underperformance and created a detailed action plan to improve performance.

The recommended corrective actions are expected to improve the annual energy output of the combined plants by approximately 3%, which translated into more than $1 million of annual revenue to the owner.

As we continue to leverage our significant O&M experience to meet customer needs, we expect that third-party wins will continue to be a key part of our growing O&M fleet.

Side 6. I'll provide an updated view of our mid- to late-stage bookings opportunity, which now total 7.3-gigawatt DC, a decrease of approximately 500 megawatts from the prior quarter primarily as a result of our strong recent bookings.

However, when factoring in the bookings for the quarter, 1.4 gigawatts of which were included as opportunities in the prior quarter, our mid- to late-stage pipeline actually grew by approximately 900-megawatt DC.

North America remains the region with the largest number of opportunities at 5.5-gigawatt DC.

However, Europe has shown a meaningful increase since the prior quarter, driven by resurgent markets in France and Spain.

Opportunities in Asia Pac region have remained relatively stable.

Even with the more than 300 megawatts of recent systems bookings, our potential systems opportunities remain strong at 1.8-gigawatt DC.

These potential systems bookings are comprised of projects in the U.S. and over 300 megawatts in Japan.

Continuing on to Slide 7, I'll next provide an update on our Series 6 capacity rollout.

The most notable achievement to highlight since our prior earnings call is the start of Series 6 production at our second Vietnam factory, our fourth Series 6 factory in total.

As mentioned previously, production commenced in early January, several weeks ahead of our target start date.

Similar to our first Vietnam factory, the initial ramp has been accelerated relative to the previous facilities by applying accumulated learnings, which including starting production with an improved module framing tool.

Construction is continuing at our second Series 6 factory in Ohio.

As announced previously, we expect to start production in early 2020, and construction is on track to our schedule.

Once completed, we will have 5 factories with annual Series 6 capacity of 5.6 gigawatts, an impressive accomplishment since announcing the transition to Series 6 in November of 2016.

Since the third quarter earnings call, we have seen steady improvement in our Series 6 throughput and wattage across our entire fleet.

When comparing February's month-to-date performance to the month of October, you can see the significant improvements made.

Note, our second Vietnam factory is excluded from this comparison as it was not operational in the base comparison period.

Megawatts produced per day is up 65%.

Capacity utilization has increased 30 percentage points.

The production yield is up 7 percentage points.

And finally, the average watt per module has increased 2 bins or 10 watts.

Since October, the percentage of modules with antireflective coating has increased 33 points.

These significant accomplishments can be credited to the outstanding work of our engineering and manufacturing associates.

We are encouraged by the meaningful progress we have made over the last months of 2018 and how we started 2019.

We continue to plan for full year production of between 5.2 and 5.5 gigawatts.

As a reminder, this targeted production includes approximately 2 gigawatts of Series 4 modules.

In order to meet these production commitments, we continue to roll out 2 upgrades and optimize the production line throughput across the various sites.

This is a dynamic process that continues to incorporate learnings from each of the factories as we have ramped and it is moving according to schedule.

I would like to make one final point before I hand the call over to Alex.

As I mentioned in October, First Solar was a sponsor to an innovative study by E3, which highlights the value of flexible solar to utilities in the form of expected reduced fuel and maintenance costs for conventional generation, reduced curtailment of solar output and reduced air emissions.

Since the study has been published, we have been pleased with the positive response and feedback from across the industry.

For example, Public Utilities Fortnightly, a leading industry publication, recognized the study as one of their 2018 Top Innovators.

Our efforts to demonstrate our thought leadership are not only limited to the United States.

Recently, we supported a study by SolarPower Europe that provides evidence to support the benefits of utilizing low-cost, utility-scale solar to keep the European grid stable and reliable.

Efforts such as this will take on increasing importance in order for the European Union to meet its 2030 renewable energy targets, and we look forward to remaining engaged in that process.

Whether in United States, Europe or other regions, we will continue to provide support and thought leadership to advance the understanding of how utility-scale solar enhances the reliability of power grids around the world.

I'll now turn the call over to Alex, who will provide more detail on our fourth quarter financial results and discuss updated guidance for 2019.

Thanks, Mark.

Before reviewing the financials for the quarter in detail, I'll first provide additional context around the factors that led to the 2018 results falling below our guidance.

There were 4 key issues that impacted our ability to meet earnings guidance.

Firstly, 2018 net sales were $100 million lower than the midpoint of our guidance due to the timing of module sales and delays in systems revenue recognition.

The lower systems revenue was associated with inclement weather and also material delivery delays in some projects.

Secondly and thirdly, as Mark mentioned earlier, we experienced increased EPC costs across several U.S. projects, partially driven by schedule acceleration to achieve year-end customer milestones.

And we experienced elevated inbound freight costs to expedite raw materials of Series 6 production.

And fourthly, 2018 ramp and related costs were $113 million compared to our guidance of $100 million.

So with that context in mind, I'll begin by discussing some of the income statement highlights for the fourth quarter and full year on Slide 9.

Net sales in the fourth quarter were $691 million, an increase of $15 million compared to the prior quarter.

The higher net sales were primarily a result of the sale of 2 projects in Japan.

For full year 2018, net sales were $2.2 billion.

And as mentioned, relative to our guidance expectations, net sales were lower due to the timing of both module sales and delays in system revenue.

As a percentage of total quarterly net sales, our systems revenue in Q4 was 83%, which was nearly flat compared to Q3.

For the full year 2018, 78% of net sales came from our systems business compared to 73% in 2017.

Gross margin was 14% in the fourth quarter and was impacted by ramp charges of $44 million as well as inbound freight costs and EPC acceleration costs.

For the full year, gross margin was 18% and included $113 million of ramp and related charges, which equates to a 5 percentage point impact.

The systems segment margin was 22% in the fourth quarter, and the module segment margin was a negative 25%.

And as it relates to the module segment gross margin, keep in mind that sales were composed almost entirely of Series 4 volume as Series 6 volume continues to be allocated almost entirely to our systems business.

However, the module segment cost of sales was composed of both Series 4 cost of sales and Series 6 ramp-related costs of $44 million.

Adjusting for the impact of ramp-related costs, Series 4 module gross margin was in line with our expectations.

Operating expenses were $87 million in the fourth quarter, an increase of $17 million compared to Q3.

Q3 OpEx benefited from a reduction to our module collection and recycling liability, while Q4 was impacted by higher SG&A from project-related expenses.

For 2018, operating expenses were $352 million, near the midpoint of our guidance range.

Now highlighting our efficient management OpEx in 2018.

Our combined SG&A and R&D expense decreased approximately $30 million or 10% versus 2017.

Operating income was $11 million in the fourth quarter and $40 million for the full year.

Compared to our guidance for the year, op income was lower than planned as a result of the lower revenue and higher cost of sales, as discussed.

Other income was $32 million in the fourth quarter from the gain on sale of certain restricted investments.

Investments sold were associated with our module collection and recycling program and were sold in part to reimburse overfunded amounts.

Note that a smaller sale of restricted investments for similar purposes was completed earlier this year in 2019, and we'll reflect it in our first quarter results.

We recorded a tax benefit of $4 million in the fourth quarter.

For the full year, we recorded tax expense of approximately $3 million.

Fourth quarter earnings per share was $0.49 compared to $0.54 in the third quarter.

For the full year, earnings per share was $1.36.

EPS was below the low end of our guidance range due to the timing of revenue recognition for certain module and system sales and the higher EPC airfreight and ramp costs discussed earlier.

I'll next turn to Slide 10 to discuss select balance sheet items and summary cash flow information.

Our cash and marketable securities balance at year-end was $2.5 billion, a decrease of approximately $183 million from the prior quarter.

Our net cash position decreased by a similar amount to $2.1 billion, at the midpoint of our guidance range.

The decrease in our cash balance is primarily related to capital investments in Series 6 manufacturing capacity, factory ramp activities and the timing of cash receipts from certain systems project sales.

Total debt at the end of the fourth quarter was $467 million, virtually unchanged from the prior quarter.

Within the quarter, project debt issued to fund another project construction in Japan and Australia was essentially offset by liabilities assumed by the buyers of 2 Japan projects sold.

Nearly all of our outstanding debt continues to be project related and will come off our balance sheet when the projects are sold.

Net working capital in Q4, which includes the change in noncurrent project assets and excludes cash and marketable securities, increased by $178 million versus the prior quarter.

The change was primarily due to an increase in module inventories, which is related to our capacity ramp and unbilled accounts receivable.

Cash flows used in operations were $186 million in the fourth quarter and $327 million for the full year.

As a reminder, when we sell an asset with project-level debt that is assumed by the buyer, the operating cash flow associated with the sale is less than if the buyer had not assumed the debt.

In Q4, buyers of our projects assumed $124 million of liabilities related to these transactions, and for the full year, that total is $241 million.

Capital expenditures were $129 million in the fourth quarter compared to $238 million in the prior quarter due to the timing of spending on Series 6 capacity.

For the full year, capital expenditures were $740 million compared to $662 million invested in Series 6 capacity expansion.

Cumulatively, Series 6 expenditures incurred at the end of 2018 were $1.1 billion.

Continuing on Slide 11, I'll next discuss the updated assumptions associated with our 2019 guidance.

We're largely maintaining our guidance ranges for the year, with minor adjustments to ramp and start-up costs, which have an offsetting impact on gross margin and operating expenses.

Whilst these changes are relatively small, there are a couple of important points to highlight.

Firstly, there's been recently significant focus around the PG&E bankruptcy and impacts to companies that have contracted offtake agreements with PG&E.

First Solar has one 75-megawatt AC project where PG&E is the contracted offtaker.

However, we believe any risk associated with this asset is limited given the project size, total development capital invested to date and competitive PPA price.

Where First Solar could potentially have greater exposure is on several unsold projects where SCE is the contracted offtaker.

We're currently in the process of marketing some of these assets for sale.

And to the extent that buyers of these projects assume any increased risk premium associated with SCE as the offtaker, this could result in lower project value.

So while we don't see this as a significant risk to the sale of either of these projects, given their competitive PPA prices and the keen market interest in contracted solar assets that we've seen in recent competitive sale processes, it is an item we think should be highlighted.

Secondly, we're lowering our gross margin guidance by 50 basis points to a revised range of 19.5% to 20.5% as a result of higher expected ramp costs.

Offsetting the decrease in gross margin is a $15 million reduction to start-up costs within our operating expense guidance.

The increase in ramp costs and offsetting decrease in start-up costs are a result of the earlier-than-planned start of production of our second Vietnam factory.

The revised range of ramp-related charges is now $35 million to $45 million, and planned start-up is $75 million to $85 million.

Combined ramp and start-up costs of $110 million to $130 million are unchanged from our prior forecasts.

Thirdly, as we emphasized during our December outlook call, the profile of earnings is expected to be weighted towards the second half of the year.

Slide 12 contains 2 charts which illustrate, from a revenue and cost perspective, some of the factors that are expected to impact the quarterly earnings distribution.

In both cases, we are not providing the actual volumes sold or actual module cost per watt, only the relative percentages.

The first chart shows Series 6 third module -- third-party module sales by quarter.

Notably, only 10% of the volumes sold is in the first quarter and only 25% in the first half of the year.

Not surprisingly, as the supply increases over the course of the year, we expect to see the volumes of sales increase in Q3 and Q4.

The second graphic shows the quarterly profile of our Series 6 module cost per watt produced relative to the 2019 full year average.

The data illustrates the cost per watt for the first quarter of 2019, which has the lowest throughput and module wattage levels for the year, which is projected to be approximately 30% higher than the 2019 full year average.

Module cost per watt is expected to improve in the second quarter but will still be 5% higher than the average.

The greatest benefit of our improved ramp and efficiency is anticipated to be in the second half of the year.

In the third quarter, the cost per watt is expected to be 5% below and the fourth quarter 10% below the 2019 full year average.

In addition to the Series 6 sales and cost per watt profile, there are 2 additional factors which we expect to contribute to lower earnings in the first half of the year.

The first is the timing of ramp and start-up charges, which are heavily weighted to Q1 and Q2.

We expect more than $40 million of combined ramp and start-up in the first quarter.

The second factor is the timing of project development sales.

Similar to our expectation at the time of our December outlook call, project development sales are expected to be weighted to the second half of the year.

And we also expect to close the sale of our Ishikawa project in Japan in the fourth quarter.

Taking all of these factors to account points to why we expect both a loss in the first quarter as well as lower earnings in Q2, with the majority of earnings coming in the second half of the year.

For the full year, we still see EPS guidance in the range of $2.25 to $2.75, driven by Series 6 production ramp and cost per watt improvements as the technology continues to scale.

And finally, I'll summarize our fourth quarter and 2018 progress on Slide 13.

First, we had earnings per share of $1.36 and year-end net cash of $2.1 billion.

Secondly, we had continued success adding to our contracted pipeline in 2018, with net module bookings of 5.6 gigawatts.

With year-to-date 2019 module bookings of approximately 1.3 gigawatts, we're off to a positive start for the year.

Thirdly, we continue to make good progress on our Series 6 capacity road map.

Early this year, we started production of our second Vietnam factory ahead of schedule, and we continue to make steady improvements in both throughput and module wattage at our other Series 6 facilities.

Our progress thus far in '18 and in '19 indicates we remain on track to our combined Series 4 and Series 6 production target of 5.2 to 5.5 gigawatts.

And lastly, a 5% net neutral movement between ramp and start-up costs between COGS and OpEx and maintaining our financial guidance ranges for the year, including our EPS range for 2019 of $2.25 to $2.75.

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

Operator?

(Operator Instructions) Your first question comes from Philip Shen with Roth Capital Partners.

Just wanted to check in with you on your shipments to customers now that you're shipping externally.

Some of our checks indicated that you may be falling 5 watts per module short in your shipments to customers versus contractual requirements or obligations, and this may be resulting in extra costs.

We could be wrong on this one, but wanted to just check in with you on this.

Can you comment on whether or not this may or may not be happening?

And if true, can you provide some color on this and perhaps talk about how long the issue may endure ahead?

Yes.

So I think the premise of the question is we want to make clear that falling short of contractual obligations, we're not falling short of any of our contractual obligations relative to commitments to the customers on any of the product which we need to ship to them.

We have -- as we said before, we have bin adders and bin deducters.

So we have a contracted commitment that we anchor around.

And to the extent the bin is actually higher or lower, then there's an adjustment to the price accordingly for that delta, could be up or could be down.

So I just wanted to make sure that that's clear.

There's nothing that we're doing that would say that we're falling short of our contractual obligation.

But to the extent we do deliver a bin that's -- a bin down would be 5 watts, then there would be a bin adjustment to the price.

And that is happening in some cases.

And part of it was -- I think we indicated on prior calls is that the early production in particular, we've been struggling to see the increased penetration of ARC.

And so -- and without ARC, you're going to lose almost 2 bins of volume.

And one of the things that we said on the call was our ARC penetration has increased now 33 percentage points.

So we're seeing a much better utilization for ARC.

And as a result of that, if we go forward and we continue to ramp across the balance of the fleet, some of the early launch issues that we had will be subsiding, and we'll be able to make sure that we hit the committed bin that we initially structured around.

But I want to make sure it's clear and you understand that to the extent that the bin is slightly above or below, the contract allows for that, and there's appropriate adjustments to the ASP.

Your next question comes from Colin Rusch with Oppenheimer.

This is Kristen on for Colin.

You talked a little bit about this in your prepared remarks, but can you provide some additional color on the geographic diversity of the backlog on an annual basis, just sort of the mix of domestic versus international?

And then what opportunities are you seeing to pick up broken projects for the systems business in the U.S.?

Sort of corollary to that, what's the expertise in integrating -- your expertise in integrating solar with storage to your pricing strategy for modules?

Okay, a lot there.

When you look at the geographic diversity of our shipments for -- and some of this will come out in the Q -- I'd say, actually, it will come out tomorrow.

You'll see that there's a high concentration of module shipments that occurred within -- in the U.S. And the range of 70% or so of the shipments last year were in the U.S., and the balance were in international markets.

And again, it's largely reflective of where the strength of the demand is.

And if you look at our pipeline as you carry forward of the 7 -- a little over 7 gigawatts of mid- to late-stage opportunities, about 5.5 of that sits within the U.S. The volumes at which we booked this quarter were largely U.S. We had some volume with a European customer, but most of the, call it, 1.6 since the last earnings call was focused around the U.S. And it largely has to do with where our customers are willing to commit.

And I think it's important to understand that, of the large order that came through this year, gigawatt of the 1.3, that volume is to be shipped in '21, '22 and '23.

You'll see customers in the U.S. because of certainty around the ITC and wanting to safe harbor.

You'll see customers having a greater appetite to commit forward and to procure materials that would go out that far in the horizon.

When you look at some of the international markets, we don't see as many customers willing to start procuring in '21, '22 and '23 partly because they have lack of certainty on the underlying projects for those modules and where they would go.

And so what we said on the call is that partly, the -- when you look at the bookings for the year, we started off great, but we still look to have a 1:1 book-to-ship ratio, which is to say that we're targeting to book somewhere between 5.5 or 6 gigawatts this year.

We may see some of that be more back-end loaded because I mean, I do see more diversity of the bookings as we progress throughout the year, being more opportunities in our international markets because we're getting to a horizon that -- towards the end of '19, and we're looking to ship into customers and starting in '21 that we can see that international customer participating in that opportunity.

So I would expect our bookings, as we progress throughout the year, to improve, having more of a diversity to U.S. versus international.

But at the same time, as long as we are still relatively capacity constrained, while it's important that we continue to grow and develop our international markets, if we have opportunities to capture better value in the U.S. markets, we'll prioritize the U.S. market.

And we may prioritize some of the international markets that give us better opportunities to capture higher ASPs, and we'll focus there first before maybe we chase some of the other markets that we know traditionally have been very low-ASP markets.

On the storage question -- let me go to the systems question first.

I think and particularly in the U.S., there is a lot that's in the market right now.

As you can see, there's a lot of, I'll call it, smaller developers and others that are trying to actively market and to sell their development pipeline.

Some with contracted assets, some not.

And I do think that some of that could be related to the capacity of some of the smaller developers to make the investments to capture the ITC safe harbor.

We indicated in our last call, we'll be investing somewhere, call it, $300 million to $400 million to secure, call it, 5 gigawatts of opportunities between now and 2023.

That's a big investment, and I think some of the smaller developers may be constrained with making those investments.

And I think they understand that if they don't make those investments, they'll be less competitive as they're competing for projects that have CODs that go through the end of 2023.

So I can see a lot coming to market, and we're trying to at least get engaged and evaluate and see if some of those opportunities make sense for us.

And clearly, we've got a great development team, and we've proven ourselves with our ability to make acquisitions and integrate development assets and contract them and realize meaningful value associated with that.

So that's a good opportunity for us.

And then storage, we are actively involved, our largest storage deal that we announced a few quarters ago with APS.

We've got a couple of other projects.

We've been recently awarded a project with a utility in Florida to do a pilot for them, a small addition of storage onto their grid.

We've done some work with a utility in Nevada around the same type of opportunity where customers are exploring and learning and wanting to know more about storage and how it can be effectively integrated.

And it's an area of emphasis and focus for us.

I look at it, it's somewhat an extension to our normal systems business and just part of our offer.

And we can add enhanced value through our power plant controls and optimization of how we charge the battery and dispatch the battery.

And we've proven some capabilities there that has been very interesting to some of our customers in that regard.

So it's still early innings.

We've picked up some wins, and I see more momentum as we move forward as it relates to storage.

Your next question comes from Julien Dumoulin-Smith with Bank of America Merrill Lynch.

Perhaps just to pick up where you left off, if you can clarify a little bit your comments just now about the securing of the backlog here from an ITC perspective.

A, how do you think about that accelerating into year-end '19 given that is the time line that you need to meet to get to qualify that ITC?

And then secondly, I think you alluded to a gigawatt utility customer in the quarter who they were trying themselves to try to lock up some supply.

So maybe as you think about the potential orders from what you haven't locked in from an ITC perspective, is that another source of bookings acceleration into the back half?

Yes.

I'll just start with what we are looking at from a safe harbor perspective ourselves.

And it's similar today to what we've talked about on our guidance call in December.

So we're still looking at somewhere between $325 million and $375 million of spend this year.

We haven't specifically talked about what we're going to spend that on.

It's less likely to be on the module side just given the constraints we have in module supply.

As Mark said, we're largely sold out for the rest of the year.

So we will look at the rest of the balance of plant.

There'll be some projects that we're far enough along that we can use the physical work test.

And so a small piece of that midpoint $350 million number that I talked about will be associated with physical work incurred.

But that will probably be in the range of $25 million to $50 million.

The rest, we'll look to spend on, as mentioned, balance of plant with projects that go out into 2021 from a contracted perspective.

And then on the uncontracted side, some projects go out to be on that 2021 time frame.

The other thing we said from our perspective is that if there is opportunity to spend more and to the point that we are able to pick up projects where other developers are constrained from a capital perspective and securing safe harbor material, it's somewhere we'll be very happy to invest additional capital and believe the returns are good.

So somewhere where if we see the right opportunities, we're willing to spend more than that $375 million top line that we talked about.

And from a customer standpoint, Julien, I mean, the order that we secured here was with one customer.

It's a common conversation that our team is having with a lot of our customers and thinking about the safe harbor and how to -- what their particular strategy is and engaging in conversations with us around that and how we can try to evolve that.

In some cases, on this customer, it's interesting, they already had a commitment with some volume for this year.

So we didn't have to -- it wasn't an issue of not having the supply, but what we were able to do is that simply -- or contractually we had volume on the book for this customer.

We then engaged with them, "Well, let's leverage that as your safe harbor anchor and then commit to volumes that -- out in the horizon and when you will construct the project in '21, '22 and '23." So Alex is right.

We are constrained as it relates to available supply now, starting up Vietnam a little bit faster than -- a little faster to get a little bit of supply.

If we continue to ramp, accordingly, we may see a little bit of opportunity there.

Those are small in the rounding.

The bigger opportunity I see is, how do we talk to customers today that have contracted volume that's on the books?

And then how do we position that as the anchor for the ITC and then contractually commit to the volume that would sit out and deliver in the '21, '22 and '23 time frame?

So we're having a number of conversations with customers in that regard.

Your next question comes from Ben Kallo with Baird.

So I have 3 questions.

First of all, like Slide 12 is kind of confusing.

Could you help me through that?

And then just talk about the cost reductions versus the 40% that you said back at Analyst Day.

Like you are plus or minus $0.01 or $0.02 from there.

Number two, I understand the cost pull-forward, but then I also see megawatts going up.

And then number three, could you just talk about how you're pricing some of these out-year contracts just because we have a hard time going with ASPs that will go out to 2022?

So how do you think about pricing those?

Yes.

To explain the graph in a bit more detail, this here, so the graph on the left-hand side of Slide 12 is showing you the Series 6 third-party volumes.

If you think about the guidance we gave, that 5.2 to 5.5 gigawatts for the year, you take out a couple of gigawatts to Series 4 and then you go take out the systems piece.

So you're left with what is Series 6 true third-party module deliveries.

And when you take that total number, we're saying, "This is the breakdown per quarter of the delivery of those modules." So about 10% of that third-party Series 6 volume is delivered in Q1, 15% in Q2, 30% in Q3 and 45% in Q4.

This is really trying to show that on a third-party module delivery basis, we're back-ending the profile pretty significantly in the year.

On the right-hand side, looking at the costs.

So to the question you had around costs, we talked in the guidance call around long-term -- or end-of-year Series 6 costs being approximately 40% lower than our 2016 benchmark for Series 4 with a roughly $0.01 adder associated with increased costs around the frame.

So if you take that point over the end of the year and say, "That's the year-end point," you can look at what you think the full year average is.

We're trying to make the point that on the average basis for '19, you're going to see whatever that average is be significantly higher in Q1 as soon as that module is delivered.

It comes down to Q2, and if you're trying get to Q3, you're fractionally under the year average.

By Q4, you're 10% under the year average.

So again, it's trying to say, when you combine these 2, the left-hand side lower volume beginning in the year, the right-hand side higher cost relative to the average, you're going to see pretty negative impacts on results for Q1 and Q2, and you start to see that reverse out when you have much higher volume and much lower costs Q3 and Q4.

Yes.

And I think what Alex said there is, Ben, your question about our view around the 40% off of our Series 4 reference point.

But for the $0.01 or so, $0.01 or $0.02, for framing a piece, there's a couple of smaller components.

That's effectively where we anticipate it to be, and nothing's changed there.

And we're working on opportunities where we can even revise the frame and even take more cost out there because between the frame and the strategical asset is really where the vast majority of the build material is.

And the team's working pretty aggressively on finding the road map to figure out how we get everything back to the full entitlement of what we have.

And there's some encouraging work being done from that standpoint.

The other thing I'll say about that slide is that one of the biggest levers that moves you from where -- whatever the number is, 20%, 30% higher in the first quarter versus the average, and that trends down to being 10% lower than the average, a big piece of that is the throughput, right?

Because there's still a significant amount of underutilization that sits in the first half of the year.

And then as we drive that utilization down, we're at full entitlement across the entire fleet because we're starting up another factory now.

And so we're going to be -- utilization, while it's significantly higher upon March, after the first month or so of production relative to our other factories, it's still going to be driving this down, and there'll be some underutilization cost that's going to be weighing down on the overall average across the fleet.

So that's a piece of it.

And then the other is the efficiency improvement.

So we'll continue to see improvements as we progress from where we are now to the end of the year.

And you'll pick up close to another 2 bins from the launching point where we are right now to the exit rate that is close to that from Q1 to Q4.

So those are the 2 big drivers that are driving that cost forward down.

The contracts for the outer year and the pricing around that, Ben, we look at -- we capture that at fair value, right?

And pricing, as we go out into '21, '22 and '23, we have a road map of where -- we'll know where our costs -- we know where our efficiency is going to be.

We know what the energy advantage is going to be at that point in time.

We'll price it accordingly.

And I'm very happy with the -- we have now quite a bit of volume.

Obviously, a lot of supply that sits out '21, '22 and '23, but I'm pretty happy with the pricing that our team has been able to capture in that window.

It's above what my expectations would have been relative to the business case we put together for Series 6. So we're pretty pleased from that standpoint.

Your next question comes from Brian Lee with Goldman Sachs.

Two for me.

I guess, first on that sort of capacity point, you mentioned in mid-December when you gave the guidance for '19 that you're putting Malaysia 1 conversion to Series 6 on hold.

I mean, you've mentioned capacity constraints, and now you're talking about 2023 deliveries throughout this call.

So given that backdrop, what's sort of the decision process around bringing that back into the capacity expansion road map here?

And then second question is just on Slide 12, super helpful with the cadence.

Alex, can you help us think about how that average line moves into 2020 with some of the utilization effects starting to fall off and then getting fuller entitlement around the efficiency targets and so forth and so on?

I'll take the capacity, and then I'll also take the other one.

So Brian, look, as we said when we -- at the end of this year, we'll ramp down 2 of our factories in Malaysia.

We'll merely start the transition of one of them.

The other one is continuing to be evaluated, and it's really being evaluated based off of market demand and our ability to capture the bookings that we need in 2021 to get to a higher level of confidence or ability to sell through that volume.

And so it's really -- it's demand related, demand driven.

And as we continue to book, then we'll somewhat crystallize our decision around that, and we'll get more and more comfortable.

What I will say, though, is that every one of those factories that comes up, in essence, creates pricing power because it creates scale.

And that scale enables us to enhance our competitive position and then allows us to capture volume in other markets that we may not be participating in today.

So I'm very motivated to get that factory up and running, but it's highly dependent upon our ability to clear the market at acceptable margins.

And as we continue to do that, then I think the likelihood of starting that conversion on that second plant and really be on our third Series 6 factory in Malaysia will start to crystallize.

And Brian, we can't give you guidance that far.

So what I can say, I guess, is that as Mark mentioned that a lot of the cost, majority of the cost sits between the 2 pieces of glass on the frame.

So that's where we're going to be spending a lot of our time.

On both -- so on the frame, we're impacted by the tariff.

We are looking to optimize the frame further, so we add some movements in the frame in terms of design from where we originally came out with Series 6 in some of the modules we produce.

So we're looking at, can we optimize the design to use less aluminum in that frame?

On the glass side, we mentioned in our guidance call in December that we have some projects that we're looking at that may impact start-up.

And one of those that we talked a little bit about was trying to optimize some of the glass where we today pay for a specialized processing on that glass, and is that something that we can either bring in-house or try and optimize pricing.

So we're continuing to work that route on the glass side and the frame side both.

And then beyond that, we're continuing to work the rest of the materials.

But a lot of this will just come from increased scale.

So -- and with scale, we get pricing power, and we get efficiency in our supply chain as well.

Your next question comes from Paul Coster with JPMorgan.

Yes.

So a couple of questions.

You see -- saw some revenue recognition slip to 2019, but you didn't raise the revenue numbers for '19.

And I'm wondering if it's something to do with PG&E and SC&E or whether it's supply constraints.

Perhaps you can just talk us through the puts and takes there as to why you didn't increase the 2019 revenue guidance.

And the other question I've got is the ramp cost seems to be increasing at least since the first guidance you gave for 2019.

What changed?

If you can just sort of talk us through the process by which we've gotten here.

Yes, sure.

So on the guidance piece, we've got a broad range in the guidance, as we talked about on Slide 12.

A significant amount of the revenue and margin is back ended for the year.

So obviously, that is in the fact that we have a guidance range, but you can see that small changes in timing could have large impacts to results on the back end of the year.

There is some risk around SCE.

When we think about SCE, I don't think it's a significant risk for us.

It's hard to evaluate.

You've got to look at what's happening with PG&E itself, how California and FERC and the bankruptcy courts will deal with that and then how that specifically applies to the facts and circumstances around SCE in their territory.

So we're monitoring that.

We do have assets that we're selling this year.

We have 3 assets in the -- we're currently running compared to the process for them.

We're seeing high demand for those.

If you look at SCE's credit today, the bonds still rate an investment grade.

You haven't seen -- the yields have widened incrementally.

We haven't seen a gap like you have on PG&E.

So I think we've got good confidence, but there is still risk around those processes.

So that's a piece of it.

But then the other piece is we just -- we're only 8 weeks into the year.

So it's early to make a change in terms of overall guidance.

We'll continue to evaluate the guidance as we go through 2019.

On the ramp piece specifically, all you're seeing is a change in geography from start-up moving into ramp, and it's a function of the timing of us bringing up the Vietnam factory.

So effectively, we've decreased start-up bringing that up early, but there's increased ramp.

And you see that in the 0.5 percentage point change in the gross margin guidance, and that's offset by a $15 million decrease in the start-up costs in the OpEx.

So those 2 net out to a 0 change to guidance.

It's just geography based on the timing of the Vietnam plant coming up.

Your next question comes from Michael Weinstein with Crédit Suisse.

This is Maheep Mandloi on behalf of Michael.

Given your shipment visibility, can we talk about how much of the third-party sales is fixed?

Or is that fixed versus floating prices for the year?

And the second question is on the Series 6 cost structure.

Can you talk about when you expect to achieve the target cost structure?

Is it still a Q4 target?

So as it relates to shipment visibility and the pricing, all of the -- anything that we recognize as a booking has a firm price associated with it.

The only impact that has is -- and we've referenced this before, if we deliver a bin that's higher than what we initially anchored towards, right, so the contract will say -- let's use as an example.

You have to deliver a 420-watt module.

We can go down 2 bins to 410, and we can go up 2 bins to 430 or we can average to a 420, whatever the math ends up working out to.

And those will be -- there'll be subtle price deltas as you move across.

In some cases, it's like $0.0025 for each bin.

In some cases, it's slightly higher than that.

So there can be slight movements in the realized ASP from what the center point of that contract is, but it's a firm fixed price.

So they all have firm fixed price.

There is no floating but for wherever the final delivery is of the product.

On the Series 6 cost structure, as we said in the last call, as we exit this year, we'll be within a couple of pennies from our targeted 40% cost reduction.

And that's important.

And when we get there, we still have an issue with the frames not fully optimized in the glass.

So we've got issues, and we've got a path on how to improve that.

And the other is we're not at the average efficiency that we had targeted for Series 6, right?

So we knew it was going to take us a couple of years, and we even showed a slide, I think, in the Analyst Day of kind of where that average efficiency would be.

And then -- and we showed more of a midterm objective of where we want to go with the real wattage for the product.

So a combination of optimizing around the glass, the frame and driving the efficiency, we will be in a much better position as we exit 2020, should be relatively in line with what our original targeted cost reduction was when we launched Series 6. And again, we launched it in November of 2016.

So it's only a little over 3 years since -- or 2 years, I guess, a little over 2 years.

We're not even 3 years into the journey.

So let's just put that in perspective, and I think there's tremendous progress that's been made over that horizon.

Your final question comes from Joseph Osha with JMP Securities.

I wanted to go back to the margin comments you made about the systems versus the module business.

In particular, the comments about Series 4. I understand that obviously, you've got more 6 allocated to your systems business.

But I'm wondering if there was any underloading on the 4 business that's weighing on those margins and also how might -- that might play out as you ramp the business down.

Yes.

So you're not seeing any underloading on the Series 4. What you're seeing is just the impacts of the fact that the Series 6 business is really still nearly all being allocated over to the systems segment from a revenue perspective and from a core comps perspective.

But you're seeing all of the ramp costs coming through in the module segment.

So you're seeing a blend of what looks like Series 4, but all are Series 6 kind of noncore costs coming through as well.

So that's what's happening there.

It's not a function of there being any underutilization on the S4 piece.

This concludes today's conference call.

You may now disconnect.