Earnings Call Transcript - FSLR Q2 2021

Operator, Operator

Good afternoon, everyone, and welcome to First Solar's Second Quarter 2021 Earnings Call. This call is being webcast live on the Investors section of First Solar's website at investor.firstsolar.com. I would now like to turn the call over to Mr. Mitch Ennis from First Solar Investor Relations. Mr. Ennis, you may begin.

Mitchell Ennis, Investor Relations

Thank you. Good afternoon, everyone, and thank you for joining us. Today, the company issued a press release announcing its second quarter 2021 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 2021. Following remarks, we'll open the call 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, including, among other risks and uncertainties, the severity and duration of the effects of the COVID-19 pandemic. 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?

Mark Widmar, CEO

Thank you, Mitch. Good afternoon, and thank you for joining us today. Beginning on Slide 3, I would like to start by thanking the First Solar team for their passion, continuing excellence, and many achievements in the second quarter. Operationally, we have started site preparation for the recently announced 3.3 gigawatt factory in Ohio, which will further cement our position as the largest PV module manufacturer in the Western Hemisphere. Additionally, I'm pleased to announce that contingent upon permitting and approval of government incentives satisfactory to First Solar, we are intending to invest approximately $680 million to add 3.3 gigawatts of manufacturing capacity in India. These next-generation factories represent a significant leap forward in our technology roadmap and will produce our most competitively advantaged modules with an expected lower cost per watt and environmental footprint compared to our existing fleet. Commercially, market demand for our CdTe technology is at a record level. Seven months into the year, we have already booked 9 gigawatts, exceeding our prior annual record of 7.7 gigawatts in 2017. From a technology standpoint, our production lines are manufacturing record modules. To illustrate this point, samples produced during our regular production process were submitted for external verification and confirmed by the National Renewable Energy Laboratory at a world record 19.2% glass area efficiency for a CdTe module. For reference and in comparison to our previous aperture area record of 19% efficiency, our new record equates to a 19.7% aperture area efficiency. Additionally, our advanced research team has been creating new optionality in our R&D roadmap. For example, we recently deployed prototypes of early-stage bifacial modules at a test facility and are pleased with the initial results. In summary, the momentum we have cultivated, paired with an increased favorable policy environment, represents a compelling growth opportunity in the near to midterm. However, before discussing these opportunities, I will first provide near-term COVID-19 supply chain cost and market updates. Please turn to Slide 4. As a global company with manufacturing operations in the United States, Malaysia, and Vietnam, the health and safety of our associates is our top priority, with a steadfast commitment to adhering to applicable COVID-19 protocols. As part of this effort, we are working with local governments to facilitate on-site testing and vaccination for our associates. I would also like to express immense gratitude to our Vietnam manufacturing associates who have to date elected to remain on site in order to maintain manufacturing continuity. While this clearly is a challenging time, we acknowledge your incredible resiliency, ingenuity, and leadership to deliver your operational plan commitments. While we have been permitted and able to maintain manufacturing operations in Malaysia and Vietnam to date, the rise of COVID-19 cases and potential government and other restrictions present risks to our production, supply chain, and technology implementation plans. As it relates to our CuRe program, the factory updates and tool implementations at our Vietnam site require international travel from both third-party equipment installers as well as our U.S.-based associates. But we continue to work with relevant agencies in Vietnam who support this essential travel in a safe manner. Delays resulting from government and other COVID-related restrictions, or an increase in case rates may impact the timing of our cure transition in Vietnam. Despite this uncertainty, we continue to execute and navigate the current environment as reflected by the manufacturing performance metrics on Slide 4. As highlighted previously, the global shipping environment remains challenging due to port congestion, limited container availability, an increase in cancellation of shipments by logistic providers, scheduled reliability issues, and other events. Since the April earnings call, shipping rates have continued to rise. Additionally, COVID-19 outbreaks and restrictions have caused disruptions in China and Southeast Asia, the impacts which have reverberated across the global logistics market. These challenges, coupled with strong global demand have led to a significant increase in the cost of transoceanic freight. We have partially mitigated the effects of higher shipping cost per watt through improvements in our module efficiency, implementation of Series 6 Plus, expansion of our distribution network strategy in the United States, and forward contracts. However, we have seen and expect to continue to see for the remainder of 2021 adverse impacts on our financial results. For context, spot rates for routes between Asia and the United States have increased 200% to 300% from Q2 2020 to Q2 2021. Over this period, sales rates reduced our module segment gross margin by 9 percentage points in Q2 of 2021 or 3 percentage points higher year-on-year. We continue to anticipate near-term challenges, including elevated fuel cost, average vessel delays of 2 weeks, and constrained container availability, impacting our ability to use space secured on vessels. Although these factors contribute to lower-than-anticipated shipments in Q2 and higher freight costs, we have a number of near-term and long-term strategies intended to improve our competitive position regarding sales. Near term, we are working closely with our customers to limit our exposure to inflated sales freight costs. In certain situations, we have accommodated requests for delayed module shipments, which provide opportunities to mitigate higher freight costs. Given current vessel schedule reliability, we are adding scheduled buffers to better meet our customers' commitments and provide greater resiliency in our shipment plan. Average sales freight from Malaysia and Vietnam to our U.S. customers increased $0.05 per watt quarter-on-quarter, ending Q2 was approximately triple that of shipments from Ohio. Long term, this reinforces the strategic thesis for locating additional manufacturing capacity near to demand. Contractually, for certain new bookings, we have employed structures that mitigate sales freight costs in excess of prenegotiated levels. As we continue to secure bookings for deliveries 2 to 3 years in the future, this type of contractual arrangement will help de-risk the expected value of our contracted backlog. I would next like to discuss the key components of our bill and material spend, approximately 2/3 of which is made up of glass and frame costs. From a glass perspective, we have largely hedged the cost through long-term fixed price agreements with domestic suppliers that have volumetric pricing benefits as we achieve higher levels of production. With regards to aluminum, in August of 2020, we entered into a commodity swap contract to hedge a portion of our U.S. cash flows for purchases of aluminum frames, which ends in Q4. While we anticipate some impacts as the hedge rolls off, we intend to partially mitigate the cost per watt impact through reduced aluminum for module uses, firstly, by differentiating between interior and exterior modules; and secondly, by redesigning the frame. Finally, the cost of lumber, which is used for our shipping and packing process, was approximately 70% higher on an index basis in Q2 compared to the start of the year. This impacted our Q2 results by approximately $2 million. Since then, lumber costs have significantly declined, and as a result, we currently do not expect it to impact our 2021 exit rate cost per watt target. In summary, while cost and uncertainties remain around bill of material items, we are tracking to achieve a 9% cost per watt produced reduction between where we ended 2020 and expect to end 2021. Note while core production costs are largely on track, the 2 percentage point decrease in our year-over-year cost per watt reduction relative to the previous expectation is largely due to the effects of higher inbound freight costs for raw materials. On a cost per watt sold basis, due to the challenging near-term sales rate environment, our revised year-over-year reduction target is 3%. Note, as a reminder, sales rate is included in our cost of sales, whereas many of our module peers report sales rate as a separate operating expense. For comparison purposes, we encourage you to consider this fact when benchmarking our module gross margin percentages relative to our peers. Turning to Slide 5, I would like to provide some context on the ASP trajectory for the year. As a reminder, 2 years ago, on the Q2 2019 earnings call, we indicated approximately 4 gigawatts of our 2021 module supply was booked or contracted subject to conditions precedent. In other words, a significant portion of the volumes sold this year had an ASP agreed to 2 years prior to model delivery. Heading into 2020 and 2021, we were largely sold out of our available supply for the forward year. As a result, we've had limited exposure to the spot market. We believe there is a strong strategic rationale for forward contracting deliveries in this manner, which provides value for both First Solar and our customers. From our perspective, contracting for future deliveries provides us confidence in our ability to sell through our expected supply and visibility into an expected profit per watt in a TV market that is typically highly price competitive. From our customers' perspective, these arrangements provide clarity and certainty of pricing, product availability, and delivery timing, enabling them to underwrite PPAs from a position of strength, with a lower risk to their expected project returns. Being able to provide certainty to both buyer and seller is a strategic differentiator for First Solar. From a U.S. policy perspective, both near and long-term pricing for all solar modules is also impacted by uncertainty over legislation related to forced labor in China, tariffs, manufacturing tax credits, investment tax credits, and other restrictions and incentives. Given the current lack of clarity over the form, structure, and duration of potential policy changes, the near-term and long-term impacts of these on both demand and pricing also remain uncertain. Moreover, this lack of clarity needs to be balanced with the significant capacity expansions announced by our competitors. From First Solar's perspective, we aim to continue to work with capable, well-financed counterparties that have high certainty in the quality and execution of projects. We also look to establish and maintain deep relationships and partnerships with our customers, delivering solutions at a fair pricing level that meets their needs and also enables attractive returns for First Solar relative to our expected future cost per watt. At the time of the previous earnings call, we indicated that the ASP across the volume of potential deliveries in 2022 was 11% lower than the volume to be shipped in 2021. Including our incremental bookings since the previous earnings call, the year-on-year decline is largely unchanged. Looking into 2023, we are very pleased with the demand and pricing we are seeing for our CdTe modules as we continue to drive to higher wattage and efficiency levels. Although there remains significant uncontracted volume to be booked, the ASP across the contracted volume for planned deliveries in 2023 is only 1% lower than that volume planned for 2022. Note, while we have yet to commence the sales process for our next-generation PV modules to be produced by our recently announced factories, they are expected to have ASP advantages due to their anticipated higher efficiency and superior balance of system cost per watt profile. In summary, as we have seen a significant increase in the desire to work with First Solar due to our demonstrated value proposition. While pricing negotiations in the market remain competitive, we continue to secure volume with customers that value our points, with the potential for ASP catalysts in the future. Relative to this objective, we are very pleased with our record year-to-date net bookings of 9 gigawatts, which includes 4.1 gigawatts since April's earnings call. After accounting for shipments of approximately 1.8 gigawatts during the second quarter, our future expected shipments would extend into 2024 or 17.2 gigawatts. Including our year-to-date bookings, we are largely sold out for 2021 and 2022, and have 3.4 gigawatts of planned deliveries in 2023 and 4 to 5 gigawatts in 2024. This long-term demand further supports the investment thesis behind our third Ohio factory and our first factory in India. Additionally, and as reflected on Slide 6, from an opportunities perspective, our pipeline of future opportunities remains robust. Note, our capacity expansion in India and the related increase in available supply to meet projected domestic demand expand our booking opportunities in the country. Accordingly, our potential bookings in India exceed 7 gigawatts. We'd also like to address the reported use of forced labor in the crystalline silicon PV manufacturing industry, highlighted by the recent withhold and release order issued by the U.S. Customs and Border Protection; the Xinjiang Supply Chain Business Advisory from the U.S. government; the Weaver Forced Labor Protection Act which passed the U.S. Senate with unanimous consent; and an investigation by the United Kingdom and other countries in the EU. Climate change is among the most pressing issues facing society today. Fortunately, the challenges of decarbonization of the global electric mix can largely be addressed with commercially available technologies, including solar, wind, energy storage, and green hydrogen. Unfortunately, the crystalline silicon supply chain is tainted by the purported use of forced labor and human rights abuses in China, which necessitates urgent action. However, it must be understood that our global collective response to forced labor does not need to conflict with long-term global climate objectives. While there are commercial solutions to ensure supply chain continuity, we've acknowledged the near-term supply challenges presented by the withholding release issue by the U.S. Customs and Border Protection. These challenges are exacerbated by the overly complex and opaque nature of crystalline silicon manufacturing. While the issue of forced labor represents an urgent ethical imperative that must be addressed, it also presents a strategic opportunity to drive change and an opportunity for the United States and like-minded nations to achieve energy security and technological independence through the promotion of a PV domestic manufacturing industry. Relative to this, we strongly support the proposed Solar Energy Manufacturing in America Act, introduced by Georgia Senator Jon Ossoff, co-sponsored by Senators Warnock, Bennett, and Stabenow. We believe that if enacted, it will help accelerate the transition to clean energy using domestically produced technology, support American energy independence, and create high-quality manufacturing jobs. By creating tax incentives for vertically integrated manufacturers and for each step of the crystalline supply chain, we can establish a level playing field where all PV technologies compete on their merits and establish a domestic capacity to support America's climate objectives. We believe the Biden-Harris administration has a unique opportunity to adopt a long-term industry policy for solar, which could include a mix of manufacturing tax credits and extension of the investment tax credit with a domestic content requirement among other strategies. Through a long-term strategic approach to policy, the administration has an opportunity to create an environment that fosters innovation for the next generation of PV. While legislative outcomes for U.S. infrastructure and solar remain uncertain, we are broadly encouraged by the legislative sentiment and willingness to support U.S. PV manufacturing to enable energy independence, security, and climate global imperatives. Turning to Slide 7. Looking forward, we believe strong demand for Series 6, a compelling technology roadmap, a strong balance sheet, and a largely fixed operating expense cost structure and an increasingly favorable policy environment for domestic PV manufacturing in the United States and India are catalysts as we evaluate capacity expansion. With respect to the United States, as announced in June, we are more than doubling our manufacturing capacity in the United States, adding 3.3 gigawatts at an implied CapEx per watt of approximately $0.20. This greenfield expansion financed by cash on hand represents an opportunity, unbound by the legacy Series 4 constraints, to optimize each parameter of the factory and product design. Accordingly, this enables us to develop a new product at the intersection of efficiency, energy yield, optimized form factor, cost competitiveness, and advantaged environmental attributes. Starting in 2023, this factory of the future is expected to commence production of our next-generation module, which is expected to lead the fleet in terms of efficiency, module wattage, cost per watt, and environmental footprint. Our next-generation modules, building upon our CuRe program, are expected to push the boundaries of our CadTel platform in several ways. Firstly, in the midterm, we anticipate this module can achieve efficiency in excess of 20% and with an optimized form factor enable module wattage in excess of our current midterm target. Secondly, we optimized the form factor anticipating benefits for balance of system cost per watt, and consequently, module ASP. Thirdly, through an optimization of the module's mounting interface and an increase in automation, this factory is expected to achieve a lower cost per watt produced than our existing fleet, despite being located in a higher cost labor market. Finally, by locating this factory domestically, we reduce our reliance on transoceanic freight costs and anticipate reducing sales freight per watt for U.S. deliveries. Our third factory in Ohio is expected to commence commercial production in the first half of 2023, scale to over 3 gigawatts of nameplate capacity by the end of the year, and 3.3 gigawatts in 2025. Internationally, we have been evaluating the expansion of our manufacturing presence in India. Our technology is uniquely advantaged in the market due to our temperature coefficient and spectral response advantages, which result in higher energy per watt installed compared to crystalline silicon due to the effects of heat and humidity. As we stated previously, we believe CuRe significantly increases our competitiveness against bifacial modules. The India PV market is predominantly monofacial due to generally low capital costs and the additional real estate needed for bifacial plants. However, given the expected lifetime energy benefit of our CuRe modules, we can achieve this with no increase in balance of system costs or other project costs, and we are well-positioned to capture the value of CuRe in the Indian market. We also applaud the steps India has taken to foster a healthy domestic PV manufacturing industry, including a combination of federal and state incentives and national barriers. This includes, among others, a $600 million production-linked incentive scheme with preference given to vertically integrated PV manufacturers who produce modules with an advantaged temperature coefficient. In addition to domestic incentives, India announced a solar tariff policy starting in April 2022, which includes 25% and 40% duties on imported modules respectively. Through its strategic approach, India has combined its clean energy targets with effective trade and industrial policy designed to enable self-sufficient domestic manufacturing and true energy security. As previously indicated, the factors in evaluating future capacity expansion include geographic proximity to solar demand where First Solar has a competitive advantage, which could mitigate freight-related costs. Secondly, the ability to export cost-competitively into other markets. Thirdly, cost-competitive labor, low energy costs, and low real estate costs. Fourthly, competitive supply chains for raw materials and components. Finally, domestic and international policies to ensure such expansion is well-positioned. In summary, we believe India meets these criteria. With the strong demand for our CdTe technology, we are eager to grow our manufacturing capacity to meet this market demand. With our expansion in the United States and India and optimization of our existing fleet, we anticipate our nameplate manufacturing capacity will double to 16 gigawatts in 2024, with the new factories combining 2 to 3 gigawatts of production in 2023. Moving on to technology, there were several noteworthy accomplishments since the previous earnings call. Firstly, following the implementation of Series 6 Plus in our 2 factories in Ohio, we are now consistently producing 450-watt modules in Ohio and Malaysia, increasing our fleet-wide average watt per module to 4.49 for July month-to-date. Secondly, our commercial production lines are manufacturing record modules, as previously discussed. Finally, our CuRe product has been certified as meeting UL and IEC standards, representing an achievement of the robust quality, reliability, and safety requirements. As we look to extend our advantages in the utility-scale market, we recently deployed prototypes of early-stage bifacial CadTel modules at a test facility and are pleased with the initial results, demonstrating real-world bifaciality. While this is only early-stage research, we believe there is a path to increase bifacial performance, which has the potential to improve upon our existing temperature coefficient, spectral response, partial shade, and long-term degradation energy advantages. As we've previously stated, we believe CuRe significantly increases our competitiveness against bifacial modules. By potentially unlocking CadTel bifacial capabilities, we have the opportunity to further improve our existing energy advantage in ground-mounted applications. In the residential and C&I markets, we recognize the value of high-efficiency, aesthetically pleasing, and domestically manufactured products. As stated previously, we continue to evaluate the prospects of leveraging the high band gap advantages of CadTel and a disruptive high-efficiency, low-cost tandem or multi-junction device. We strongly believe that a thin-film semiconductor is essential to achieving the highest-performing tandem PV modules and that CadTel, which benefits from many innovations of our technology roadmap and has a proven commercially scaled track record, is ideally placed to enable this leap forward in high-performance modules. In the midterm, we believe there is a path to achieve a 25% efficient multi-junction PV module. As we seek to grow our presence and competitive position in the residential and C&I markets, we believe this type of module has the potential to be disruptive and provide us with a competitive edge. I'll now turn the call over to Alex, who will discuss our second quarter financial results and 2021 guidance.

Alexander Bradley, CFO

Thanks, Mark. Before discussing our Q2 results and 2021 financial guidance, I'd like to reiterate our core operating principle of endeavoring to create shareholder value through a disciplined decision-making framework, balancing growth, liquidity, and profitability. As it relates to growth, we anticipate doubling our nameplate manufacturing capacity from approximately 8 gigawatts today to 16 gigawatts in 2024 through adding additional factories in Ohio and India, as well as optimizing our existing fleet. Beyond that, we continue to evaluate the potential for further expansion in the United States as the policy environment develops. While our liquidity position has been a strategic differentiator in an industry that has historically prioritized growth without regard to long-term capital structure, we anticipate we'll be able to continue to self-fund the capacity expansion and strategic investments in our technology, whilst maintaining a strong differentiated balance sheet, which we believe is a meaningful competitive differentiator. While the strength of our balance sheet provides this flexibility, as we expand internationally, we may elect to utilize debt to mitigate currency risk and optimize returns on our international expansion. As it relates to profitability, our technology and capacity roadmap are expected to enhance our long-term earnings potential. Despite a long-term PV industry trend of declining ASPs, we anticipate revenue growth through capacity expansion. From a pricing perspective, although there remains significant uncontracted volume yet to book, we're pleased with the pricing levels we've secured to date for 2023 deliveries, which in aggregate are only 1% lower than that of volume planned for delivery in 2022. From a margin perspective, continued progress towards our midterm cost objectives is expected to enhance our profit potential. Furthermore, we've yet to book 2023 volumes for our next-generation PV modules, which are expected to be both ASP advantageous due to their higher efficiency and optimized form factor, creating value for customers as well as cost per watt advantages. Combined with the benefits of locating supply near to demand and reducing the cost of sales rate, these factories are expected to increase gross margin per watt by approximately $0.01 to $0.03 relative to our existing fleet. Overall, we believe a combination of capacity growth, technology enhancements, and reducing our cost per watt, coupled with an operating cost structure that is 80% to 90% fixed, will drive meaningful contribution margin as we scale. Before reviewing our overall financial results for the quarter, I'll first discuss the legacy system license that benefited revenue and margin during the period. In 2014, we sold a project that was eligible for a 30% cash grant payment under Section 1603 of the American Recovery and Reinvestment Act. The indemnification arrangement in September of 2017 saw us indemnify the project poster following the underpayment of anticipated cash flow and proceeds by the U.S. government. In 2018, the project entity commenced legal action seeking full payment of the previously expected cash grants. In Q2 of this year, a settlement was reached pursuant to which the U.S. government made a payment in Q3 to the project entity, a portion of which we are entitled to. Accordingly, we recognized systems segment revenue of approximately $65 million during the quarter, which directly benefited gross margin. Starting on Slide 8, I'll cover the income statement highlights for the second quarter. Net sales in Q2 were $629 million, a decrease of $174 million compared to the prior quarter. The decrease in net sales was primarily due to the sale of the Sun Streams 2, 4, and 5 projects in the prior quarter, partially offset by the aforementioned settlement agreement. On a segment basis, our module segment revenue in Q2 was $543 million compared to $535 million in the prior quarter. Total gross margin was 28% in Q2 compared to 23% in Q1. The systems segment gross margin of $65 million was largely driven by the previously mentioned settlement agreement. Despite the aforementioned delays in certain module deliveries as well as higher-than-expected logistics costs, our Q2 module segment gross margin increased to 20% from 19% in the prior quarter. While we continue to navigate and partially mitigate the effects of the dislocated shipping market, higher freight costs impacted our financial results for the quarter. In Q2, sales rate totaled approximately $50 million or 9 percentage points of module gross margin. Along with module warranty expense of approximately $2 million, sales freight and warranty reduced our module gross margin by approximately 10 percentage points. As mentioned, we're in the process of implementing Series 6 Plus and CuRe in 2021, which requires downtime resulting in lower production and underutilization. In Q2, our module segment gross margin was impacted by $7 million of underutilization. In total, sales rate, module warranty, and underutilization impacted our Q2 module gross margin by approximately 11 percentage points. SG&A and R&D expenses totaled $60 million in the second quarter, a decrease of approximately $12 million compared to the prior quarter. In Q2, we had a $3 million reduction in expected credit losses benefiting SG&A expense. Production startup, which is included in operating expenses, totaled $2 million in Q2, a decrease of $10 million compared to the prior quarter. This decrease was driven by the start of commercial production at our second Series 6 factory in Malaysia in Q1. Q2 operating income was $110 million, which included depreciation and amortization of $66 million, $65 million related to the aforementioned settlement agreement, $9 million related to underutilization and production start-up expense and share-based compensation of $5 million. We recorded a tax expense of $20 million in the second quarter compared to $46 million in Q1. The decrease in tax expense for Q2 is largely attributable to lower pretax income. The combination of the aforementioned items led to second quarter earnings per share of $0.77 and $2.73 for the first 2 quarters of 2021 on a diluted basis. Next, turning to Slide 9, I'll discuss fixed balance sheet items and summary cash flow information. Our cash, cash equivalents, marketable securities, and restricted cash balance ended the quarter at $2.1 billion, an increase of $255 million compared to the prior quarter with several factors impacting our quarter-end cash balance. Firstly, in Q1, we sold certain restricted marketable securities associated with our module collection and recycling program for total proceeds of $259 million. We intend to reinvest these proceeds, at which point they will be considered restricted marketable securities, which are not included in our measure of total cash. Secondly, in early April, we received proceeds from the sale of our U.S. project development business. Finally, our operating cash flows during the quarter were partially offset by capital expenditures. Total debt at the end of the second quarter was $279 million, an increase of $22 million from the end of Q1. This increase is due to a loan drawdown on the credit facility for a Japanese systems project. As a reminder, all of our outstanding debt continues to be project-related and will come off the balance sheet when the corresponding project is sold. Our net cash position, which includes cash, cash equivalents, restricted cash, and marketable securities less debt, increased by $233 million to $1.8 billion as a result of the aforementioned factors. Net working capital in Q2, which includes noncurrent project assets and excludes cash and marketable securities, decreased by $176 million compared to the prior quarter. This decrease was primarily driven by the collection of proceeds from the sale of our U.S. project development business and an increase in current liabilities due to an increase in down payments from module customers. Net cash generated by operating activities was $177 million in the second quarter. Finally, capital expenditures were $91 million in the second quarter compared to $90 million in the prior quarter. Continuing on Slide 10, I'll next discuss 2021 guidance. Firstly, starting with our systems business. We recognized a $65 million benefit in Q2 related to the previously mentioned settlement agreement and have incorporated this into our systems revenue and gross margin guidance. Secondly, we're evaluating whether to continue holding our Losour Norte asset in Chile or pursue a sale of this project. The fee such a sale will require coordination with the project lenders and could result in an impairment charge in the future if we are unable to recover our net carrying value in the project. No impact from any possible sale of this project is included in our guidance for the year. As it relates to our module business, there are several key updates. As highlighted on the previous 2 earnings calls, we continue to anticipate elevated shipping costs in 2021. Despite near and long-term strategies to mitigate the impact, the cost of shipping has continued to rise since the April earnings call. As a result of elevated rates, port congestion, limited container availability, and schedule reliability issues, sales rates are expected to adversely impact our 2021 results by an incremental $60 million relative to our previous expectations. For the full year 2021, we anticipate sales rate and warranty will reduce our module segment gross margin by 10 to 11 percentage points, a 250 basis point increase from the previous earnings call. While we continue to manage our core manufacturing costs, we also anticipate a shipping-related variable cost headwind of approximately $20 million, primarily due to elevated inbound freight costs for raw materials. Additionally, Q2 shipments were lower than expected due to vessel delays, constrained customer container availability, and accommodating certain customer requests. We're currently tracking to achieve full year 2021 shipments of 7.6 to 8 gigawatts, which represents a 0.2 gigawatt decrease to the low end of the guidance range. We also acknowledge that the current logistics environment presents risk to our 2021 shipment plan. As it relates to capacity expansion, our recently announced factories in Ohio and India are anticipated to commence production in 2023 and increased 2021 capital expenditures by approximately $400 million. Related to this expansion, we anticipate incurring an additional $700 million of capital expenditures in 2022, with the remainder in 2023. With these factors in mind, we're updating our 2021 guidance as follows: our module segment revenue guidance of $2.4 billion to $2.55 billion represents a $50 million decrease to the low end of the guidance range to account for our current expectations on shipment timing. Our updated net sales guidance of $2.875 billion to $3.1 billion reflects an increase in systems revenue on both the high and low end due to the aforementioned settlement agreement. Additionally, we've increased the low end of our guidance systems range to account for clarity on project sale accounts. Our module segment gross margin guidance is $485 million to $535 million. While our previous guidance represented an $80 million reduction to the high end due to a $60 million increase in expected sales rate and $20 million increase in expected inbound rate. Revised low end also represents an $80 million decrease relative to our previous guidance due to a 0.2 gigawatt reduction in the low end of our shipments guidance and an increase in expected sales and inbound freight costs, which are partially offset by the risk accounted for in our previous guidance range. As a result of these factors, we anticipate our module gross margin will be approximately 20% to 21% for the full year. For the full year 2021, we anticipate sales rate and warranty will reduce our module segment gross margin by 10 to 11 percentage points. Additionally, we expect the impact of ramp underutilization and reduced throughput to total $41 million. Our systems segment gross margin guidance is $210 million to $225 million, reflecting a $65 million increase due to the aforementioned settlement agreement and a $15 million increase to the low end due to clarity on project sale economics. We anticipate that most of our remaining full year systems segment revenue and gross margin will be recognized in the fourth quarter of the year. Our revised total gross margin guidance is $695 million to $760 million, which reflects a $15 million decrease in the high end of the range. SG&A and R&D expenses of $265 million to $275 million, production start-up expense of $20 million to $25 million, and operating expenses of $285 million to $300 million combined are unchanged. We revised operating income guidance to a range of $545 million to $625 million and includes anticipated depreciation and amortization of $262 million, share-based compensation of $20 million, $61 million to $66 million related to ramp underutilization, reduced throughput and production start-up expense, and a gain on the sale of our U.S. project development and North American O&M businesses of $149 million. Turning to non-operating items, we expect interest income, interest expense, and other income to net negative $15 million, an increase of $5 million compared to our previous guidance due to higher net interest expense and foreign exchange losses. Our tax guidance of $100 million to $120 million is unchanged. Our revised earnings per share guidance is $4 to $4.60 per share. As a reminder, there are a number of items impacting our EPS guidance for 2021. Firstly, ramp underutilization, reduced throughput, and production start-up expense driven by factory upgrades are expected to contribute to a $0.50 EPS headwind in 2021. Secondly, these upgrades will require approximately 3 weeks of planned downtime across the fleet, which is expected to contribute to lower production. Finally, sales rate and inbound freight both remain significantly elevated compared to historic levels. Our capital expenditure guidance has increased by $400 million, driven by our recently announced expansion plan to a revised range of $825 million to $875 million. As a result of additional CapEx in 2021 and high logistics costs, we decreased our year-end 2021 net cash guidance to a revised range of $1.35 billion to $1.45 billion. Lastly, our shipment guidance is 7.6 to 8 gigawatts, representing a 0.2 gigawatt reduction to the low end of the guidance range. Turning to Slide 11, I will summarize the key messages from the call today. From a financial perspective, our net cash position of $1.8 billion remains strong, we delivered year-to-date EPS of $2.73, and we revised our 2021 EPS guidance to account for the current freight market. Operationally, we started flight preparation for our recently announced factory in Ohio and announced our manufacturing expansion into India. As a result of this expansion and optimization of our existing fleet, we anticipate our nameplate manufacturing capacity will reach 16 gigawatts in 2024. Finally, demand for Series 6 is at a record-high level with 9 gigawatts of year-to-date net bookings, which includes 4.1 gigawatts since the previous earnings call. And with that, we will conclude our prepared remarks and open the call for questions.

Operator, Operator

Our first question comes from Philip Shen from ROTH Capital Partners.

Philip Shen, Analyst

The first one is on Vietnam and Malaysia with the COVID situation there. I think, Mark, you mentioned that people are working hard and maybe even living at the facility to maintain utilization. Can you talk about how you expect utilization to trend ahead? Is there a risk for a shutdown of production at any point in time in the future? And how is this impacting your ability to roll out new updates and so forth? And then secondarily, in terms of bookings, you guys have had some nice bookings here. There's still a bunch available for 2023. I think you mentioned maybe 3 gigawatts. When do you expect that to possibly get booked? I mean, could we see that booked later this year? Or do you think that might carry into 2022?

Mark Widmar, CEO

Yes, Phil. So, obviously, we've got to comply with all the requirements of what's going on in both those countries. And in some cases, there have been restrictions around movement control orders in Malaysia. Fortunately, Malaysia has been deemed essential, which continues to allow us to operate, and we will ensure compliance with all local requirements. We've also started the process in both facilities to get our associates vaccinated. Most of our associates in both facilities have received the first shot, and we expect to provide the second shot soon. So that's helpful as well. Vietnam is trending more significantly in terms of challenges. Relative to their historical numbers, the increase in cases and fatalities is significant. The government has imposed requirements for running our factory that includes quarantining on-site. We have made accommodations for our associates there to quarantine, following a manageable schedule, allowing us to rotate associates through over time. The team has done an outstanding job to continue operations and hit performance metrics despite the challenges. As I sit here today, while we look across our supply chain—both in Malaysia and Vietnam—as well as our facilities, we are able to manage the current situation. However, if conditions continue to worsen, we'll need to assess and evaluate our ability to continue operations. It's a clearly challenging environment, but we have managed it well. As it relates to technology rollout, we highlighted earlier that the COVID situation is affecting our CuRe rollout. We had initially planned to roll out in Ohio first, then Malaysia, and finally Vietnam. Some upgrades in Malaysia have already been accommodated, enabling the CuRe product release when we start production. However, the necessary upgrades in Vietnam have yet to happen, and there are significant travel restrictions and quarantines in place, which raises the risk of delays in the CuRe rollout in Vietnam. Regarding bookings, we have about 3.4 gigawatts for 2023. With the two new factories, we anticipate adding close to 3 gigawatts of incremental volume in 2023. There is significant engagement with customers, and we have several large deals in the pipeline both in the U.S. and India. Currently, we have about 7 gigawatts of opportunities in India that we aim to execute. So, demand is strong, and we do not anticipate a significant slowdown in booking momentum in the coming quarters, with a strong second half of the year expected for booking 2023 and 2024.

Operator, Operator

Our next question comes from the line of Eric Lee from BOFA.

Unidentified Analyst, Analyst

Congrats on the bookings. For the bookings specifically, could you comment on the average ASPs as you look into '23? Is that still in the $0.20 - high $0.20 per watt range? And can you talk about where you're booking into that '23-plus timeframe?

Mark Widmar, CEO

Yes. So what we've indicated on the call is that our current bookings for 2023 are essentially flat. They've decreased by about 1%. Pricing for our current bookings and negotiations is generally firming. We observe this trend in both 2022 and 2023 deliveries. As we book more volume, we remain somewhat capacity constrained, even with the two new factories; the new volume won’t start shipping until 2023. Given that we have constrained capacity, it begins to firm up pricing in the market. We're satisfied with this development. As stated in our prepared remarks, we aim for a balanced perspective to reach an ASP that is attractive for both parties. The project economics must work while meeting our required returns. We have also been implementing modifiers in our contracts around shipping costs, meaning that if there are additional sales freight expenses, there would be a mechanism that reflects variable pricing to the customer. This factor is also added in our future bookings.

Operator, Operator

Our next question comes from the line of Ben Kallo from Baird.

Benjamin Kallo, Analyst

Could you talk about a little bit about what went into your guidance, the assumptions? The small change in the low end seems minimal. I just want to understand what assumptions went into there regarding shipping costs especially, and the timing of any other plant shutdowns or costs like that? My second question is just on the ASPs. What I heard you just say was that ASPs are up, where you last talked to us about your negotiations. Can you discuss the reasons for that, particularly if related to supply chain? You also mentioned a potential advantage in ASPs with the new technology. Can you provide more detail on that?

Alexander Bradley, CFO

Yes, Ben. Starting with the guidance, you're not seeing a shift on the low end of the guidance range. The impact of the settlement agreement, which contributed $65 million in revenue and flowed straight through to gross margin, is the reason you don’t see significant change. If you analyze the module side of gross margin, we are down about $15 million due to lower shipment volumes and approximately $65 million on freight. So, it significantly impacted revenue. The bottom line is that we had approximately $80 million in freight costs and $60 million in sales rates impacting unit costs, and about $20 million related to inbound shipping. We also had about $15 million as a risk. Thus, the net impact is approximately $65 million. But keep in mind that the $65 million settlement also benefited the revenue line. On ASPs, we are observing increasing ASPs, and we differentiate ourselves through our strong capabilities and technology. Moreover, recent uncertainties related to the crystalline silicon supply chain, both in the U.S. and abroad, has created anxiety amongst customers, making them eager to align with us to ensure their module supply chain is secure. First Solar operates independently from the Chinese crystalline silicon supply chain, which provides us with a different opportunity with customers. In addition to technology, the evolution of CuRe is enhancing our competitive edge in the market. Our upcoming products, set to be produced from our Perrysburg 3 factory and the factory in India, will have higher efficiency than the existing fleet and will be optimized accordingly. With all these facets, it creates incremental value beyond our current offerings, along with an expectation of obtaining an additional $0.01 to $0.03 in gross margin realization compared to the Series 6 Plus CuRe.

Operator, Operator

Our next question comes from the line of Brian Lee from Goldman Sachs.

Brian Lee, Analyst

I have two additional questions regarding modeling. First, can you provide insight into the cash flow trajectory for the next few years? What net cash balance do you feel comfortable with, and when do you expect to return to positive free cash flow? Is that expected in 2024? Given the significant capital expenditures related to the expansions in Ohio and India, I'm curious about what the ideal free cash flow trajectory might be. My second question is regarding operating expenses. Alex, you mentioned that a significant portion of the operating expenses is fixed, which we have observed over the years. However, we usually see start-up and production costs associated with new fabs. How should we anticipate these costs in 2022 and 2023 for Ohio and India?

Alexander Bradley, CFO

On the cash side, our prior guidance estimated a year-end cash balance of $1.8 million to $1.9 million, which is now adjusted to a midpoint of $1.4 million after considering the anticipated $400 million CapEx this year associated with the expansion. Over the next couple of years, we expect significant cash generation from operations with our existing six factories. However, I cannot pinpoint a minimum comfortable cash number. We should generate substantial organic cash flow, enabling financing for the construction of U.S. factories without raising debt. Still, there might be an opportunity to maximize the balance sheet's structuring, such as considering leveraging the factory in India. This might help optimize capital structure, better align revenue streams with expenses, and take advantage of some financing benefits for importing equipment. As for the OpEx aspect, we are working through the specifics but expect startup costs for the new factories to be higher due to their larger scale than previous facilities. Historically, factory setups cost around $1.2 million, increasing to about $1.5 to $1.6 million for the larger factories. Normally, planning for the larger fixed costs will imply about $30 million to $40 million for the new setups. For the Ohio factory, significant startup costs will likely hit in 2022, with around 60% to 70% occurring in that period with the remaining balancing out in 2023. The India factory will have a delay, so you might see 25% to 50% in 2022 and then 50% to 75% in 2023. It's important to remember that even with the integration of these new factories and the significant scaling we anticipate, approximately 80% to 90% of our operating costs are fixed. So, this should result in an increase in contribution margins as we expand our capacity.

Operator, Operator

This concludes today's conference call. Thank you again for participating. You may now disconnect.